1

2

BUEHLER® SUM-MET™

The Science Behind Materials PreparationA Guide to Materials Preparation & Analysis

3

BUEHLER® SUM-MET™ - The Science Behind Materials PreparationCopyright © 2004 BUEHLER LTD.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmit-ted in any form or by an means, electronic, mechanical, photocopying, recording or otherwise, withoutpermission of the copyright holder.

ISBN Number: 0-9752898-0-2

FN NUMBER: FN01255

Printed in the United States of America

First Printing: 1M0204

This book was produced digitally by Buehler Ltd.

TrademarksAll terms mentioned in this book that are known to be trademarks or service marks have been appropriatelycapitalized.

Warning and DisclaimerEvery effort has been made to make this book as complete and as accurate as possible, but no warrantyor fitness is implied. The information provided is on an “as is” basis. The authors and the publisher shallhave neither liability nor responsibility to any person or entity with respect to any loss or damages arisingfrom the information contained in this book.

4

Table of Contents ........................................... 4

Introduction .................................................... 6

Sampling ......................................................... 7

Goals of Specimen Preparation .................... 9

Method Development ..................................... 9

Sectioning ..................................................... 10

Abrasive Wheel Cutting ........................... 10

Precision Saws ....................................... 12

Mounting of Specimens ............................... 14

Clamp Mounting ...................................... 14

Compression Mounting ........................... 14

Castable Resins for Mounting ................. 15

Edge Preservation .................................. 17

Grinding ........................................................ 20

Grinding Media ........................................ 20

Grinding Equipment ................................. 22

Polishing ....................................................... 23

Mechanical Polishing ............................... 24

Electrolytic Polishing ............................... 24

Manual “Hand” Polishing ......................... 24

Automatic Polishing ................................. 25

Polishing Clothes ..................................... 25

Polishing Abrasives ................................. 27

Examples of Preparation Procedure .......... 29

The Traditional Method ............................ 29

Contemporary Methods ........................... 29

Procedures for Specific Materials .............. 32

Periodic Table of Elements ...................... 33

Light Metals: Al, Mg and Be ..................... 34

Aluminum ......................................... 34

Magnesium ...................................... 36

Beryllium .......................................... 37

Low Melting Point Metals:

Sb, Bi, Cd, Pb, Sn and Zn ................ 38

Refractory Metals: Ti, Zr, Hf, Cr, Mo,

Nb, Re, Ta, V and W ......................... 40

Titanium ........................................... 40

Zirconium and Hafnium .................... 41

Other Refractory Metals: Cr, Mo, Nb,

Re, Ta, V and W ............................... 43

Ferrous Metals ........................................ 45

Copper, Nickel & Cobalt .......................... 48

Copper ............................................. 45

Nickel ............................................... 49

Cobalt .............................................. 50

Precious Metals ...................................... 51

Thermally-Spray Coated Specimens ...... 53

Sintered Carbides ................................... 55

Ceramics ................................................. 57

Composites ............................................. 59

Printed Circuit Boards ............................. 61

Eletronic Materials ................................... 62

Polymers ................................................. 65

Etching .......................................................... 67

Etching Procedures ................................ 67

Selective Etching .................................... 68

Electrolytic Etching and Anodizing .......... 71

Heat Tinting ............................................. 71

Interference Layer Method ...................... 72

Commonly Used Etchants for

Metals and Alloys .................................... 73

Table of Contents

5

Table of Contents

Written by George F. Vander Voort, Director, Research & Technology, with contributions from Wase Ahmed, George Blann, Paula Didier, Don Fisher,Scott Holt, Janice Klansky and Gabe Lucas.

Light Optical Microscopy ............................ 78

The Light Microscope .............................. 79

Microscope Components ........................ 79

Resolution ............................................... 82

Depth of Field .......................................... 83

Imaging Modes ........................................ 83

Microindentation Hardness Testing ............ 87

The Vickers Test ..................................... 87

The Knoop Test ....................................... 88

Factors Affecting Accuracy, Precision

and Bias .................................................. 89

Automation .............................................. 91

Image Capture & Analysis ........................... 93

Acquisition ............................................... 93

Clarification ............................................. 96

Operator Interactive Measurements ....... 96

Automated Measurements ...................... 96

Thresholding ........................................... 96

Binary Operations ................................... 98

Common Applications ............................. 99

Laboratory Safety ....................................... 101

Laboratory Equipment ........................... 102

Personal Protective Equipment (PPE) .. 102

Chemicals, Storage and Handling ......... 103

Etchants ................................................ 103

Solvents ................................................ 104

Acids ..................................................... 106

Bases .................................................... 110

Other Chemicals ................................... 110

Summary ..................................................... 112

References .................................................. 113

Appendices ................................................. 115

ASTM Metallography Standards ............... 123

ISO Standards ............................................ 124

Other National Standards .......................... 125

Buehler Trademarks ................................... 128

Index ............................................................ 129

Worldwide Sales Offices ............................ 132

6

INTRODUCTION

Proper specimen preparation is essential if thetrue microstructure is to be observed, identi-fied, documented and measured. Indeed, whenmetallographers have difficulty seeing or mea-suring microstructures, these problems canalmost always be traced to improper specimenpreparation procedures, or failure to properlyexecute the preparation steps. When a speci-men is prepared properly, and the structureis revealed properly, or a phase or constituentof interest is revealed selectively and withadequate contrast, then the actual tasks ofexamination, interpretation, documentation andmeasurement are generally quite simple. Ex-perience has demonstrated that getting therequired image quality to the microscope is byfar the greatest challenge. Despite this, thespecimen preparation process is often treatedas a trivial exercise. However, specimen prepa-ration quality is the determining factor in thevalue of the examination. This is in agreementwith the classic computer adage, “garbage in =garbage out.”

Historically, people tend to divide specimenpreparation (after mounting) into two steps,grinding and polishing, further sub-divided,somewhat arbitrarily, into coarse and finestages. Years ago, these divisions were quiteclear. Both processes involve the same mecha-nism, chip formation by micro-machining. Grindingis abrasive material removal from the specimensurface using randomly oriented, fixed abrasiveson a substrate such as paper or cloth. The abra-sive size range has generally been in the rangeof about 150 to 5 μm, with SiC paper used exclu-sively until recently.

Historically, polishing has been defined in termsof industrial polishing rather than metallo-graphic polishing, simply stating that it is asmoothing of the surface to a high luster. A pol-ished surface was defined as “a surface thatreflected a large proportion of incident light in aspecular manner.” These definitions are true asfar as they go; but they are inadequate for met-allography. Polishing also produces micro-machining action, but scratch depths are shal-lower in polishing, as the abrasives are smaller.They are not fixed initially to a substrate, but areapplied as a paste, slurry or spray. After a shorttime, the abrasive particles become fixed in the

cloth producing cutting action. The cloths used inpolishing exhibit a range of resilience dependingon whether they are non-woven polymeric orchemo-textiles, woven napless materials, ornapped materials. Resilience influences edgeretention, relief control and cutting rate (“aggres-siveness”). Polishing must remove the damageremaining after grinding, with the final step reduc-ing the damage to where it is insignificant, or canbe removed by the etchant. In this way, the truemicrostructure is revealed. For most metals andalloys, and many non-metallic materials, polish-ing down to a 1-, or even a 3-μm diamond finishmay yield a surface suitable for routine examina-tion, although finer abrasives are usually neededfor more critical work, such as failure analysis.

Historically, polishing was performed with abra-sive slurries, most commonly using alumina, butwith limited use of MgO, Cr2O3 or CeO for spe-cific materials. Diamond abrasives wereintroduced in the late 1940s. Cloths such as can-vas, billiard, felt, and cotton are not used as muchtoday as they tend to promote relief and edgerounding, or other artifacts. Modern preparationmethods have focused on reducing the numberof SiC grinding steps and replacing SiC paperwith other abrasives and surfaces, althoughthere are some materials that respond better togrinding with a series of finer and finer SiC pa-pers. But, most materials can be prepared withone grinding step, often called planar grindingwhen an automated system is used, and two ormore polishing steps. The second step uses rela-tively hard, non-resilient surfaces with a relativelycoarse diamond size (15- or 9-μm, for example)and the abrasive is added as a slurry or spray.Initially, the particles move between the specimensurface and the substrate. Technically, this is lap-ping. But, after a short time, the diamond particlesbecome embedded in the surface and producecutting action. We tend to call this polishing, al-though the historic definitions are not as clear here.If the substrate is a rigid grinding disk, the actionmay be more correctly called grinding than pol-ishing, as the surface finish will be dull. The sameabrasive when used on a hard, woven surface suchas an UltraPol silk cloth, will yield a much morelustrous appearance, but with less stock removaland shallower scratches. Napped cloths, if theyare used, are generally restricted to the final stepwith the finest abrasives.

Introduction

7

SAMPLING

The specimens selected for preparation must berepresentative of the material to be examined.Random sampling, as advocated by statisticians,can rarely be performed by metallographers. Agood exception is the testing of fasteners where aproduction lot can be sampled randomly. But, alarge forging or casting cannot be sampled at ran-dom as the component would be renderedcommercially useless. Instead, test locations arechosen systematically based on ease of sampling.Many material specifications dictate the samplingprocedure. In failure analysis studies, specimensare usually removed to study the origin of the fail-ure, to examine highly stressed areas, to examinesecondary cracks, and so forth. This, of course,is not random sampling. It is rare to encounter ex-cessive sampling as testing costs are usuallyclosely controlled. Inadequate sampling is morelikely to occur.

Normally, a specimen must be removed from alarger mass and then prepared for examination.This requires application of one or moresectioning methods. For example, in a manu-facturing facility, a piece may be cut from a barof incoming metal with a power hacksaw, or anabrasive cutter used dry, i.e., without a coolant.The piece is then forwarded to the laboratorywhere it is cut smaller to obtain a size moreconvenient for preparation. All sectioningprocesses produce damage; some methods(such as flame cutting or dry abrasive cutting)produce extreme amounts of damage. Tradi-tional laboratory sectioning procedures, usingabrasive cut-off saws, introduce a minor amountof damage that varies with the material beingcut and its thermal and mechanical history. It isgenerally unwise to use the original face cut inthe shop as the starting point for metallographicpreparation as the depth of damage at this loca-tion may be quite extensive. This damage mustbe removed if the true structure is to be revealed.However, because abrasive gr inding andpolishing steps also produce damage, wherethe depth of damage decreases with decreas-ing abrasive size, the preparation sequencemust be carefully planned and performed; oth-erwise, preparation-induced artifacts will beinterpreted as structural elements.

Many metallographic studies require more thanone specimen and sectioning is nearly alwaysrequired to extract the specimens. A classicexample of multiple specimen selection isthe evaluation of the inclusion content of steels.One specimen is not representative of the wholelot of steel, so sampling becomes important.ASTM standards E 45, E 1122 and E 1245 giveadvice on sampling procedures for inclusion stud-ies. To study grain size, it is common to use asingle specimen from a lot. This may or may notbe adequate, depending upon the nature of thelot. Good engineering judgment should guidesampling, in such cases. In many cases, a prod-uct specification may define the procedurerigorously. Grain structures are not alwaysequiaxed and it may be misleading to selectonly a plane oriented perpendicular to the de-formation axis, a “transverse” plane, for such astudy. If the grains are elongated due to pro-cessing, which does happen, the transverseplane will usually show that the grains areequiaxed in shape and smaller in diameter thanthe true grain size. To study the effect of defor-mation on the grain shape of wrought metals, aminimum of two sections are needed — oneperpendicular to, and the other parallel to, thedirection of deformation.

Sampling

HELPFUL HINTS FOR

SAMPLING

When developing a samplingplan for an unfamiliar part or component, de-termine the orientation of the piece relativeto the original wrought or cast starting ma-terial and prepare sections on thelongitudinal, transverse and planar surfaces,or radial and transverse surfaces, to revealthe desired information. Remember that themicrostructure may look much more homo-geneous than it is on a transverse planecompared to a longitudinal or planar surface.If the starting material was cold worked, ornot fully recrystallized after hot working, grainstructures will appear to be equiaxed andsmaller on a transverse plane, but non-equiaxed and larger on a longitudinal orplanar surface.

8

GOALS OF SPECIMENPREPARATION

The preparation procedure and the prepared speci-men should have the following characteristics toreveal the true microsctructure*:

• Deformation induced by sectioning, grinding andpolishing must be removed or be shallow enoughto be removed by the etchant.

• Coarse grinding scratches must be removed;very fine polishing scratches may be tolerablefor routine production work.

• Pullout, pitting, cracking of hard particles, smear,and other preparation artifacts, must be avoided.

• Relief (i.e., excessive surface height variationsbetween structural features of different hard-ness) must be minimized; otherwise portionsof the image will be out of focus at high magni-fications.

• The surface must be flat, particularly atedges (if they are of interest) or they cannotbe imaged.

• Coated or plated surfaces must be kept flat ifthey are to be examined, measured or photo-graphed.

• Specimens must be cleaned adequatelybetween preparation steps, after preparation,and after etching.

• The etchant chosen must be either general orselective in its action (reveal only the phase orconstituent of interest, or at least producestrong contrast or color differences betweentwo or more phases present), depending uponthe purpose of the investigation, and must pro-duce crisp, clear phase or grain boundaries,and strong contrast.

If these characteristics are met, then the true struc-ture will be revealed and can be interpreted,measured and recorded. The preparation methodshould be as simple as possible, should yield con-sistent high quality results in a minimum of timeand cost and must be reproducible.

Preparation of metallographic specimens [1-3]generally requires five major operations: (a) sec-tioning, (b) mounting (optional), (c) grinding,(d) polishing and (e) etching (optional).

Method DevelopmentThe methods presented in this book use times thatare conservative so that the vast majority of com-positions of alloys or materials of a specific typecan be prepared to yield the true microstructurewithout artifacts or observable scratches. Theywere developed using an 8-inch (200 mm) diam-eter platen system with six 1.25-inch (30 mm)diameter mounted specimens. For each category,except beryllium, a wide variety of specimens withdifferent processing histories were prepared us-ing the stated methods.

For any metal or material category, there arecompositions or processing conditions that makethose specimens easier, or more difficult, to pre-pare compared to the “average” specimen. Ingeneral, even the most difficult specimens of agiven type can be prepared perfectly using thesemethods with no more than one etch and re-polish (repeat last step) cycle. For perfect coloretching results, it may be necessary to followthe cycle with a brief vibratory polish, depend-ing upon the material.

When working with easier to prepare specimens,or for routine work where the same degree ofperfection is not required, the user can reducethe times, or eliminate one or two steps, or thefinal step, and get satisfactory results. Modifica-tions to the recommended procedures are left tothe user based upon their knowledge of the de-gree of difficulty in preparing their specimens andthe desired quality of preparation.

Goals of Specimen Preparation

* Not all of these apply to every specimen.

9

SECTIONING

Bulk samples for subsequent laboratory section-ing may be removed from larger pieces usingmethods such as core drilling, band or hack-sawing, flame cutting, or similar methods.However, when these techniques are used, pre-cautions must be taken to avoid alteration of themicrostructure in the area of interest. Labora-tory abrasive-wheel cutting is recommended toestablish the desired plane of polish. In the caseof relatively brittle materials, sectioning may beaccomplished by fracturing the specimen at thedesired location.

Abrasive-Wheel CuttingThe most commonly used sectioning device inthe metallographic laboratory is the abrasivecut-off machine, Figure 1. All abrasive-wheelsectioning should be performed wet. An ampleflow of coolant, with an additive for corrosionprotection and lubrication, should be directed intothe cut. Wet cutting will produce a smoothsurface finish and, most importantly, will guardagainst excessive surface damage causedby overheating. Abrasive wheels should beselected according to the manufacturer’srecommendations. Table 1 summarizes Buehler’srecommendations for our machines. Specimensmust be fixtured securely during cutting, andcutting pressure should be applied carefully toprevent wheel breakage. Some materials, suchas CP (commercial purity) titanium, Figure 2, arequite prone to sectioning damage.

Wheels consist of abrasive particles, chiefly alu-mina or silicon carbide, and filler in a bindermaterial that may be a resin, rubber, or a mix-ture of resin and rubber. Alumina (aluminumoxide) is the preferred abrasive for ferrous al-loys and silicon carbide is the preferred abrasivefor nonferrous metals and minerals. Wheelshave different bond strengths and are recom-mended based on the suitability of their bondstrength and abrasive type for the material to besectioned. In general, as the hardness of a ma-terial increases, abrasives become dull morequickly, and the binder must break-down andrelease the abrasives when they become dullso that fresh abrasive particles are available tomaintain cutting speed and efficiency. Conse-quently, these wheels are called “consumable”wheels because they wear away with usage. Ifthey do not wear at the proper rate, dull abra-sives will rub against the region being cutgenerating heat and altering the existing truemicrostructure. If this heat becomes excessive,it can lead to grain or particle coarsening, soft-ening or phase transformations, and in extremecase, to burning or melting. Different materialshave different sensitivities to this problem. But,the need to balance the wheel break-down ratewith the hardness of the piece being sectioned,produces the various recommendations listedfor cutting different materials and metals withdifferent hardnesses, such as steels.

The size of the cut-off machine also affects wheelparameters. As the diameter of the wheel isincreased, to permit sectioning of larger speci-mens, the wheel thickness is generally increased.Even so, for a given diameter wheel, there maybe a range of thicknesses available. Thickerwheels are stronger but remove more material inthe cut. This may, or may not, be a problem. But,

Sectioning

Figure 1. Delta Automatic Orbital Cutter

Figure 2. Cutting damage (top) and a “burr” aftersectioning of an annealed CP titanium specimen(mod. Weck’s reagent, 100X, polarized light plussensitive tint).

1 0

thicker wheels do generate more heat during cut-ting than a thinner wheel, everything else beingheld constant. Consequently, for cases where thekerf loss or heat loss must be minimized, selectthe thinnest available wheel of the proper bondstrength and abrasive. Buehler’s AcuThin Cut-offWheels offer the thinnest possible wheels for deli-cate abrasive sectioning. Buehler also hasdiamond cut-off blades with resin bonded orRimlock1 metal bonded blade forms in sizes from8-inch (203 mm) to 12-inch (305 mm) diameters.Resin-bonded diamond blades are ideal for cut-ting very hard cemented carbide specimens;Rimlock or continuous rim blades are recom-mended for cutting rocks and petrographicspecimens.

Historically, the most common cutter design hasbeen the so-called “chop” cutter. Basically, theblade is attached to a motor and the operatorpulls a handle to drive the blade downward intothe work piece. Because of the design, the blademoves through an arc as it is pulled downward,Figure 3a. For efficient cutting, the piece must

be oriented to minimize the contact area with thewheel. For small parts, this is generally easy todo, but for larger parts, it may not be possible toorient the specimen properly for optimal cutting.When a round wheel is used to cut a round bar inchop mode, the contact area is initially very small.As the cut progresses, the cut widens until themaximum diameter is reached. After this, the con-tact area decreases until sectioning is completed.Under application of a constant load, the pressureon the abrasive particles in the cut decreases asthe contact area increases. If the pressure appliedto the grains is inadequate for cutting, then heat isgenerated which may not be dissipated by the cool-ant, which causes deformation damage, phasechanges and possibly burning or melting.

A chop cutter may be set up to pulse as the loadis applied, that is, the wheel is fed into thespecimen, and then feeding is halted momen-tarily, Figure 3b. Pulsing causes the wheel to bestress shocked which removes both dull andsharp abrasive from the wheel as new abrasiveis exposed to the cut. While a better cut may be

1Rimlock is a registered trademark of Felker Operations.

Sectioning

Table 1. Buehler’s Abrasive Cutting Recommendations

Available DiametersRecommended Use Bond Abrasive (Inches) (mm)

General Usage Blades

Tools Steels 60 HRC and Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,Above Carburized Steels 14, 16, 18 350, 400, 455

Hard Steel 50 HRC Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,14, 16, 18 350, 400, 455

Medium Hard Steel Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,35-50 HRC 14, 16, 18 350, 400, 455

Soft or Annealed Steel Rubber Al2O3 9, 10, 12, 230, 250, 300,15-35 HRC 46-90 HRB 14, 16, 18 350, 400, 455

Medium Hard Nonferrous Materials, Rubber SiC 9, 10, 12, 230, 250, 300,Uranium, Titanium, Zirconium 14, 16, 18 350, 400, 455

Soft Nonferrous Materials Rubber SiC 9, 10, 12, 230, 250, 300,Aluminum, Brass, etc. 14, 16, 18 350, 400, 455

Superalloys Rubber Al2O3 10, 12, 250, 300,14, 16, 18 350, 400, 450

Thin Blades to Minimize Kerf Loss and Cutting Deformation

Tool, Hard Steel, Rubber Al2O3 5, 9, 10*, 12 130, 230,≤ 45 RC 250*, 300

Medium Hard, Soft Steel Rubber Al2O3 5, 7, 9, 130, 180, 230,≥ 45 HRC 10, 12, 14 250*, 300, 350

Hard or Soft Nonferrous Materials Rubber SiC 7 180

*Rubber Resin BondRefer to Buehler’s Consumables Buyers Guide for ordering information, and exact dimensions of arbor sizes, outer diameter, and thickness ofBuehler Metabrase, Delta, and Acu-Thin Abrasive Cut-off Wheels

1 1

Wheel Path

Sample

Chop Cutting with PulsingWheel contact still governed by specimensize. The pulsing action gives a shock tothe wheel causing abrasives to breakaway. Wheel wear is generally high withthis method.

3b

Chop CuttingThe traditional form of machineoperation. Wheel contact arc isgoverned by specimen size. Generallya struggle with large/difficult parts.

3a

OscillationThe wheel is reciprocated over asmall distance. This minimizes the contactarc. Available only on manual labor-intensive cutters.

3c

Traverse and IncrementThe wheel contact arc can be preciselycontrolled via depth increment. Thetraverse stroke must always exceed thepart length to avoid a change in wheelcontact arc area. Machine needs to be setfor each part. Action is slow.

3d

OrbitalSimilar in action to traverse and incrementbut on a curved path. Simpler and quickerin operation. Part size is irrelevant as theorbital action produces a minimum contactarc area during cutting.

3e

CUTTING STYLE AND WHEEL PATH

obtained, cutting time and wheel wear areincreased.

Optimal cutting, with the least damage to the speci-men, is obtained by keeping the pressure on eachabrasive particle in the wheel constant and as lowas possible by keeping the contact area betweenwheel and specimen constant and low. This ideais called the minimum area of contact cutting(MACC) method. Oscillation style cutters, suchas the OscilliMet, minimize the contact area byoscillating perpendicular to the cut contact area,Figure 3c. This approach has the advantage of

permitting larger size pieces to be cut properlycompared to a chop cutter with the same wheeldiameter.

Another option for MACC is the “traverse and in-crement” cutter design, Figure 3d. A strip of materialis removed by moving either the wheel or the workpiece perpendicular to the cut direction. After this,the cutter must move back to the side where thecut began. The wheel is dropped down a fixed in-crement and another strip is removed. In thisapproach, the wheel must move to the end of thestroke, return to the starting point, and then be

Figure 3. Illustration of cutter types and the wheel path during sectioning

Sectioning

1 2

moved an increment downwards in each cycle.Consequently, cut times are longer. The wheeltraverse length must be greater than the speci-men width. This machine must be programmed foreach size part to obtain cutting.

Orbital cutting, Figure 3e, combines the bestfeatures of chop, oscillation and traverse-and-increment cutters, while minimizing cutting time.As the wheel is fed into the specimen, the arboris also rotated in a small ellipse planar with thewheel permitting the wheel to cut a small strip ofmaterial from the specimen. But no time is lost,as with the traverse and increment cutter, as thewheel is ready to cut on the original side. Withthe elliptical orbit and a controlled feed, theconditions for minimum area of contact cuttingare obtained, irrespective of work piece size.Special programming is not needed. The orbitalcutting concept is available in the Delta cutter fromBuehler. These also feature Smartcut, which willsense motor overloading during a cut, if it occurs,and reduce the feed rate automatically. When it issafe to return to the original feed rate, Smartcutwill do so.

Precision SawsPrecision saws, Figure 4, are commonly used inmetallographic preparation to section materialsthat are small, delicate, friable, extremely hard,or where the cut must be made as close aspossible to a feature of interest, or where the cutwidth and material loss must be minimal. As thename implies, this type of saw is designed tomake very precise cuts. They are smaller in sizethan the usual laboratory abrasive cut-off sawand use much smaller blades, typically from 3-to 8-inch (7.6 to 20.3 mm) in diameter. Theseblades can be of the non-consumable type, madeof copper-based alloys and copper plated steelwith diamond or cubic boron nitride abrasivebonded to the periphery of the blade; or, theycan be consumable blades using alumina orsilicon carbide abrasives with a rubber-basedbond. Blades for the precision saws are muchthinner than the abrasive wheels used in anabrasive cutter and the load applied during cuttingis much less. Consequently, less heat is gener-ated during cutting and damage depths arereduced. While pieces with a small section size,that would normally be sectioned with an abrasive

cutter can be cut with a precision saw, the cuttingtime will be appreciably greater but the depth ofdamage will be much less. Precision saws arewidely used for sectioning sintered carbides, ce-ramic materials, thermally sprayed coatings,printed circuit boards, electronic components,bone, teeth, etc. Table 2 lists selection criteria forprecision saw blades. Buehler’s blades for preci-sion saws are available with different abrasivesizes and bonds to provide optimum cutting for awide variety of applications.

Sectioning

Figure 4. IsoMet 5000 Linear Precision Saw

HELPFUL HINTS FOR

SECTIONING

When cutting a diff icultspecimen with the recommended consum-able abrasive wheel, if the cutting action hasbecome very slow, pulse the applied force.This will help break down the abrasive bond-ing, exposing fresh, sharp abrasives toenhance the cutting action. However, if youare using a resin- bonded diamond blade tocut cemented carbides, or other very hardmaterials, do not pulse the applied force, asthis will shorten wheel life.

1 3

Table 2. Selection Criteria for Precision Saw Blades

3″ x 0.006″ 4″ x 0.012″ 5″ x 0.015″ 6″ x 0.020″ 7″ x 0.025″ 8″ x 0.035″(75mm x (100mm x (125mm x (150mm x (180mm x (200mm x

BLADE SERIES 0.2mm) 0.3mm) 0.4mm) 0.5mm) 0.6mm) 0.9mm)

Diamond Wafering Blades

Series 30HC Diamond, for use X** X** Xwith plastics, polymers, and rubber

Series 20HC Diamond, for aggressive X* X Xgeneral sectioning of ferrous andnonferrous materials includingtitanium alloys

Series 15HC Diamond, for general X X X X X Xuse with ferrous and nonferrousalloys, copper, aluminum, metalmatrix, composites, PC boards,thermal spray coatings, andtitanium alloys

Series 20LC Diamond, for use with X* Xhard/tough materials structuralceramics, boron carbide, boronnitride, and silicon carbide

Series 15LC Diamond, for use with X X X X X Xhard/brittle materials structuralceramics, boron carbide, boronnitride, and silicon carbide

Series 10LC Diamond, for use with X X X* Xmedium to soft ceramics, electronicpackages, unmounted integratedcircuits, GaAs, AIN and glass fiberreinforced composites

Series 5LC Diamond, for use with X Xsoft friable ceramics, electronicpackages, unmounted integratedcircuits, composites with finereinforcing media, CaF2, MgF2,and carbon composites

IsoCut Wafering Blades

Cubic Boron Nitride (CBN) abrasiveblades work well for many toughmaterials giving significantly shortercut times

For iron, carbon steels, high alloy X X X X X Xsteels, cobalt alloys, nickel super-alloys, and lead alloys

General Usage Abrasive Cut-off Bond/Wheels 0.03″ (0.8mm) thick Abrasive

For ferrous materials, stainless R/Al2O3 Xsteels, cast irons, andthermal spray coatings

For tough nonferrous metals, R/SiC Xaluminum, copper, titanium,uranium, zirconium

AcuThin Abrasive Cut-off Bond/Wheels 0.019″ (0.5mm) thick Abrasive

For sectioning small, delicatespecimens or where minimaldeformation and kerf loss is theprimary concern

Tool, hard steel, ≥ 45 HRC R/Al2O3 X

Medium hard, soft steel ≤ 45 HRC R/Al2O3 X

* Alternate blade thickness of 0.020″ (0.5 mm) ** Alternate blade thickness of 0.030″ (0.8 mm)

For a complete listing of Buehler consumable supplies for use with the Isomet Precision Saws, please refer to Buehler’s Consumables Buyers Guide.

Sectioning

1 4

MOUNTING OF SPECIMENS

The primary purpose of mounting metallographicspecimens is for convenience in handling speci-mens of difficult shapes or sizes during thesubsequent steps of metallographic preparationand examination. A secondary purpose is to pro-tect and preserve extreme edges or surfacedefects during metallographic preparation. Themethod of mounting should in no way be injuriousto the microstructure of the specimen. Pressureand heat are the most likely sources of injuriouseffects.

Phenolic plastics were introduced to metallogra-phy in 1928 for encapsulating specimens by hotcompression mounting. Prior to that time, speci-mens were prepared unmounted, or mounted inflours of sulfur, wax or low-melting point alloys,such as Wood’s metal. These “mounting com-pounds” were not without their problems.Introduction of polymers was a big improvementover these methods. Subsequently, many poly-mers were evaluated for use as mountingcompounds, as they were introduced to the mar-ket. Development of castable resins in the 1950’sadded new resins to the metallographer’s toolchest. The simplicity of mounting without using apress, and their low curing temperatures, madecastable resins an attractive alternative.

Clamp MountingClamps have been used for many years tomount cross sections of thin sheet specimens.Several specimens can be clamped convenientlyin sandwich form making this a quick, convenientmethod for mounting thin sheet specimens. Whendone properly, edge retention is excellent, andseepage of fluids from crevices between speci-mens does not occur. The outer clamp edgesshould be beveled to minimize damage to pol-ishing cloths. If clamps are improperly used sothat gaps exist between specimens, fluids andabrasives can become entrapped and will seepout obscuring edges and can cause cross con-tamination. This problem can be minimized byproper tightening of clamps, by use of plasticspacers between specimens, or by coating speci-men surfaces with epoxy before tightening.

Compression MountingThe most common mounting method usespressure and heat to encapsulate the specimenwith a thermosetting or thermoplastic mountingmaterial. Common thermosetting resins includephenolic (PhenoCure resin), diallyl phthalate andepoxy (EpoMet resin) while methyl methacrylate(TransOptic resin) is the most commonly usedthermoplastic mounting resin. Table 3 lists the char-acteristics of the hot compression mounting resins.

Mounting of Specimens

Table 3. Characteristics of Buehler’s Compression (Hot) Mounting Resins

Best Edge Retention; Near ZeroVery Low Shrinkage; Best Edge electrical

Least Fine Particle Size; Retention; Very Resistance;Materials Expensive Fills Smallest Crevices Low Shrinkage SEM - EDS/WDS Clear

Ceramics PhenoCure EpoMet F EpoMet G ProbeMet TransOptic

Steels PhenoCure EpoMet F EpoMet G ProbeMet TransOptic

Plated PhenoCure EpoMet F EpoMet G ProbeMet TransOpticLayers

Aluminum PhenoCure ProbeMet TransOptic

Copper/ PhenoCure ProbeMet TransOpticBrass

Black, Red, Black Black Bronze TransparentColor or Green

Temperature 300°F/150°C 300°F/150°C 300°F/150°C 300°F/150°C 300°F/180°C

4200 psi/ 4200 psi/ 3000-4200 psi/ 4200 psi/ 3800 psi/Pressure 290 bar 290 bar 210-290 bar 290 bar 260 bar

1 5

Both thermosetting and thermoplastic materialsrequire heat and pressure during the moldingcycle; but, after curing, mounts made of thermo-plastic resins must be cooled under pressure toat least 70 °C while mounts made of thermoset-ting materials may be ejected from the mold at themaximum molding temperature. However, coolingthermosetting resins under pressure to near am-bient temperature before ejection will significantlyreduce shrinkage gap formation. Never rapidlycool a thermosetting resin mount with water afterhot ejection from the molding temperature. Thiscauses the metal to pull away from the resin pro-ducing shrinkage gaps that promote poor edgeretention, see Figure 5, because of the differentrates of thermal contraction. EpoMet resin, a ther-mosetting epoxy, provides the best edge retention,Figure 6, of these resins and is virtually unaffectedby hot or boiling etchants while phenolic resinsare badly damaged.

Mounting presses vary from simple laboratoryjacks with a heater and mold assembly to full au-tomated devices, as shown in Figure 7. Anadvantage of compression mounting is productionof a mount of a predicable, convenient size andshape. Further, considerable information can beengraved on the backside – this is always moredifficult with unmounted specimens. Manual “hand”polishing is simplified, as the specimens are easy

to hold. Also, placing a number of mounted speci-mens in a holder for semi- or fully-automatedgrinding and polishing is easier with standardmounts than for unmounted specimens. Mountedspecimens are easier on the grinding/polishing sur-faces than unmounted specimens.

Castable Resins for MountingCold mounting materials require neither pre-ssure nor external heat and are recommendedfor mounting specimens that are sensitive to heatand/or pressure. Acrylic resins, such as VariDurand SamplKwick resins, are the most widely usedcastable resins due to their low cost and shortcuring time. However, shrinkage can be aproblem with acrylics. Epoxy resins, althoughmore expensive than acrylics, are commonlyused because epoxy will physically adhere tospecimens, have low shrinkage, and can bedrawn into cracks and pores, particularly if avacuum impregnation chamber, Figure 8, isemployed and a low viscosity epoxy, such asEpoThin resin, is used. Epoxies are very suitablefor mounting fragile or friable specimens and cor-rosion or oxidation specimens. Dyes or fluorescentagents, such as FluorMet dye, may be added to

Figure 5. Edge retention of this improperly carbur-ized 8620 alloy steel was degraded by a shrinkagegap between the specimen and the phenolic mount:a) top, 500X; b) bottom, 1000X (2% nital).

b

a

Mounting of Specimens

Figure 6. Excellent edge retention of a borided42CrMo4 alloy steel specimen mounted in EpoMetresin (1000X, 2% nital).

Figure 7. SimpliMet 3000 Automatic Mounting Press

1 6

epoxies for the study of porous specimens suchas thermal spray coated specimens. Most epoxiesare cured at room temperature, and curing timescan vary from 2 to 8 hours. Some can be cured atslightly elevated temperatures in less time, as longas the higher temperature does not adverselyaffect the specimen. Table 4 lists the characteris-tics of Buehler’s castable resins.

Cast epoxy resins provide better edge retentionthan cast acrylic resins, mainly due to the betteradhesion of epoxy to the specimen and theirlower shrinkage. Acrylics usually do not bond tothe specimen and, because they shrink moreduring curing, a gap is inevitability formed be-tween specimen and mount. When a shrinkagegap is present, edge retention is usuallypoor. To improve edge retention with castablemounts, the specimen can be plated, for example,with electroless nickel using the EdgeMet kit, or

Flat Edge Filler particles can be added to the resin.To obtain electrical conductivity, Conductive Fillerparticles can be added to the castable resin.

When preparing castable resin mounts, particu-larly epoxy mounts by manual (“hand”) methods,the metallographer will observe that the surfacetension between the mount and the workingsurface is much higher than with a compressionmount. This can make holding the mount morechallenging. If automated devices are used, themetallographer may hear “chatter” (noise) dur-ing rough grinding due to the greater surfacetension. The chatter can be reduced or stoppedby changing to contra mode (head and platen ro-tate in opposite directions.)

Acrylics do generate considerable heat duringcuring and this can be strongly influenced bythe molding technique used. Nelson [4] mea-sured the exotherm produced by polymerizingan acrylic resin using two procedures: a glassmold on a glass plate (insulative) and an alumi-num mold on an aluminum plate (conductive).Polymerization produced a maximum exothermof 132°C using the insulative approach but only42°C using the conductive approach. Note that132°C is not much less than the 150°C tem-perature used in hot compression mounting!Nelson also measured the exotherm producedwhen an epoxy resin was cured in a phenolic ringform placed on a pasteboard base. Although this

Table 4. Characteristics of Buehler’s Castable Resins

Peak Shore D Cure RecommendedName Type Temperature Hardness* Time Mold Comments

EpoThin Epoxy 80 °F 78 Best Edge 9 Any Low viscosity, low shrinkage,(27 °C) Retention hours transparent, best for

vacuum impregnation

EpoxiCure Epoxy 82 °F 82 Best Edge 6 Any Moderate hardness,(formerly (28 °C) Retention hours transparent,Epoxide) low shrinkage

EpoKwick Epoxy 185 °F 82 Good 90 Sampl-Kup Faster epoxy, some(85 °C) Edge Retention minutes shrinkage, transparent

EpoColor Epoxy 175 °F 82 Good Edge 90 Sampl-Kup Dye-enhanced epoxy,(79 °C) Retention minutes displays red under darkfield

and polarized light

VariDur Acrylic 170 °F 85 Fair Edge 10 Any Fast cure, harder acrylic(formerly (77 °C) Retention minutes opaque, abrasive resistantUltraMount)

SamplKwick Acrylic 175 °F 80 Poor 5-8 Any Very fast cure, translucent, (79 °C) Edge Retention minutes some shrinkage

* Hardness differences appear negligible but abrasion resistance has a significant effect on edge rounding

Mounting of Specimens

Figure 8. Vacuum Impregnation Equipment

1 7

was an insulative approach, the maximum tem-perature during polymerization was only 7 °C, avast improvement over the acrylics.

Nelson’s work applies to specific acrylic andepoxy resins molded upon specific conditions.While the epoxy that he used exhibited a lowexotherm, this does not imply that all epoxyresins will exhibit such low exotherms in poly-merization. Epoxy resins that cure in short timeperiods develop much higher exotherms, that canexceed that of acrylic resins. In addition to thespeed of curing of the epoxy resin, other factorsdo influence the magnitude of the exothermduring polymerization. The larger the mass ofepoxy in the mount, the faster it will set and thegreater the exotherm. Indeed, very large mountscan generate enough heat to crack extensively.Heating the resin makes it less viscous andspeeds up curing, also generating more heatduring polymerization. The mold material alsocan influence curing time and temperature.For example, EpoxiCure cures fastest inSamplKup plastic molds, slower in phenolic ringforms, and still slower in the reuseable rubbermounting cups. Consequently, the exotherm willbe greater when using the SamplKup type moldand lowest when using the rubber mountingcups. All of these factors must be considered ifthe exotherm must be minimized.

Edge PreservationEdge preservation is a classic metallographicproblem and many “tricks” have been promoted(most pertaining to mounting, but some to grind-ing and polishing) to enhance edge flatness.These methods [2] include the use of backupmaterial in the mount, the application of coatingsto the surfaces before mounting or the additionof a filler material to the mounting resin. Plating[2] of a compatible metal on the surface to beprotected (electroless nickel, deposited usingthe EdgeMet Kit has been widely used) is gener-ally considered to be the most effective procedure.However, image contrast at an interface betweena specimen and the electroless nickel may be in-adequate for certain evaluations. Figure 9 showsthe surface of a specimen of 1215 free-machiningsteel that was salt bath nitrided. One specimenwas plated with electroless nickel; both weremounted in EpoMet G. It is hard to tell where the

nitrided layer stops for the plated specimen, Fig-ure 9a, which exhibits poor image contrastbetween the nickel and the nitrided surface. Thisis not a problem for the non-plated specimen, Fig-ure 9b.

Introduction of new technology has greatlyreduced edge preservation problems [5,6]. Mount-ing presses that cool the specimen to near ambienttemperature under pressure produce much tightermounts. Gaps that form between specimen andresin are a major contributor to edge rounding, asshown in Figure 5. Staining at shrinkage gaps mayalso be a problem, as demonstrated in Figure 10.Use of semi-automatic and automatic grinding/pol-ishing equipment, rather than manual (hand)preparation, increases surface flatness and edgeretention. To achieve the best results, however,the position of the specimen holder, relative to theplaten, must be adjusted so that the outer edge ofthe specimen holder rotates out over the edge ofthe surface on the platen during grinding and pol-ishing, particularly for 8-inch diameter (200 mm)platens. The use of harder, woven or non-woven,napless surfaces for polishing with diamond abra-sives (rather than softer cloths such as canvas,billiard and felt) maintains flatness. Final polishingwith low nap cloths for short times introduces very

Mounting of Specimens

Figure 9. Example of (a, top) poor coating edgevisibility due to a lack of contrast between theprotective nickel plating and the salt bath nitridedsurface (arrow) of 1215 free-machining carbon steel;and, (b, bottom) good contrast and visibilitybetween EpoMet resin and nitrided surface (arrow)plus excellent edge retention (1000X, 2% nital).

Ni

b

a

1 8

little rounding compared to use of higher nap, softercloths.

These procedures will produce better edgeretention with all thermosetting and thermoplasticmounting materials. Nevertheless, there are stilldifferences among the polymeric materials usedfor mounting. Thermosetting resins provide betteredge retention than thermoplastic resins. Of thethermosetting resins, diallyl phthalate provides littleimprovement over the much less expensive phe-nolic compounds. The best results are obtainedwith EpoMet G an epoxy based thermosetting resinthat contains a filler material. For comparison, Fig-ure 11 shows micrographs of the nitrided 1215specimen mounted in a phenolic resin (Figure 11a),

and in methyl methacrylate (Figure 11b), at 1000X.These specimens were prepared in the samespecimen holder as those shown in Figure 9, butneither displays acceptable edge retention at1000X. Figure 12 shows examples of perfect edgeretention, as also demonstrated in Figure 9.

Very fine aluminum oxide spheres have beenadded to epoxy mounts to improve edge reten-tion, but this is really not a satisfactory solution asthe particles are extremely hard (~2000 HV) andtheir grinding/polishing characteristics are incom-patible with softer metals placed insidethe mount. Recently, a soft ceramic shot (~775HV) has been introduced that has grinding/polish-ing characteristics compatible with metallic

b

a

Mounting of Specimens

Figure 10. Staining (thin arrows) due to etchantbleed out from a shrinkage gap (wide arrows)between the phenolic mount and the M2 high speedsteel specimen (500X, Vilella’s reagent).

Figure 11. Poor edge retention (arrows) at the saltbath nitrided surface of 1215 free-machining carbonsteel mounted in (a, top) phenolic resin and in (b,bottom) methyl methacrylate and polished in thesame holder as those shown in Figure 9 (1000X, 2%nital).

b

a

Figure 12. Excellent edge retention for (a, top)complex coated (arrow) sintered carbide insert(1000X, Murakami’s reagent) and for (b, bottom) ironnitrided (arrow points to a white iron nitride layer;an embedded shot blasting particle is to the left ofthe arrow) H13 hot work die steel specimen (1000X,2% nital). EpoMet was used in both cases.

Figure 13. Flat Edge Filler shot was added toEpoxicure resin to improve the edge retention ofthis annealed H13 hot work die steel specimen(500X, 4% picral).

1 9

specimens placed in the mount. Figure 13 showsan example of edge retention with the Flat EdgeFiller soft ceramic shot in an epoxy mount.

Following are general guidelines for obtaining thebest possible edge retention. All of these factorscontribute to the overall success, although someare more critical than others.

• Properly mounted specimens yield better edgeretention than unmounted specimens, as round-ing is difficult, if not impossible, to prevent at afree edge. Hot compression mounts yield betteredge preservation than castable resins.

• Electrolytic or electroless Ni plating (e.g., withthe EdgeMet Kit) of the surface of interestprovides excellent edge retention. If thecompression mount is cooled too quickly af-ter polymerization, the plating may be pulledaway from the specimen leaving a gap. Whenthis happens, the plating is ineffective for edgeretention.

• Thermoplastic compression mounting mate-rials are less effective than thermosetting resins.The best thermosetting resin for edge retentionis EpoMet G, an epoxy-based resin containinga hard filler material.

• Never hot eject a thermosetting resin afterpolymerization and cool it quickly to ambient(e.g., by cooling it in water) as a gap will formbetween specimen and mount due to the dif-ferences in thermal contraction rates. Fullyautomated mounting presses cool the mountedspecimen to near ambient temperature underpressure and this greatly minimizes gap forma-tion due to shrinkage.

• Automated grinding/polishing equipment pro-duces flatter specimens than manual “hand”preparation.

• Use the central force mode (defined later in thetext) with an automated grinder/polisher as thismethod provides better flatness than individualpressure mode (defined later in the text).

• Orient the position of the smaller diameter speci-men holder so that, as it rotates, its peripheryslightly overlaps the periphery of the larger di-ameter platen.

• Use PSA-backed SiC grinding paper (if SiC isused), rather than water on the platen and aperipheral hold-down ring, and PSA-backedpolishing cloths rather than stretched cloths.

• UltraPrep metal-bonded or resin-bonded dia-mond grinding discs produce excellent flatsurfaces for a wide variety of materials.

• Use “hard” napless surfaces for rough polish-ing (until the final polishing step), such as TexMet1000, UltraPol or UltraPad cloths, and fine pol-ishing, such as a TriDent cloth. Use a napless,or a low- to medium-nap cloth, depending uponthe material being prepared, for the final stepand keep the polishing time brief.

• Rigid grinding discs, such as the ApexHerculesH and S discs, produce excellent flatness andedge retention and should be used wheneverpossible.

Mounting of Specimens

HELPFUL HINTS FOR

MOUNTING

Epoxy is the only resin thatwill physically adhere to a specimen. If itsviscosity is low, epoxy can be drawn intopores and cracks by vacuum impregnation.Acrylics are too viscous for vacuum impreg-nation.

Castable resins are sensitive to shelf life,which can be extended by keeping them ina refrigerator. It is a good practice to dateyour containers when you get them.

A protective nickel plating can be separatedfrom the specimen if you eject the speci-men after the curing cycle and cool it quickly.

2 0

Other materials have also been used both forthe planar grinding stage or, afterwards, toreplace SiC paper. For very hard materialssuch as ceramics and sintered carbides, one, ormore, metal-bonded or resin-bonded diamonddisks (the traditional type) with grit sizes fromabout 70- to 9-μm can be used. The traditionalmetal- or resin-bonded diamond disc has dia-mond spread uniformly over its entire surface.An alternate type of disc, the UltraPrep disc, hasdiamond particles applied in small spots to the disksurface, so that surface tension is lessened.UltraPrep metal-bonded discs are available in sixdiamond sizes from 125- to 6-μm while UltraPrepresin bonded discs are available in three diamondsizes from 30- to 3-μm. Another approach uses astainless steel woven mesh UltraPlan cloth on aplaten charged with coarse diamond, usually inslurry form, for planar grinding. Once planar sur-faces have been obtained, there are severalsingle-step procedures available for avoiding thefiner SiC papers. These include the use of plat-ens, thick woven polyester cloths, silk, or rigidgrinding disks. With each of these, an intermedi-ate diamond size, generally 9- to 3-μm, is used.

Grinding MediaThe grinding abrasives commonly used in thepreparation of metallographic specimens aresilicon carbide (SiC), aluminum oxide (Al2O3),emery (Al2O3 - Fe3O4), composite ceramics anddiamond. Emery paper is rarely used today inmetallography due to its low cutting efficiency.SiC is more readily available as waterproofpaper than aluminum oxide. Alumina papers,such as PlanarMet Al 120-grit paper, do have abetter cutting rate than SiC for some metals [3].These abrasives are bonded to paper, polymericor cloth backing materials of various weights inthe form of sheets, discs and belts of varioussizes. Limited use is made of standard grindingwheels with abrasives embedded in a bondingmaterial. The abrasives may be used also inpowder form by charging the grinding surfaceswith the abrasive in a premixed slurry or sus-pension. SiC particles, particularly with the finersize papers, embed readily when grinding softmetals, such as Pb, Sn, Cd and Bi (see Figure14). Embedding of diamond abrasive is also aproblem with these soft metals and with aluminum,but mainly with slurries when napless cloths areused, see Figure 15.

GRINDING

Grinding should commence with the finest grit sizethat will establish an initially flat surface and re-move the effects of sectioning within a few minutes.An abrasive grit size of 180 to 240 (P180 to P280)is coarse enough to use on specimen surfacessectioned by an abrasive cut-off wheel. Hack-sawed, bandsawed, or other rough surfacesusually require abrasive grit sizes in the range of120- to 180-grit. The abrasive used for each suc-ceeding grinding operation should be one or twogrit sizes smaller than that used in the precedingstep. A satisfactory fine grinding sequence mightinvolve SiC papers with grit sizes of 220- or240-, 320-, 400-, and 600-grit (P240 or P280,P400, P600 and P1200). This sequence is usedin the “traditional” approach.

As with abrasive-wheel sectioning, all grindingsteps should be performed wet provided thatwater has no adverse effects on any constitu-ents of the microstructure. Wet grinding minimizesspecimen heating, and prevents the abrasive frombecoming loaded with metal removed from thespecimen being prepared.

Each grinding step, while producing damageitself, must remove the damage from the previ-ous step. The depth of damage decreases withthe abrasive size but so does the metal removalrate. For a given abrasive size, the depth ofdamage introduced is greater for soft materialsthan for hard materials.

For automated preparation using a multiple-specimen holder, the intital step is called planargrinding. This step must remove the damage fromsectioning while establishing a common plane forall of the specimens in the holder, so that eachspecimen is affected equally in subsequent steps.Silicon carbide and alumina abrasive papers arecommonly used for the planar grinding step andare very effective. Besides these papers, thereare a number of other options available. Oneoption is to planar grind the specimens with aconventional alumina grinding stone. Thisrequires a special purpose machine, as the stonemust rotate at a high speed, ≥1500 rpm, to cuteffectively. The stone must be dressed regularlywith a diamond tool to maintain flatness, dam-age depth is high and embedding of aluminaabrasive in specimens can be a problem.

Grinding

2 1

P280-P1200), and the electrical resistancemethod for very fine grit sizes). The grit sizenumbering systems differ above 180- (P180) grit,but equivalent sizes can be determined usingTable 5. As with many standards, they are notmandatory and manufacturers can, and do,make some of their papers to different mean par-ticle sizes than defined in these specifications.There is a philosophical difference in the two sys-tems. ANSI/CAMI papers use a wider particle size

CarbiMet SiC paper is made in the USAaccording to the ANSI/CAMI standard (B74.18-1996) while SiC papers manufactured inEurope are made according to the FEPAstandard (43-GB-1984, R 1993). Both standardsuse the same methods for sizing the abrasivesand the same standards to calibrate thesedevices (sieving for the coarsest grits, sedimen-tation grading for intermediate grits (240-600 or

Table 5. European/USA Equivalency Grit Guide

European/USA Equivalency Grit Guide

FEPA (Europe) ANSI/CAMI (USA)

Grit Number Size(μμμμμm) Grit Number Size(μμμμμm) Emery Grit

P60 269.0 60 268.0

P80 201.0 80 188.0

P100 162.0 100 148.0

P120 127.0 120 116.0

P180 78.0 180 78.0 3

P240 58.5 220 66.0 2

P280 52.2 240 51.8

P320 46.2

P360 40.5 280 42.3 1

P400 35.0 320 34.3 0

P500 30.2 360 27.3

P600 25.8 400 22.1 00

P800 21.8

P1000 18.3 500 18.2 000

P1200 15.3 600 14.5

P1500 12.6 800 12.2 0000

P2000 10.3 1000 9.2

P2500 8.4 1200 6.5

P4000* 5.0*

The chart shows the midpoints for the size ranges for ANSI/CAMI graded paper according to ANSI standard B74.18-1996 and for FEPA gradedpaper according to FEPA standard 43-GB-1984 (R1993). The ANSI/CAMI standard lists SiC particles sizes ranges up to 600 grit paper. For finergrit ANSI/CAMI papers, the particles sizes come from the CAMI booklet, Coated Abrasive (1996).

*FEPA grades finer than P2500 are not standardized and are graded at the discretion of the manufacturer. In practice, the above standard valuesare only guidelines and individual manufactures may work to a different size range and mean value.

Figure 15. 6μm diamond particles (arrows)embedded in lead (1000X).

Figure 14. SiC grit particle (arrow) embedded in a6061-T6 aluminum weldment (500X, aqueous 0.5%HF).

Grinding

2 2

on rotating “wheels”; that is, motor-driven platens,Figure 18, upon which the SiC paper is attached.

Lapping is an abrasive technique in which theabrasive particles roll freely on the surface of acarrier disc. During the lapping process, the discis charged with small amounts of a hard abrasivesuch as diamond or silicon carbide. Lapping diskscan be made of many different materials; cast ironand plastic are used most commonly. Lapping pro-duces a flatter specimen surface than grinding,but it does not remove metal in the same manneras grinding. Some platens, referred to as laps, arecharged with diamond slurries. Initially the diamondparticles roll over the lap surface (just as with othergrinding surfaces), but they soon become embed-ded and cut the surface producing microchips.

distribution (centered on the mean size) than FEPApapers. A broader size range allows cutting to be-gin faster at lower pressures than with a narrowersize range, so less heat is generated and lessdamage results. However, the broader size rangedoes produce a wider range of scratch depths;but, these should be removed by the next step inthe preparation sequence. Generation of less dam-age to the structure is considered to be moreimportant than the surface finish after a particulargrinding step, as it is the residual damage in thespecimen that may prevent us from seeing the truemicrostructure at the end of the preparation se-quence.

Grinding EquipmentStationary grinding papers, often used by students,but uncommon in industrial use, are supplied instrips or rolls, such as for use with the HandiMet2 roll grinder. The specimen is rubbed against thepaper from top to bottom. Grinding in one directionis usually better for maintaining flatness than grind-ing in both directions. This procedure can be donedry for certain delicate materials, but water is usu-ally added to keep the specimen surface cool andto carry away the grinding debris.

Belt grinders, such as the SurfMet I or the DuoMetII belt grinders, Figure 16, are usually presentin most laboratories. An alternative approach isto use a high speed disc grinder, such as theSuperMet grinder, Figure 17. These devices areused with coarse abrasive papers from 60- to 240-grit, and are mainly used for removing burrs fromsectioning, for rounding edges that need not bepreserved for examination, for flattening cutsurfaces to be macro-etched, or for removing sec-tioning damage. Manual grinding work is performed

Figure 17. SuperMet grinder

Grinding

Figure 16. DuoMet II belt grinder

Figure 18. Beta grinder/polisher

2 3

POLISHING

Polishing is the final step, or steps, in producing adeformation-free surface that is flat, scratch free,and mirror-like in appearance. Such a surface isnecessary to observe the true microstructure forsubsequent metallographic interpretation, bothqualitative and quantitative. The polishing techniqueused should not introduce extraneous structuressuch as disturbed metal (Figure 19), pitting (Fig-ure 20), dragging out of inclusions, “comet tailing”

(Figure 21), staining (Figure 22) or relief (heightdifferences between different constituents, or be-tween holes and constituents (Figure 23 and 24).Polishing usually is conducted in several stages.Traditionally, coarse polishing generally was con-ducted with 6- or 3-μm diamond abrasives chargedonto napless or low-nap cloths. For hard materi-als, such as through hardened steels, ceramics

and cemented carbides, an additional coarse pol-ishing step may be required. The initial coarsepolishing step may be followed by polishing with1-μm diamond on a napless, low nap, or mediumnap cloth. A compatible lubricant should be usedsparingly to prevent overheating or deformationof the surface. Intermediate polishing should beperformed thoroughly so that final polishing maybe of minimal duration. Manual, or “hand” polish-ing, is usually conducted using a rotating wheelwhere the operator rotates the specimen in a cir-cular path counter to the wheel rotation direction.

Mechanical PolishingThe term “mechanical polishing” is frequently usedto describe the various polishing procedures in-volving the use of fine abrasives on cloth. The clothmay be attached to a rotating wheel or a vibratory

Figure 23. Examples of freedom from relief (a, top)and minor relief (b, bottom) at the edges of largeprimary hypereutectic silicon particles in an as-castAl-19.85% Si specimen (200X, aqueous 0.5% HF).

Figure 21. “Comet tails” at large nitrides in anannealed H13 hot work die steel specimen (200X,DIC).

Figure 22. Staining (arrow) on the surface of a ofTi-6% Al-2% Sn-4% Zr-2% Mo prepared specimen(200X).

a

Polishing

Figure 19. (top) Preparation damage (arrows) inannealed CP titanium (500X, DIC, Kroll’s reagent).Figure 20. (bottom) Pitting (arrow) on the surface ofa cold-drawn brass (Cu-20% Zn) specimen (100X)

2 4

Figure 24. Differences in relief control in a brazedspecimen containing shrinkage cavities: a) (top)poor control; and, b) (bottom) good control (100X,glyceregia).

polisher bowl. Historically, cloths have been ei-ther stretched over the wheel and held in placewith an adjustable clamp on the platenperiphery, or held in place with a pressure sensi-tive adhesive (PSA) bonded to the back of the cloth.If a stretched cloth moves under the applied pres-sure during polishing, cutting will be less effective.If an automated polishing head is used, stretchedcloths are more likely to rip, especially if un-mounted specimens are being prepared. Inmechanical polishing, the specimens are held byhand, held mechanically in a fixture, or merelyconfined within the polishing area, as with theVibroMet 2 polisher.

Electrolytic PolishingElectrolytic polishing can be used to preparespecimens with deformation free surfaces. Thetechnique offers reproducibility and speed. Inmost cases, the published instructions for elec-trolytes tell the user to grind the surface to a600-grit finish and then electropolish for about 1-to 2-minutes. However, the depth of damage aftera 600- (P1200) grit finish may be severalmicrometres but most electropolishing solutionsremove only about 1-μm per minute. In this case,the deformation will not be completely removed.In general, electropolished surfaces tend to bewavy rather than flat and focusing may be

difficult at high magnifications. Fur ther,electropolishing tends to round edges associatedwith external surfaces, cracks or pores. In two-phase alloys, one phase will polish at a differentrate than another, leading to excessive relief. Insome cases, one phase may be attacked prefer-entially and inclusions are usually attacked.Consequently, electrolytic polishing is not recom-mended for failure analysis or image analysiswork, except possibly as a very brief step at theend of a mechanical polishing cycle to removewhatever minor damage may persist.Electropolishing has been most successfulwith soft single-phase metals and alloys, particu-larly where polarized light response needs to bemaximized.

Manual “Hand” PolishingAside from the use of improved polishing clothsand abrasives, hand-polishing techniques stillfollow the basic practice established manyyears ago:

1. Specimen Movement. The specimen is heldwith one or both hands, depending on theoperator’s preference, and is rotated in a direc-tion counter to the rotation of the polishing wheel.In addition, the specimen is continually movedback and forth between the center and the edgeof the wheel, thereby ensuring even distributionof the abrasive and uniform wear of the polish-ing cloth. (Some metallographers use a smallwrist rotation while moving the specimen fromthe center to the edge of one side of the wheel.)After each step, the specimen is rotated 45 to90° so that the abrasion is not unidirectional.

2. Polishing Pressure. The correct amount ofapplied pressure must be determined by experi-ence. In general, a firm hand pressure is appliedto the specimen.

3. Washing and Drying. The specimen iswashed by swabbing with a liquid detergent so-lution, rinsed in warm running water, then withethanol, and dried in a stream of warm air. Alco-hol usually can be used for washing when theabrasive carrier is not soluble in water or if thespecimen cannot tolerate water. Ultrasonic clean-ing may be needed if the specimens are porousor cracked.

b

a

Polishing

2 5

4. Cleanness. The precautions for cleanness, aspreviously mentioned, must be strictly observedto avoid contamination problems. This involves thespecimen, the metallographer’s hands, and theequipment.

Automated PolishingMechanical polishing can be automated to a highdegree using a wide variety of devices rangingfrom relatively simple systems, Figure 25, to rathersophisticated, minicomputer, or microprocessorcontrolled devices, Figure 26. Units also vary incapacity from a single specimen to a half dozenor more at a time and can be used for all grinding

face flatness and edge retention. There are twoapproaches for handling specimens. Central forceutilizes a specimen holder with each specimen heldin place rigidly. The holder is pressed downwardagainst the preparation surface with the force ap-plied to the entire holder. Central force yields thebest edge retention and specimen flatness. If theresults after etching are inadequate, the speci-mens must be placed back in the holder and theentire preparation sequence must be repeated.Instead of doing this, most metallographers willrepeat the final step manually and then re-etch thespecimen.

The second method utilizes a specimen holderwhere the specimens are held in place loosely.Force is applied to each specimen by a piston,hence the term “individual force” for this ap-proach. This method provides convenience inexamining individual specimens during the prepa-ration cycle, without the problem of regainingplanarity for all specimens in the holder on thenext step. Also, if the etch results are deemedinadequate, the specimen can be replaced in theholder to repeat the last step, as planarity isachieved individually rather than collectively. Thedrawback to this method is that slight rocking ofthe specimen may occur, especially if the speci-men height is too great, which degrades edgeretention and flatness.

Polishing ClothsThe requirements of a good polishing clothinclude the ability to hold the abrasive media,long life, absence of any foreign material thatmay cause scratches, and absence of any pro-cessing chemical (such as dye or sizing) thatmay react with the specimen. Many cloths ofdifferent fabrics, weaves, or naps are availablefor metallographic polishing. Napless or low napcloths are recommended for coarse polishingwith diamond abrasive compounds. Napless,low, medium, and occasionally high nap clothsare used for final polishing. This step should bebrief to minimize relief. Table 6 lists Buehler’sline of polishing cloths, their characteristics andapplications.

Polishing AbrasivesPolishing usually involves the use of one or moreof the following abrasives: diamond, aluminum

and polishing steps. These devices enable the op-erator to prepare a large number of specimens perday with a higher degree of quality than hand pol-ishing and at reduced consumable costs.Automatic polishing devices produce the best sur-

Polishing

Figure 25. Beta & Vector grinder/polisher andpower head

Figure 26. BuehlerVanguard 2000 fully automaticspecimen preparation system

2 6

Table 6. Cloth Selection Guide

RecommendedPremium μμμμμm Size Abrasive Cloth UsageCloths Range Types Characteristics Guide Applications

UltraPad 15 - 3 Diamond Hard Woven, Used to Replace Ferrouscloth No Nap with High Multiple SiC Materials and

Material Removal Grinding Steps ThermalSpray Coatings

UltraPol 15 - 1 Diamond Hard Woven, Excellent Surface Minerals, Coal,cloth No Nap Finish Used to Retain and Ceramics,

Flatness in Medium Inclusion Retentionto Hard Specimens in Steels, and

Refractory Metals

Nylon 15 - 1 Diamond Medium-Hard Used to Retain Ferrous Materials,cloth Woven, No Nap Flatness and Sintered Carbides

Hard Phases and Cast Irons

TexMet P 15 - 3 Diamond Hardest Perforated Used for Material Ceramics, Carbides,pad Non-woven Pad Removal and Petrographic, Hard

for High Material Flatness of Metals, Glass, MetalRemoval Hard Specimens Matrix Composites

TexMet 2000 15 - 3 Diamond, Hard Non-woven Used for Harder Hard Ferrouspad Al2O3 Pad used for Specimens and Materials

Durability Increased Flatness and Ceramics

TexMet 1000 9 - 0.02 Diamond, Soft Most Popular Cloth Ferrous andpad Al2O3, SiO2 Non-woven for Intermediate Nonferrous Metals,

Pad for Steps, Good for Ceramics, ElectronicSurface Finish Most Materials Packages, PCB’s,

Thermal SprayCoatings, Cast Irons,Cermets, Minerals,

Composites, Plastics

TriDent 9 - 1 Diamond Softer, Used to Maximize Ferrous andcloth Durable, Flatness and Nonferrous Metals,

Woven Synthetic, Retain Phases while Microelectronics,No Nap Providing Excellent Coatings

Surface Finish

WhiteFelt 6 - 0.02 Diamond, Soft and General Usage Ferrous andcloth Al2O3, SiO2 Durable Matted for Intermediate Nonferrous Materials

Wool Cloth to Fine Steps

NanoCloth 6 - 0.02 Diamond, Synthetic Cloth Softer Final Soft Ferrous andcloth Al2O3, SiO2 with Dense Polishing Nonferrous Metals

Short Nap Cloth also usedwith Diamondto MaximizeFlatness and

Surface Finish

MicroCloth 5 - 0.02 Diamond, Synthetic Cloth Most Popular Ferrous andcloth Al2O3, SiO2 with Medium Nap General Nonferrous

Usage Final Metals, Ceramics,Polishing Cloth Composites, PCB’s,

Cast Irons, Cermets, Plastics, Electronics

MasterTex 1 - 0.05 Al2O3, SiO2 Soft Synthetic Softer Final Soft Nonferrous andcloth Velvet with Polishing Cloth Microelectronic

Low Nap Packages

ChemoMet 1 - 0.02 Al2O3, SiO2 Soft Synthetic General Usage Titanium, Stainlesscloth Pad, Micro-nap Pad that Removes Steels, Lead/Tin

Smear Metal Solders, Electronicfrom Tough Packages, Soft

Materials during NonferrousChemomechanical Metals, Plastics

Polishing

Polishing

2 7

oxide (Al2O3), and amorphous silicon dioxide(SiO2) in collidal suspension. For certain materi-als, cerium oxide, chromium oxide, magnesiumoxide or iron oxide may be used, although theseare used infrequently. With the exception of dia-mond, these abrasives are normally suspendedin distilled-water, but if the metal to be polishedis not compatible with water, other suspensionssuch as ethylene glycol, alcohol, kerosene orglycerol, may be required. The diamond abra-sive should be extended only with the carrierrecommended by the manufacturer. Besides thewater-based diamond suspension, Buehler alsooffers oil-based suspensions that are needed toprepare materials sensitive to water.

Diamond abrasives were introduced to metallog-raphers commercially in the late 1940s. However,their use in metallography goes back to at leastthe late 1920s, as Hoyt [7] mentions a visit to theCarboloy plant in West Lynn, Massachusetts,where he saw sapphire bearings being polishedwith diamond dust in an oil carrier. He used someof this material to prepare sintered carbides andpublished this work in 1930. Diamond abrasiveswere first introduced in a carrier paste but lateraerosol and slurry forms were introduced. Virginnatural diamond was used initially, and is still avail-able in MetaDi diamond paste and suspensions.Later, synthetic diamond was introduced, first of

Figure 27. Comparison of monocrystalline (a, top)and polycrystalline (b, bottom) synthetic diamondgrain shapes (SEM, 450X).

the monocrystalline form, similar in morphology tonatural diamond, and then in polycrystalline form.MetaDi II diamond pastes use synthetic monoc-rystalline diamond while MetaDi Supremesuspensions use synthetic polycrystalline dia-monds. Figure 27 shows the shape differencesbetween monocrystalline and polycrystalline dia-monds. Studies have shown that cutting rates arehigher for many materials using polycrystalline dia-mond compared to monocrystalline diamond.

Colloidal silica was first used for polishing wafersof single crystal silicon where all of the damageon the wafer surface must be eliminated before adevice can be grown on it. The silica is amorphousand the solution has a basic pH of about 9.5. Thesilica particles are actually nearly spherical inshape, Figure 28, the polishing action is slow, andis due to both chemical and mechanical action.Damage-free surfaces can be produced moreeasily when using colloidal silica than with otherabrasives (for final polishing). Etchants can re-spond differently to surfaces polished with colloidalsilica. For example, an etchant that produces agrain contrast etch when polished with alumina

may instead reveal the grain and twin boundarieswith a “flat” etch when polished with colloidal silica.Color etchants frequently respond better when col-loidal silica is used producing a more pleasingrange of colors and a crisper image. But, cleaningof the specimen is more difficult. For manual work,use a tuft of cotton soaked in a detergent solution.For automated systems, stop adding abrasiveabout 10-15 seconds before the cycle ends and.For the last 10 seconds, flush the cloth surfacewith running water. Then, cleaning is simpler.Amorphous silica will crystallize if allowed toevaporate. Crystalline silica will scratch speci-mens, so this must be avoided. When opening a

b

a

Polishing

Figure 28. Amorphous silica particles in colloidalsilica (TEM, 300,000X).

2 8

bottle, clean off any crystallized particles than mayhave formed around the opening. The safest ap-proach is to filter the suspension before use.Additives are used to minimize crystallization, asin MasterMet 2 Colloidal Silica, greatly retardingcrystallization.

For routine examinations, a fine diamondabrasive, such as 1-μm, may be adequate asthe last preparation step. Traditionally, aqueousfine alumina powders and slurries, such as theMicroPolish II deagglomerated alumina powdersand suspensions, have been used for final polish-ing with medium nap cloths. Alpha alumina (0.3-μmsize) and gamma alumina (0.05-μm size) slurries(or suspensions) are popular for final polishing,either in sequence or singularly. MasterPrep alu-mina suspension utilizes alumina made by thesol-gel process, and it produces better surfacefinishes than alumina abrasives made by the tra-ditional calcination process. Calcined aluminaabrasives always exhibit some degree of agglom-eration, regardless of the efforts to deagglomeratethem, while sol-gel alumina is free of this problem.MasterMet colloidal silica suspensions (basic pHaround 9.5) are newer final polishing abrasivesthat produce a combination of mechanical andchemical action which is particularly beneficial fordifficult to prepare materials. Vibratory polishers,Figure 29, are often used for final polishing, par-ticularly with more difficult to prepare materials,for image analysis studies, or for publication qual-ity work.

Polishing

Figure 29. VibroMet 2 vibratory polisher

HELPFUL HINTS FOR

GRINDING/POLISHING

Specimens that containcracks or pores that are not filled with ep-oxy may require ultrasonic cleaning toremove abrasive and debris from the open-ings to avoid contaminating the next step.

Excessive ultrasonic cleaning vibrations candamage the structure of certain soft metalsand alloys, particularly precious metals.

To remove a worn out tightly adhering cloth,soak the platen under hot water for a fewminutes. This loosens the glue bond and al-lows you to pull it off easier.

2 9

EXAMPLES OF PREPARATIONPROCEDURES

The “Traditional” MethodOver the past forty years, a general procedurehas been developed that is quite successful forpreparing most metals and alloys. This methodis based on grinding with silicon carbide water-proof papers through a series of grits, then roughpolishing with one or more diamond abrasivesizes, followed by fine polishing with one or morealumina suspensions of different particle size.This procedure will be called the “traditional”method, and is described in Table 7.

This procedure is used for manual or automatedpreparation, although manual control of the forceapplied to a specimen would not be as consis-tent. Complementary motion means that thespecimen holder is rotated in the same direc-tion as the platen, and does not apply to manualpreparation. Some machines can be set so thatthe specimen holder rotates in the direction op-posite to that of the platen, called “contra.” Thisprovides a more aggressive action but was notpart of the “traditional” approach when auto-mated. The traditional method is not rigid, asother polishing cloths may be substitutedand one or more of the polishing steps might

be omitted. Times and pressures could bevaried, as well, to suit the needs of the work, orthe material being prepared. This is the “art”of metallography.

Contemporary MethodsNew concepts and new preparation materialshave been introduced that enable metallogra-phers to shorten the process while producingbetter, more consistent results. Much of this ef-fort has centered upon reducing or eliminatingthe use of silicon carbide paper in the grindingsteps. In all cases, an initial grinding step mustbe used, but there are a wide range of materialsthat can be chosen instead of SiC paper. Thereis nothing wrong with the use of SiC for the firststep, except that it has a short life. If an auto-mated device is used that holds a number ofspecimens rigidly (central force), then the firststep must remove the sectioning damage on eachspecimen and bring all of the specimens in theholder to a common plane. This first step is oftencalled “planar grinding.” SiC paper can be usedfor this step, although more than one sheet maybe needed. Alternatively, the metallographercould use PlanarMet Al alumina paper, conven-tional metal or resin bonded diamond discs or theUltraPrep diamond discs, the UltraPlan stainlesssteel mesh cloth (diamond is applied during use),

Table 7. The Traditional Procedure for Preparing Most Metals and Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 120- (P120) grit 6 (27) 240-300 Untilabrasive discs SiC water Comp. Plane(waterproof paper) cooled

CarbiMet 220- or 240- (P240 or P280) 6 (27) 240-300 1:00abrasive discs grit SiC water cooled Comp.

CarbiMet 320- (P400) grit 6 (27) 240-300 1:00abrasive discs SiC water cooled Comp.

CarbiMet 400- (P600) grit 6 (27) 240-300 1:00abrasive discs SiC water cooled Comp.

CarbiMet 600- (P1200) grit 6 (27) 240-300 1:00abrasive discs SiC water cooled Comp.

Canvas 6-μm MetaDi 6 (27) 120-150 2:00cloth diamond paste with Comp.

MetaDi fluid extender

Billiard or 1-μm diamond paste with 6 (27) 120-150 2:00Felt cloths MetaDi fluid extender Comp.

MicroCloth Aqueous 0.3-μm 6 (27) 120-150 2:00cloth α − alumina Comp.

MicroPolish slurry

MicroCloth Aqueous 0.05-μm 6 (27) 120-150 2:00cloth γ - alumina Comp.

MicroPolish slurry

Comp. = Complementary (platen and specimen holder both rotate in the same direction)

Examples of Preparation Procedures

3 0

the ApexHercules H and S rigid grinding discs (dia-mond is applied during use), or lapping platens ofseveral types (diamond is applied and becomesembedded in the surface during use). With thishuge array of products to choose from, how canthe metallographer decide what to use? Each ofthese products has advantages and disadvan-tages, and this is only the first step.

One or more steps using diamond abrasives onnapless surfaces usually follow planar grinding.PSA-backed silk, nylon or polyester cloths arewidely used. These give good cutting rates, main-tain flatness and minimize relief. Silk cloths, suchas the UltraPol cloth, provide the best flatness andexcellent surface finishes relative to the diamondsize used. UltraPad cloth, a thicker, hard, wovencloth, is more aggressive, gives nearly as good asurface finish, similar excellent flatness, and longerlife than an UltraPol cloth. Synthetic chemotextilepads, such as the TexMet1000 and 2000 cloths,give excellent flatness and are more aggressivethan silk. They are excellent for retaining secondphase particles and inclusions. Diamond suspen-sions are most popular with automated polishersas they can be added easily during polishing; al-though it is still best to charge the cloth initiallywith diamond paste of the same size to get polish-ing started quickly. Final polishing could beperformed with a very fine diamond size, such as0.1-μm diamond, depending upon the material, yourneeds and personal preferences. Otherwise, final

polishing is performed with MasterMet colloidalsilica or with MicroPolish or MasterPrep aluminaslurries using napless, or low to medium nap cloths.For some materials, such as titanium and zirco-nium alloys, an attack polishing solution is addedto the slurry to enhance deformation and scratchremoval and improve polarized light response.Contra rotation (head moves in the direction op-posite to the platen) is preferred as the slurry stayson the cloth better, although this will not work if thehead rotates at a high rpm. Examples of genericcontemporary preparation practices follow inTables 8 to 10. Specific procedures for a widerange of materials are given in the next section.

The starting SiC abrasive size is chosen basedupon the degree of surface roughness and depthof cutting damage and the hardness of the ma-terial. Never start with a coarser abrasive thannecessary to remove the cutting damage andachieve planar conditions in a reasonable time.

A similar scheme can be developed usingrigid grinding discs, such as the ApexHerculesHDisc. These discs are generally restricted to ma-terials above a certain hardness, such as 175 HV,although some softer materials can be preparedusing them. This disc can also be used for the pla-nar grinding step. An example of such a practice,applicable to nearly all steels (results are marginalfor solution annealed austenitic stainless steels)is given in Table 9.

Examples of Preparation Procedures

Table 8. Generic Four-Step Contemporary Procedure for Many Metals and Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 120- to 240- 6 (27) 240-300 Untilabrasive discs (P120 to P280) Comp. Plane***(waterproof paper) grit SiC water cooled

UltraPol 9-μm MetaDi 6 (27) 120-150 5:00cloth Supreme diamond Comp.

suspension*

TexMet 1000 or 3-μm MetaDi 6 (27) 120-150 4:00TriDent cloth Supreme diamond Comp.

suspension*

MicroCloth, ~0.05-μm 6 (27) 120-150 2:00NanoCloth or MasterMet Contra**ChemoMet colloidal silica orcloths MasterPrep

alumina suspensions* Plus MetaDi fluid extender as desired** With heads that rotate at >100 rpm, contra may spray the abrasive above the splash guard

*** Until Plane = All specimens in the holder are ground to a common plane and the damage from sectioning has been removedComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

3 1

The planar grinding step could also beperformed using a 45-μm metal bonded, or a30-μm resin bonded, UltraPrep disc or with theApexHercules H rigid grinding disc and 15- or30-μm diamond, depending upon the material. Rigidgrinding discs contain no abrasive; they must becharged during use and suspensions are the easi-est way to do this. Polycrystalline diamondsuspensions are favored over monocrystalline syn-thetic diamond suspensions for most metals andalloys due to their higher cutting rate.

The ApexHercules S rigid grinding disc, designedfor soft metals and alloys, is used in a similar man-ner. This disc is quite versatile and can be used toprepare harder materials as well, although its wear

rate will be greater than the H disc when used toprepare very hard materials. A generic four steppractice is given below for soft metals and alloys.

The planar grinding step can be performed withthe 30-μm resin bonded diamond disc, or witha second ApexHercules S disc and 15- or30-μm diamond, depending upon the metal oralloy. For some very difficult metals and alloys,a 1-μm diamond step on a TriDentcloth (similar tostep 3, but for 3 minutes) could be added, and/ora brief vibratory polish (use the same cloths andabrasives as for step 4) may be needed to pro-duce perfect publication quality images. Four stepsmay suffice for routine work.

Examples of Preparation Procedures

Table 10. Four-Step Contemporary Procedure for Nonferrous Metals Using a Rigid Grinding Disc

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/ Direction (min:sec)

CarbiMet 220- to 320- 5 (22) 240-300 Untilabrasive discs (P280 or P400) grit Comp. Planewaterproof paper SiC water cooled

ApexHercules S 6-μm MetaDi 5 (22) 120-150rigid grinding disc Supreme diamond Comp. 5:00

suspension*

TriDent 3-μm MetaDi 5 (22) 120-150cloth Supreme diamond Comp. 4:00

suspension*

MicroCloth, ~0.05-μm 5 (22) 120-150NanoCloth or MasterMet Contra 2:00ChemoMet colloidal silicacloths or MasterPrep

alumina suspensions

*MetaDi fluid extender as desiredComp.= Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Table 9. Four-Step Contemporary Procedure for Steels Using a Rigid Grinding Disc

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 120- to 240- 6 (27) 240-300 Untilabrasive discs (P120 to P280) Comp. Planewaterproof paper grit SiC, water cooled

ApexHercules H 9-μm MetaDi 6 (27) 120-150 5:00rigid grinding disc Supreme diamond Comp.

suspension*

TexMet 1000 or 3-μm MetaDi 6 (27) 120-150TriDent cloth Supreme diamond Comp. 4:00

suspension*

MicroCloth, ~0.05-μm 6 (27) 120-150 2:00NanoCloth or MasterMet ContraChemoMet colloidal silicacloths or MasterPrep

alumina suspensions

*Plus MetaDi fluid extender as desiredComp.= Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

3 2

PROCEDURES FOR SPECIFICMATERIALS

Following are our recommendations for preparinga wide range of materials arranged, in the case ofmetals, according to common characteristics,basically in accordance with their classification inthe periodic table of the elements. This break downis further modified, especially in the case of ironbased alloys, as required due to the wide range ofproperties that can be experienced.

The periodic table of the elements categorizesmetallic, nonmetallic and metalloid elementsaccording to similarities in atomic structure and isa good starting point, aided by similarities in physi-cal properties and behavior, for grouping elementsand their alloys that have reasonably similar prepa-ration procedures. The periodic table (pg. 29) hasthe various groups color coded for convenience.

Of course, metallographers prepare materialsother than metals. These have been grouped andsub-divided according to the nature of these ma-terials, as follows:

• Sintered Carbides

• Ceramics

• Composites: Metal Matrix, Polymer Matrix andCeramic Matrix

• Printed Circuit Boards

• Microelectronic Devices

• Plastics and Polymers

The procedures discussed in this book were de-veloped and tested extensively using machineswith an 8-inch (200 mm) diameter platen with six1.25-inch (30 mm) diameter specimens in a 5-inch(125 mm) diameter holder. These procedures pro-duced the same results when the specimens wereprepared using a 10-inch (250 mm) or a 12-inch(300 mm) diameter platen. When using the 8-inch(200 mm) platen and a 5-inch (125 mm) diameterspecimen holder, the machine’s head position wasadjusted so that the specimens rotate out over theouter periphery of the surface of the platen. Thisprocedure makes maximum use of the workingsurface and improves edge retention. This align-ment practice is less critical with the largerdiameter platens when using the same diameterholder. However, if the 7-inch (175 mm) diameterspecimen holder is used with the 12-inch (300 mm)diameter platen, the head position should be ad-

justed so that the specimens rotate out over theplaten’s edge.

In general, it is difficult to make accurate predic-tions about variations in time and load when usingdifferent format platens and mount sizes. Differ-ent mount sizes per se should not be used as thebasis for calculations as the area of the speci-mens within the mount is more important than themount area itself. For example, the same speci-men cross sectional area could be present in 1-,1.25- or 1.5-inch (25, 30 or 40 mm) diameter poly-mer mounts. Further, the number of specimens inthe holder can vary substantially. It is necessaryto put at least three mounts in a central-force holderto obtain a proper weight balance. But, two speci-mens could be dummy mounts (mounts withoutspecimens). In general, as the area of the speci-men in the mount increases, as the diameter andthe number of specimens increases, the time andpressure must increase to obtain the same de-gree of preparation perfection.

If the area of the mounts is used, rather than thearea of the embedded specimens, and we assumethat we are preparing six 1-inch (25 mm), six 1.25-inch (30 mm) or three 1.5-inch (40 mm) diameterspecimens, then we can adjust the pressures rela-tive to the standard values in the tables (for six1.25-inch (30 mm) diameter specimens) using an8 inch (200 mm) diameter platen format machine.A comparison of the cross-sectional areas wouldsuggest that for the six 1-inch (25 mm) specimens,the pressures could be reduced to 65% of the stan-dard values while for three 1.5-inch (40 mm)diameter specimens, the pressures and timescould be reduced to about 70% of the standardvalues.

Changing the platen diameter increases the dis-tance that the specimens travel in a given time.For example, for the same rpm values, specimenstravel 1.25 and 1.5 times as far using a 10- or 12-inch (250 or 300 mm) diameter platen,respectively, compared to an 8-inch (200 mm)platen. This suggests that the times could be re-duced to 80 or 67% of that used for the 8-inch(200 mm) diameter platen when using 10- or 12-inch (250 or 300 mm) diameter platens. Whilepreparation times may be shorter using a largerformat machine, consumable cost is greater andlarger machines are needed only when the speci-men throughput needs are very high or with largespecimens.

Procedures for Specific Materials

3 3

PER

IOD

IC T

AB

LE O

F E

LEM

EN

TS

Procedures for Specific Materials

3 4

4Be

12Mg

13Al

LIGHT METALS: Al, Mg and Be

LIGHT METALS: Al, Mg and Be

AluminumAluminum is a soft, ductile metal. Deformation in-duced damage is a common preparation problemin the purer compositions. After preparation, thesurface will form a tight protective oxide layer thatmakes etching difficult. Commercial grades con-tain many discrete intermetallic particles with avariety of compositions. These intermetallic par-ticles are usually attacked by etchants before thematrix. Although the response to specific etchantshas been used for many years to identify thesephases, this procedure requires careful control.Today, energy-dispersive analysis is commonlyperformed for phase identification due to its greaterreliability.

Light Metals

Five and four-step practices for aluminum alloysare presented below. While MgO was the preferredfinal polishing abrasive for aluminum and its al-loys, it is a difficult abrasive to use and is notavailable in very fine sizes. Colloidal silica has re-placed magnesia as the preferred abrasive for thefinal step and is finer in size. For color etchingwork, and for the most difficult grades of alumi-num, a brief vibratory polish may be needed tocompletely remove any trace of damage orscratches. The five-step practice is recommendedfor super pure (SP) and commercially pure (CP)aluminum and for wrought alloys that are difficultto prepare.

240- or 320-grit (P280 and P400) SiC waterproofpaper may also be used for the planar grindingstep. An UltraPol cloth produces better surface

Table 11. Five-Step Procedure for Aluminum Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 220 to 320abrasive discs (P240 to P400) 5 (22) 240-300 Until(waterproof paper) grit SiC water cooled Comp. Plane

UltraPol or 9-μm MetaDi 5 (22) 120-150 5:00UltraPad cloths Supreme diamond* Comp.

TriDent cloth 3-μm MetaDi 5 (22) 120-150 4:00Supreme suspension* Comp.

TriDent cloth 1-μm Metadi 5 (22) 120-150 2:00paste* Comp.

MicroCloth, ~0.05-μm 5 (22) 120-150 1:30NanoCloth or Mastermet ContraChemoMet cloth colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

3 5

Light Metals

finishes than an UltraPad cloth, but the UltraPadcloth has a longer useful life. A ChemoMet cloth isrecommended when edge retention is critical. Purealuminum and some alloys are susceptible to em-bedment of fine diamond abrasive particles,especially when suspensions are used. If this oc-curs, switch to diamond in paste form, which ismuch less likely to cause embedding. SP and CPaluminum can be given a brief vibratory polish(same products as last step) to improve scratchcontrol, although this is generally not required.MasterPrep alumina suspension has been foundto be highly effective as a final polishing abrasivefor aluminum alloys, however, the standard alu-mina abrasives made by the calcination processare unsuitable for aluminum.

Synthetic napless cloths may also be used for thefinal step with colloidal silica and they will intro-duce less relief than a low or medium nap cloth,but may not remove fine polishing scratches aswell. For very pure aluminum alloys, this proce-dure could be followed by vibratory polishing toimprove the surface finish, as these are quite dif-ficult to prepare totally free of fine polishingscratches.

Dendritic segregation in as cast 206 aluminumrevealed by color etching with Weck’s reagent(polarized light, 200X).

For many aluminum alloys, excellent results canbe obtained using a four step procedure, such asshown below in Table 12. This procedure retainsall of the intermetallic precipitates observed in alu-minum and its alloys and minimizes relief.

Deformed, elongated grain structure of extruded6061-F aluminum after shearing revealed byanodizing with Barker’s reagent (polarized light,100X).

Table 12. Four-Step Procedure for Aluminum Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/ Direction (min:sec)

CarbiMet 220- to 320- (P240 5 (22) 240-300 Untilabrasive discs to P400) grit SiC Comp. Plane(waterproof paper) water cooled

UltraPol 6-μm MetaDi 6 (27) 120-150 6:00cloth Supreme Comp.

diamondsuspension*

TriDent cloth 3-μm MetaDi 6 (27) 120-150 4:00Supreme diamond Comp.

suspension*

MicroCloth ~0.05-μm 6 (27) 120-150 2:00NanoCloth or MasterMet ContraChemoMet colloidal silicacloth or MasterPrep

alumina suspensions

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

HELPFUL HINTS FOR

ALUMINUM

Many low-alloy aluminumspecimens are difficult to polish to a perfectfinish using standard methods. A vibratorypolisher can be used quite effectively withMasterMet colloidal silica to produce defor-mation-free, scratch-free surfaces foranodizing or color etching and examinationwith polarized light or Nomarski DIC.

3 6

Light Metals

MagnesiumPreparation of magnesium and its alloys is ratherdifficult due to the low matrix hardness and thehigher hardness of precipitate phases that lead torelief problems, and from the reactivity of the metal.Mechanical twinning may result during cutting,grinding, or handling if pressures are excessive.Final polishing and cleaning operations shouldavoid or minimize the use of water and a varietyof solutions have been proposed. Pure magne-sium is attacked slowly by water while Mg alloysmay exhibit much higher attack rates. Some au-thors state that water should not be used in anystep and they use a 1 to 3 mixture of glycerol toethanol as the coolant even in the grinding steps.Always grind with a coolant, as fine Mg dust is afire hazard. Because of the presence of hard in-termetallic phases, relief may be difficult to control,especially if napped cloths are used. Following isa five-step procedure for magnesium and its al-loys, see Table 13.

MasterPolish is nearly water-free and yields ex-cellent results as the final abrasive. After the laststep, wash the specimens with ethanol. Cleaningafter the last step, without using water, is difficult.Holding the specimen under running water forabout a second eased the cleaning problem anddid not appear to harm the microstructure. Cos-metic cotton puffs can scratch the surface whenswab etching. For best results, etch-polish-etchcycle may be needed. Magnesium has a hexago-nal close-packed crystal structure and will respondto polarized light. To enhance the response, use abrief vibratory polish with the materials used in thelast step.

Mechanical twinning in deformed high-puritymagnesium (99.8% Mg) (acetic-picral etch, crossedpolarized light plus sensitive tint, 50X).

As-cast microstructure of a magnesium alloy with2.5% of rare earth elements, 2.11% Zn and 0.64% Zrshowing a film of the rare earth elements in thegrain boundaries, alloy segregation within grains(“coring” revealed by color variations within grains)and a few mechanical twins in the grains (acetic-picral etch, crossed polarized light plus sensitivetint, 100X).

Table 13. Four-Step Procedure for Magnesium Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 220- to 320- (P240 5 (22) 200-250 Untilabrasive discs to P400) grit SiC* Comp. Plane(waterproof paper) water cooled

TexMet 1000 9-μm MetaDi oil-based 5 (22) 120-150 6:00pad diamond slurry Contra

TexMet 1000 3-μm MetaDi oil-based 5 (22) 120-150 5:00pad diamond slurry Contra

TexMet 1000 1-μm MetaDi oil-based 5 (22) 120-150 4:00pad diamond slurry Contra

ChemoMet ~0.05-μm 3(13) 120-150 1:30-3:00cloth MasterPolish or Contra

MasterPrep alumina

*SiC surfaces were coated with wax to minimize embedmentComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

3 7

Light Metals

BerylliumBeryllium is also a difficult metal to prepare andpresents a health risk to the metallographer. Onlythose familiar with the toxicology of Be, and areproperly equipped to deal with these issues, shouldwork with the metal. The grinding dust is extremelytoxic. Wet cutting prevents air contamination butthe grit must be disposed of properly. As with Mg,Be is easily damaged in cutting and grinding form-ing mechanical twins. Light pressures are required.Although some authors claim that water cannotbe used, even when grinding Be, others report nodifficulties using water. Attack-polishing agents arefrequently used when preparing Be and many arerecommended [2]. Table 14 shows a four-step prac-tice for beryllium.

For the final step, mix hydrogen peroxide (30%concentration – avoid physical contact!) with theMasterMet colloidal silica in a ratio of one part hy-drogen peroxide to five parts colloidal silica. Oxalicacid solutions (5% concentration) have also beenused with alumina for attack polishing. For optimalpolarized light response, follow this with vibratorypolishing using a one-to-ten ratio of hydrogen per-oxide to colloidal silica.

Grain structure of wrought, P/M beryllium(unetched, cross-polarized light, 100X).

Table 14. Four-Step Procedure for Beryllium

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 220- to 320- (P240 4 (18) 240-300 Untilabrasive discs to P400) grit SiC Comp. Plane(waterproof paper) water cooled

UltraPol 6-μm MetaDi 4 (18) 120-150cloth Supreme Comp. 5:00

diamondsuspension*

TriDent 3-μm MetaDi 4 (18) 120-150 4:00cloth Supreme Comp.

diamondsuspension*

MicroCloth ~0.05-μm 3 (13) 80-120 2:00NanoCloth or MasterMet ContraChemoMet colloidal silicacloth

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

HELPFUL HINTS FOR

MAGNESIUM

For polishing, apply diamondpaste and use lapping oil, or use Buehler’soil-based MetaDi diamond suspensions,available down to a 0.10-μm size, with amedium nap cloth.

HELPFUL HINTS FOR

LIGHT METALS

Small diamond abrasive sizesare prone to embedment when applied as asuspension or by spraying. Charge the clothusing diamond paste to eliminate embedding.Another technique is to use both a fine dia-mond paste and MasterPrep aluminasuspension – simultaneously – to preventembedding.

3 8

Low Melting Point Metals

30Zn

48Cd

50Sn

51Sb

82Pb

83Bi

LOW-MELTING POINT METALS:Sb, Bi, Cd, Pb, Sn and Zn

LOW-MELTING POINT METALS:Sb, Bi, Cd, Pb, Sn and Zn

As pure metals, antimony, bismuth, cadmium, lead,tin, and zinc are all very soft and difficult to pre-pare. Pure antimony is quite brittle, but alloyscontaining Sb are more common. Bismuth is a softmetal, but brittle, and not difficult to prepare. How-ever, retaining bismuth par ticles in freemachining steels is difficult. Cadmium and zinc,both with hexagonal close-packed crystal struc-tures, are quite prone to mechanical twin formationif sectioning or grinding is performed too aggres-sively. Zinc is harder than tin or lead and tends tobe brittle. Zinc is widely used to coat sheet steel(galvanized steel) for corrosion protection, and isa common metallographic subject. Pure zinc isvery difficult to prepare. Lead is very soft and duc-tile and pure specimens are extremely difficult toprepare; however, lead alloys are considerablyeasier. Tin, which is allotropic with a body-cen-tered tetragonal crystal structure at roomtemperature, is soft and malleable and less sensi-tive to twinning. Pure tin, like pure lead, is very

difficult to prepare. Due to their low melting points,and low recrystallization temperatures, cold set-ting resins are usually recommended asrecrystallization may occur during hot compres-sion mounting. Some of these metals in the pure,or nearly pure form, will deform under the pres-sures used in compression mounting. Alloys ofthese metals are easier to prepare, as they areusually higher in hardness. Heating of surfacesduring grinding must be minimized. Grinding ofthese metals is always difficult, as SiC particlestend to embed heavily. Many au

Proeutectic dendrites of Cu2Sb surrounded by aeutectic mixture of antimony and Cu2Sb, in an as-cast Sb – 30% Cu specimen (unetched, polarizedlight, 200X).

(top) Cadmium dendrites and a Cd-Bi eutectic in anas-cast Cd – 20% Bi alloy (unetched, crossedpolarized light plus sensitive tint, 50X). (bottom)As-cast microstructure of Zn – 0.1% Ti – 0.2% Cualloy exhibiting mechanical twins and a three-phaseeutectic (alpha-Zn, Cu-Zn and Zn15Ti) in a cellularpattern (Palmerton reagent, crossed polarized lightplus sensitive tint, 200X).

3 9

Low Melting Point Metals

thors have recommended coating the SiC papersurface with bees wax, but this does not solvethe problem. Paraffin (candle wax) is much betterfor reducing embedding. Embedding is most com-mon with the finer grit size papers. Diamond is nota very effective abrasive with these metals. Alu-mina works quite well. Following is a procedurefor these alloys, see Table 15.

For best results, follow this with a vibratory polishusing MasterMet colloidal silica on MicroCloth orNanoCloth pads for times up to 1-2 hours. Thiswill improve polarized light response for the hcpCd and Zn and the rhombohedral Bi. The 1-μm alu-mina step will remove any embedded siliconcarbide particles much more effectively than a dia-mond abrasive step.

Table 15. Six-Step Procedure for Low-Melting Point Metals

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 320- (P400) grit SiC 4 (18) 150-250 Untilabrasive discs water cooled Comp. Plane(waterproof paper) (wax coated)*

CarbiMet 400- (P600) grit SiC 4 (18) 150-250 0:30abrasive water cooled Comp.discs (wax coated)*

CarbiMet 600- (P1200) grit SiC 4 (18) 150-250 0:30abrasive water cooled Comp.discs (wax coated)*

CarbiMet 800- (P1500) grit SiC 4 (18) 150-250 0:30abrasive water cooled Comp.discs (wax coated)*

MicroCloth or 1-μm 5 (22) 120-150 4:00-5:00NanoCloth MicroPolish II Comp.pads deagglomerated

alumina suspension

MicroCloth or MasterPrep 4 (18) 80-150 4:00NanoCloth 0.05-μm alumina Contrapads suspension**

*Rub candle wax lightly across revolving disc prior to grinding**See vibratory polish recommendation in textComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

HELPFUL HINTS FOR

LOW-MELTING POINT

METALS

Embedment of fine abrasive particles is acommon problem when preparing soft, low-melting point specimens. To reduceembedment of SiC abrasive particles, coatthe paper with wax before grinding. Johnson(Wear, Vol. 16, 1970, p. 351-358) showed thatcandle wax is far more effective than paraf-fin. However, if the layer applied is too thick,the metal-removal rate in grinding will be re-duced drastically.

4 0

Refractory Metals

50Sn

REFRACTORY METALS:Ti, Zr, Hf, Cr, Mo, Nb, Re, Ta, V, and W

22Ti

23V

24Cr

40Zr

41Nb

42Mo

72Hf

73Ta

74W

75Re

REFRACTORY METALS: Ti, Zr,Hf, Cr, Mo, Nb, Re, Ta, V and W

TitaniumPure titanium is soft and ductile, but is very easilydamaged by twinning in sectioning and grinding.Preparation of commercially pure titanium, whichis a popular grade, is very difficult, while prepara-tion of the alloys is somewhat easier. Some authorshave stated that titanium alloys should not bemounted in phenolic resins as the alloys can ab-sorb hydrogen from the resin. Further, it is possiblethat the heat from mounting could cause hydridesto go into solution. This is also possible withcastable resins if the exothermic reaction of poly-merization generates excessive heat. If the hydridephase content is a subject of interest, then thespecimens must be mounted in a castable resin

with a very low exotherm (long curing times favorlower heat generation, and vice versa). Titaniumis very difficult to section and has low grinding andpolishing rates. The following practice for titaniumand its alloys demonstrates the use of an attack-polishing agent added to the final polishing abrasive

Table 16. Three-Step Procedure for Titanium and Titanium Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 320- (P400) grit 6 (27) 240-300 Untilabrasive discs SiC, water cooled Comp. Plane(waterproof paper)

UltraPol 9-μm MetaDi 6 (27) 120-150 10:00cloth Supreme diamond Contra

suspension*

MicroCloth or ~0.05-μm 6 (27) 120-150 10:00NanoCloth MasterMet Contrapads colloidal silica plus

attack polish agent**

*Plus Metadi fluid extender as desired **See text for agentComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Alpha at the surface of heat treated (1038 ºC, waterquench) Ti – 3% Cr alloy after tint etching withBeraha’s reagent (polarized light, 500X).

4 1

Refractory Metals

Table 17. Four-Step Procedure for Zirconium and Hafnium

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 320- (P400) 5 (22) 200-250 Untilabrasive discs grit SiC Comp. Plane(waterproof paper) water cooled

UltraPol or 9-μm MetaDi Supreme 5 (22) 150-200 5:00UltraPad cloths diamond suspension* Contra

TriDent cloth 3-μm MetaDi Supreme 5 (22) 150-200 3:00diamond suspension* Contra

MicroCloth, ~0.05-μm 6 (27) 120-150 7:00NanoCloth or MasterMet ContraChemoMet colloidal silica pluscloths attack polish agent**

*Plus MetaDi fluid extender as desired **See text for agentComp. = Complementary (platen and specimen holder both rotate in the same directionContra = Platen and specimen holder rotate in opposite directions

Basket-weave alpha-beta structure of as-cast Ti –6% Al – 4% V revealed by heat tinting (polarizedlight, 100X).

to obtain the best results, especially for commer-cially pure titanium, a rather difficult metal toprepare free of deformation for color etching, heattinting and/or polarized light examination of thegrain structure. Attack polishing solutions addedto the abrasive slurry or suspension must betreated with great care to avoid burns. Use good,safe laboratory practices and it is advisable to wearprotective gloves. This three step practice couldbe modified to four steps by adding a 3- or 1-μmdiamond step, but this is usually unnecessary, seeTable 16.

A number of attack polishing agents have beenused. The simplest is a mixture of 10 mL hydro-gen peroxide (30% concentration – avoid skincontact) and 50 mL colloidal silica. Some metal-lographers add either a small amount of Kroll’sreagent to this mixture, or a few mL of nitric andhydrofluoric acids (avoid contact). These latter ad-ditions may cause the suspension to gel. Ingeneral, these acid additions do little to improvethe action of the hydrogen peroxide (the safer 3%

concentration is not effective). Polarized light re-sponse of CP titanium can be improved byfollowing this procedure with a brief vibratory pol-ish using colloidal silica.

Zirconium and HafniumPure zirconium and pure hafnium are soft, ductilehexagonal close-packed metals that can deformby mechanical twinning if handled aggressively insectioning and grinding. As with most refractorymetals, grinding and polishing removal rates arelow and eliminating all polishing scratches anddeformation can be difficult. It may even be pos-sible to form mechanical twins in compressionmounting. Both can contain very hard phases thatmake relief control more difficult. To improve po-larized light response, it is common practice tochemically polish specimens after mechanicalpolishing. Alternatively, attack polishing additionscan be made to the final polishing abrasive slurry,or vibratory polishing may be employed. Table 17is a four-step procedure that can be followed byeither chemical polishing or vibratory polishing.

Several attack polishing agents have been usedfor Zr and Hf. One is a mixture of 1-2 parts hydro-gen peroxide (30% concentration – avoid all skincontact) to 8 or 9 par ts colloidal sil ica.Another is 5 mL of a chromium trioxide solution(20 g CrO3 to 100 mL water) added to 95 mL col-loidal silica or MasterpPrep alumina slurry.Additions of oxalic, hydrofluoric or nitric acids have

4 2

Refractory Metals

also been used. All of these attack polishing addi-tions must be handled with care as they are strongoxidizers. Skin contact must be avoided. Chemi-cal polishing solutions are reviewed in [2]. Cain’shas been popular. Use under a hood and avoidskin contact. Ann Kelly developed an excellentchemical polish for refractory metals, such as Zr,Hf, and Ta. It consists of 25 mL lactic acid, 15 mLnitric acid and 5 mL hydrofluoric acid. Swab vigor-ously for up to 2 minutes. To prepare ultra-pure Zrand Hf, the above method is unsatisfactory and aprocedure, such as in Table 15 (add a 5- or 3-μmalumina step) is needed, as diamond is ineffec-tive. Use the chromium trioxide attack polish withthe alumina and conclude with Kelly’s chemicalpolish.

(top) Equiaxed grain structure of wrought zirconium(unetched, crossed polarized light, 100X). (middle)Equiaxed grain structure of wrought hafnium(unetched, crossed polarized light plus sensitivetint, 100X). (bottom) Fine grain recrystallizedmicrostructure of wrought Zr - 1.14% Cr strip in theas-polished condition at 200X using polarized lightplus sensitive tint.

HELPFUL HINTS FOR

REFRACTORY METALS

Due to their very low rate ofgrinding and polishing, the more aggressivecontra rotation gives suitable surfaces inless time.

Refractory metal preparation is aided by us-ing attack-polishing additives. Polarized lightresponse may be improved by followingpreparation by swabbing with a chemical pol-ishing solution.

(top) Cold worked microstructure of wroughtzirconium alloy XL (Zr - 3% Sn - 0.8% Nb - 0.8% Mo)in the as-polished condition at 200X with polarizedlight plus sensitive tint. (bottom) Microstructure (ααααα-Zr + AlZr3) of wrought Zr - 8.6% Al in the as-polished condition at 200X with polarized light.

4 3

Other Refractory Metals

OTHER REFRACTORY METALS:Cr, Mo, Nb, Re, Ta, V and W

These refractory metals have body-centeredcubic crystal structures (except for rheniumwhich is hexagonal close packed) and are softand ductile when pure, but some are brittle in com-mercial form. They can be cold worked easily,although they do not work harden appreciably, soit may be difficult to get completely deformation-free microstructures.

Pure chromium is soft and brittle; but, when en-countered commercially, for example, as a platedlayer, it is hard and brittle. Chromium alloys arerelatively easy to prepare, although difficult to etch.Molybdenum may be tough or brittle dependingupon composition. It is susceptible to deformationdamage in sectioning and grinding. Pure niobium(columbium) is soft and ductile and difficult to pre-pare while its alloys are harder and simpler to

Microstructure of powder-made W – 5.6% Ni – 2.4%Fe revealing tungsten grains surrounded by anickel-iron matrix (unetched, Nomarski DIC, 200X).

Table 18. Four-Step Procedure for Refractory Metals

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 320- (P400) 6 (27) 150-250 Untilabrasive discs grit SiC Comp. Plane(waterproof paper) water cooled

UltraPol or 9-μm MetaDi 6 (27) 150-200 10:00UltraPad Supreme diamond Contracloths suspension*

TexMet 1000 3-μm MetaDi 6 (27) 150-200 8:00TriDent cloth Supreme diamond Contra

suspension*

MicroCloth, ~0.05-μm 6 (27) 120-150 5:00NanoCloth or MasterMet ContraChemoMet colloidal silica pluscloths attack polish agent**

*Plus Metadi fluid extender as desired **See text for agent

Comp. = Complementary (platen and specimen holder both rotate in the same direction)

Contra = Platen and specimen holder rotate in opposite directions

prepare. Grinding and polishing rates have beenreported to vary with crystallographic orientation.Rhenium is very sensitive to cold work and willform mechanical twins. Tantalum is softer than nio-bium and more difficult to prepare as it easily formsdamaged layers in sectioning and grinding. Tanta-lum may contain hard phases that promote reliefcontrol problems. Vanadium is a soft, ductile metalbut may be embrittled by hydrogen; otherwise itcan be prepared much like a stainless steel. Tung-sten is not too difficult to prepare, although grindingand polishing rates are low. Hard carbides and ox-ides may be present in these metals that introducerelief control problems.

Mechanical polishing often incorporates anattack-polishing agent in the final step or is fol-lowed by vibratory polishing or chemical polishing.Manual preparation of these metals and their al-loys tends to be very tedious due to their lowgrinding and polishing rates. Automated ap-proaches are highly recommended, especially ifattack polishing is performed. Following is a ge-neric four step practice suitable for thesemetals and their alloys, see Table 18.

Many attack-polishing additives [2] have been sug-gested for these metals and their alloys.A good general purpose attack polish consistsof a mixture of 5 mL chromium trioxide solution(20 g CrO3 in 100 mL water) to 95 mL MasterMetcolloidal silica. Avoid skin contact as this is astrong oxidizing solution. A number of chemical

4 4

Other Refractory Metals

polishing solutions have been suggested [2]. ForNb, V and Ta, use a solution consisting of 30 mLwater, 30 mL nitric acid, 30 mL hydrochloric acidand 15 mL hydrofluoric acid. Swab or immerse atroom temperature. An alternative chemical polishfor Nb, V and Ta consists of 120 mL water, 6 gferric chloride, 30 mL hydrochloric acid and 16 mLhydrofluoric acid. Eary and Johnson recommendimmersing the specimen in this solution 1 minutefor V, 2 minutes for Nb and 3 minutes for Ta. Vibra-tory polishing is also very helpful for these metalsand their alloys.

(top) Alpha grains in Mo – 47.5% Re. The smallspots are sigma phase (Murakami’s reagent, 200X).Deformed alpha grains in W – 25% Re containingsigma phase (black spots) revealed using (middle)bright field and (bottom) Nomarski DIC whichreveals the cold work much better (Murakami’sreagent, 500X).

(top) Non-recrystallized grain structure of wroughtpure molybdenum, longitudinal plane (500X,polarized light, etch of water, hydrogen peroxide(30% conc) and sulfuric acid in 7:2:1 ratio).(bottom) Equiaxed alpha grain structure in wroughtpure vanadium (200X, etch: glycerol-nitric acid-hydrofluoric acid, 1:1:1ratio).

(top) Wrought tungsten - 10 atomic % Ti containinga small amount of alpha-Ti, beta-Ti-W eutectic andgrains of beta-Ti,W of varying composition andcrystal orientation (500X, Kroll’s reagent/Murakami’s reagent at room temperature). (bottom)Fine grain boundary precipitates (not identified) inwrought, cold worked Fan Steel 85-03 alloy (Nb -28% Ta - 10.5% W - 0.9% Zr), longitudinal plane(500X, etchant: lactic acid-nitric acid-hydrofluoricacid, 30:10:5 ratio).

4 5

Ferrous Metals

FERROUS METALS: Fe

26Fe

FERROUS METALS

Iron-based alloys account for a large portion of allmetals production. The range of compositions andmicrostructures of iron-based alloys is far widerthan any other system. Pure iron is soft and duc-ti le. Development of scratch-free anddeformation-free grain structures is difficult. Sheetsteels present the same problem, which can becomplicated by protective coatings of zinc, alumi-num or Zn-Al mixtures. In general, harder alloysare much easier to prepare. Cast irons may con-tain graphite, which must be retained in preparation.Inclusions are frequently evaluated and quantified.Volume fractions can vary from nearly 2% in a freemachining grade to levels barely detectable in apremium, double vacuum melt alloy. A wide rangeof inclusion, carbide and nitride phases has been

identified in steels. Addition of 12 or more percentchromium dramatically reduces the corrosion rateof steels, producing a wide range of stainless steelalloys. Tool steels cover a wide range of compo

Table 19. Four-Step Procedure for Hard Steels using the BuehlerHercules H disc

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 45-μm 6 (27) 240-300 Untilmetal-bonded diamond, Comp. Planedisc water cooled

ApexHercules H 9-μm MetaDi Supreme 6 (27) 120-150 5:00rigid grinding disc diamond suspension* Comp.

TriDent cloth 3-μm MetaDi Supreme 6 (27) 120-150 3:00diamond suspension* Comp.

MicroCloth, MasterPrep 6 (27) 120-150 2:00NanoCloth or 0.05-μm alumina ContraChemoMet suspension,cloths or MasterMet

colloidal silica

*Plus Metadi fluid extender desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Lath martensitic microstructure of anunconsolidated powder metallurgy steel gearcontaining substantial porosity (Klemm’s I reagent,polarized light, 200X).

4 6

Ferrous Metals

sitions and can attain very high hardnesses.Preparation of ferrous metals and alloys is quitestraightforward using the contemporary methods.Edge retention (see guidelines on pages 13-15)and inclusion retention are excellent, especially ifautomated equipment is used. The following pro-cedures are recommended and are adequate forthe vast majority of ferrous metals and alloys. Table9 presented a variation of the method given in Table19.

This practice is also useful for cast iron speci-mens including graphitic cast irons. 240- (P280)grit SiC paper can be substituted for the UltraPrepdisc or another ApexHercules H disc can be usedfor planar grinding with 30-μm MetaDi Supremediamond suspension. Due to their high silicon con-

tent and the potential for staining problems, it isbest to use the MasterPrep alumina suspensionfor the final polishing step.

For softer steels, use the 30-μm resin bondedUltraPrep disc, or 240- (P280) grit SiC paper, forstep 1. The ApexHercules S disc can be used inplace of the ApexHercules H disc but this is notusually necessary. However, with either disc, use6-μm MetaDi Supreme for the second step insteadof 9-μm, as shown in Table 20.

This practice is well suited for solution annealedaustenitic stainless steels and for soft sheet steels.UltraPol or UltraPad cloths could be substitute forthe rigid grinding discs, if desired. For perfect pub-lication quality images, or for color etch

Table 20. Four-Step Procedure for Soft Steels using the BuehlerHercules S disc

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 30-μm 6 (27) 240-300 Untilresin-bonded diamond, Comp. Planedisc water cooled

ApexHercules S 6-μm MetaDi 6 (27) 120-150 5:00rigid grinding disc Supreme diamond Comp.

suspension*

TriDent cloth 3-μm MetaDi 6 (27) 120-150 3:00Supreme diamond Comp.

suspension*

MicroCloth, MasterPrep 6 (27) 120-150 2:00NanoCloth or 0.05-μm alumina ContraChemoMet suspension,cloths or MasterMet

colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Table 21. Three-Step Procedure for Steels

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 120- to 320- 6 (27) 240-300 UntilAbrasive Discs (P120 to P400) Comp. Plane(Waterproof Paper) grit SiC, water cooled

ApexHercules H or S 3-μm MetaDi 6 (27) 120-150 5:00Rigid Grinding Disc Supreme Comp.

Diamond Suspension*

MicroCloth MasterPrep 0.05-μm 6 (27) 120-150 5:00NanoCloth or Alumina Suspension, ContraChemoMet or MasterMetCloths Colloidal Silica

*Plus MetaDi Fluid extender as desiredComp. = Complementary (platen and specimen holder rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

4 7

Ferrous Metals

ing, follow this practice with a brief vibratory pol-ish using the cloths and abrasives in the last step.

Many steels, particularly the harder steels, canbe prepared in three steps with excellent results.A recommended practice is given in Table 21.

For soft alloys, use 240- or 320- (P280 or P400)grit SiC paper; for harder alloys use 120 (P120),180- (P180), or 240- (P280) grit SiC paper,depending upon the starting surface finish and thehardness of the alloy. Planar grinding can also beperformed using 45-μm metal-bonded or 30-μmresin-bonded UltraPrep diamond discs. For softersteels, use the ApexHercules S disc for best re-sults. The UltraPol cloth can also be used for thesecond step for steels of any hardness.

The following practice is recommended for stain-less steels and maraging steels. For solutionannealed austenitic grades and for ferriticstainless grades and for annealed maraginggrades, use the ApexHercules S disc or theUltraPol cloth for best results. Start with 120- (P120)grit SiC paper only if it is a very hard martensiticstainless steel, such as type 440C. For the mar-tensitic grades, planar grinding can be performedusing a 45-μm metal-bonded diamond UltraPrep

disc. For softer stainless steels, use the 30-μmresin-bonded diamond UltraPrep disc for planargrinding. Another alternative is a second ApexHercules disc, either H or S, depending upon thehardness of the grade, and a 30-μm MetaDiSupreme polycrystalline diamond suspension. Thesolution annealed austenitic stainless steels andthe fully ferritic stainless steels are the most diffi-cult to prepare. It may be helpful to add a 1-μmdiamond step on a TriDent cloth before the laststep, or to follow the last step with a brief vibratorypolish using colloidal silica on MicroCloth,NanoCloth or a ChemoMet cloth, see Table 22.

Table 22. Four-Step Contemporary Procedure for Stainless Steels and Maraging Steels

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 120- to 240- (P120 6 (27) 240-300 Untilabrasive discs to P280) grit SiC, Comp. Plane(waterproof paper) water cooled

ApexHercules S 9-μm 6 (27) 120-150 5:00rigid grinding disc or MetaDi Comp.UltraPol cloth Supreme

DiamondSuspension*

TriDent cloth 3-μm 6 (27) 120-150 5:00MetaDi Comp.

Supremediamond

suspension*

MicroCloth, MasterPrep 6 (27) 120-150 2:30NanoCloth or 0.05-μm alumina ContraChemoMet suspension,cloths or MasterMet

colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Microstructure of a duplex (ferrite tan, austenitewhite) stainless steel held at 816º C for 48 hourswhich forms sigma phase (orange particles)revealed by electrolytic etching with aqueous 20%NaOH at 3 V dc, 9 s (500X).

4 8

Copper, Nickel and Cobalt

27Co

28Ni

29Cu

Copper, Nickel and Cobalt

COPPER, NICKEL and COBALT

CopperPure copper is extremely ductile and malleable.Copper and its alloys come in a wide rangeof compositions, including several variants ofnearly pure copper for electrical applications tohighly alloyed brasses and bronzes and toprecipitation hardened high strength alloys.Copper and its alloys can be easily damaged byrough sectioning and grinding practices andthe depth of damage can be substantial. Scratchremoval, par ticularly for pure copper andbrass alloys, can be very difficult. Following thepreparation cycle with a brief vibratory polish us-ing colloidal silica is very helpful for scratchremoval. Attack-polish additions have been usedin the past to improve scratch removal butusually are not necessary using the contempo-

rary method followed by vibratory polishing, seeTable 23.

Planar grinding can be performed using the45- or 15-μm metal-bonded or the 30-μm resin-bonded UltraPrep discs. Use the resin-bondeddisc for the soft copper grades and copperalloys.

Table 23. Five-Step Procedure for Cu and Cu Alloys

Alpha grains containing annealing twins inphosphorous-deoxidized arsenical bronze that wasannealed and lightly cold drawn (Klemm’s I reagent,polarized light, 50X).

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet abrasive discs 220- to 320- (P240 to 5-6 240-300 Until(waterproof paper) P400) grit SiC water cooled (22-27) Comp. Plane

UltraPol cloth or 6-μm MetaDi 5-6 120-150 5:00ApexHercules S Supreme diamond (22-27) Comp.rigid grinding disc suspension*

TriDent or 3-μm MetaDi Supreme 5-6 120-150 3:00TexMet 1000 pads diamond suspension* (22-27) Comp.

TriDent 1-μm MetaDi Supreme 5-6 120-150 2:00cloth Diamond Suspension* (22-27) Comp.

MicroCloth, ~0.05-μm MasterMet 5-6 120-150 1:30-2:00NanoCloth or colloidal silica (22-27) ContraChemoMet or MasterPreppads alumina suspensions

*Plus MetaDi Fluid extender as desiredComp.= Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

4 9

Copper, Nickel and Cobalt

NickelNickel and its alloys have face-centered cubiccrystal structures and are prepared in basicallythe same way as austenitic stainless steels. Purenickel is more difficult to prepare than the alloys.The Ni-Fe magnetic alloys are rather difficult toprepare scratch free unless vibratory polishing isused. The Monel (Ni-Cu) and the highly corrosionresistant (Ni-Cr-Fe) alloys are moredifficult to prepare than the nickel-based superal-loys. Solution annealed superalloys are alwaysmore difficult to prepare than age hardened su-peralloys. Age hardened superalloys can beprepared using the ApexHercules H disc; for allother nickel alloys, use the ApexHercules S discfor best results. The following practice works wellfor nickel based superalloys (and Fe-Ni basedsuper alloys) and the highly corrosion resistantNi-Cr-Fe alloys, see Table 24.

If color etching is to be performed, follow the laststep with a brief vibratory polish using the samematerials as in the last step. This step is also help-ful for the most difficult to prepare solution annealedalloys. Alternatively, for the most difficult speci-mens, or when color etching is being performed,a 1-μm diamond step on a Trident cloth can beadded before the final step.

For pure nickel, nickel-copper and nickel-ironalloys, a five step practice is preferred, as givenbelow. The planar grinding step can be performedusing either the 30-μm resin-bonded UltraPrep dia-mond disc or with 240- (P280) or 320- (P400) gritSiC papers with equal success, see Table 25.

Attack-polishing agents are not often used withthese alloys to eliminate fine polishing scratchesor residual damage. If this is a problem, and someof these grades are very difficult to get perfectly

Table 25. Five-Step Procedure for Ni, Ni-Cu and Ni-Fe Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 30-μm diamond 5 (22) 200-300 Untilresin-bonded disc water cooled Comp. Plane

UltraPol cloth or 9-μm MetaDi 6 (27) 100-150 5:00ApexHercules S Supreme diamond Comp.rigid grinding disc suspension*

TriDent or 3-μm MetaDi Supreme 6 (27) 100-150 3:00TexMet 1000 pads diamond suspension* Comp.

TriDent 1-μm MetaDi Supreme 6 (27) 100-150 2:00cloth diamond suspension* Comp.

MicroCloth, ∼0.05-μm MasterMet 6 (27) 80-150 1:30-2:00NanoCloth or colloidal silica or ContraChemoMet pads MasterPrep

alumina suspensions

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Table 24. Four-Step Procedure for Nickel-Based Superalloys and Ni-Cr-Fe Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet abrasive discs 220- to 240- (P240 to P280) 6 (27) 240-300 Until(waterproof paper) grit SiC water cooled Comp. Plane

ApexHercules H 9-μm MetaDi 6 (27) 120-150 5:00or ApexHercules S Supreme diamond Comp.rigid grinding disc suspension*

Trident cloth 3-μm MetaDi Supreme 6 (27) 120-150 5:00diamond suspension* Comp.

MicroCloth, MasterPrep 0.05-μm 6 (27) 120-150 2:00-5:00NanoCloth alumina suspension or ContraChemoMet pads MasterMet colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

5 0

Copper, Nickel and Cobalt

free of scratches and deformation damage, a briefvibratory polish, using the same materials as inthe last step, will provide the needed improvement.MasterPrep alumina may give better results thancolloidal silica for the more pure nickel composi-tions.

CobaltCobalt and its alloys are more difficult toprepare than nickel and its alloys. Cobalt is a toughmetal with a hexagonal close-packedcrystal structure and is sensitive to deformationdamage by mechanical twinning. Grinding and pol-ishing rates are lower for Co than for Ni, Cu or Fe.Preparation of cobalt and its alloys issomewhat similar to that of refractory metals. De-spite its hcp crystal structure, crossedpolarized light is not very useful for examiningcobalt alloys compared to other hcp metals andalloys. Following is a practice for preparing Co andits alloys, see Table 26.

Two steps of SiC paper may be needed to get thespecimens co-planar. If the cut surface is of goodquality, start with 320- (P400) grit paper. Cobaltand its alloys are more difficult to cut than moststeels, regardless of their hardness. Attack pol-ishing has not been reported but chemicalpolishing has been used after mechanical polish-ing. Morral (2) has recommended two chemicalpolishing solutions: equal parts of acetic and nitricacids (immerse) or 40 mL lactic acid, 30 mL hy-drochloric acid and 5 mL nitric acid (immerse). Awide variety of Co-based alloys have been pre-pared with the above method without need forchemical polishing. The 1-μm diamond step couldbe eliminated for routine work.

Alpha grains containing annealing twins of solutionannealed and double aged Waspaloy nickel-basedsuperalloy (Beraha’s reagent, 100X).

Table 26. Five-Step Procedure for Cobalt

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 220- to 320- 6 (27) 250-300 Untilabrasive discs (P240 to P400) grit Contra Plane(Waterproof Paper) SiC water cooled

UltraPol cloth or 9-μm MetaDi 6 (27) 100-150 5:00ApexHercules S Supreme diamond Contrarigid grinding disc suspension*

TexMet 1000 3-μm MetaDi Supreme 6 (27) 100-150 5:00pad diamond suspension* Contra

TriDent or 1-μm MetaDi Supreme 6 (27) 100-150 3:00TexMet 1000 pads diamond suspension* Contra

MicroCloth, ~0.05-μm MasterMet 6 (27) 80-120 2:00-3:00NanoCloth or colloidal silica or ContraChemoMet MasterPreppads alumina suspensions

*Plus Metadi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Equiaxed grain structure of Elgiloy (Co – 20% Cr –15% Fe – 15% Ni – 7% Mo – 2% Mn – 0.15% C –0.05% Be) after hot rolling and annealing revealingannealing twins (Beraha’s reagent, crossedpolarized light plus sensitive tint, 100X).

5 1

Precious Metals

PRECIOUS METALS: Au, Ag, Ir,Os, Pd, Pt, Rh, and Ru

44Ru

45Rh

46Pd

47Ag

77Ir

78Pt

79Au

76Os

PRECIOUS METALS: Au, Ag, Ir,Os, Pd, Pt, Rh and Ru

Relatively few metallographers work with preciousmetals, other than those used in electronic devices.Preparing precious metals within an integrated cir-cuit is discussed later (see MicroelectronicDevices). The precious metals are very soft andductile, deform and smear easily, and are quitechallenging to prepare. Pure gold is very soft andthe most malleable metal known. Alloys, which aremore commonly encountered, are harder andsomewhat easier to prepare. Gold is difficult to etch.Silver is very soft and ductile and prone to sur-face damage from deformation. Embedding ofabrasives is a common problem with both gold andsilver and their alloys. Iridium is much harder andmore easily prepared. Osmium is rarely encoun-tered in its pure form, even its alloys are infrequent

subjects for metallographers. Damaged surfacelayers are easily produced and grinding and pol-ishing rates are low. It is quite difficult to prepare.Palladium is malleable and not as difficult to pre-pare as most of the precious metals. Platinum issoft and malleable. Its alloys are more commonlyencountered. Abrasive embedment is a problem

Eutectic microstructure of Ag – 28% Cu whereKlemm’s reagent has colored the copper particlesand the silver phase is uncolored (500X).

Table 27. Stewart’s Manual Procedure for Precious Metals

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet abrasive 220- to 240- (P240 to P280) moderate 400 Untildiscs (waterproof paper) grit SiC water cooled Plane

CarbiMet 320- (P400) grit SiC moderate 400 1:00abrasive discs water cooled

CarbiMet 400- (P600) grit SiC moderate 400 1:00abrasive discs water cooled

CarbiMet 600- (P1200) grit SiC moderate 400 1:00abrasive discs water cooled

MetCloth 6-μm MetaDi Supreme moderate 400 2:00pad diamond suspension

MasterTex 1-μm MetaDi Supreme moderate 400 2:00cloth diamond suspension

Comp. = Complementary (platen and specimen holder both rotate in the same direction)

5 2

Precious Metals

with Pt and its alloys. Rhodium is a hard metaland is relatively easy to prepare. Rh is sensitiveto surface damage in sectioning and grinding. Ru-thenium is a hard, brittle metal that is not too difficultto prepare.

Stewart (Tech-Notes, Vol. 2, Issue 5) has de-scribed a method for preparing jewelry alloys.Invariably, these are small pieces, due to theircost, and must be mounted. Stewart uses EpoMetG resin for most specimens. If transparency isneed, he uses TransOptic resin. Fragile beads andballs are mounted in castable resins. Grinding andpolishing was conducted at 400 rpm. His practiceis shown in Table 27.

This can be followed by a brief vibratory polishusing colloidal silica on MasterTex, ChemoMet,MicroCloth or NanoCloth pads to further enhancethe quality of the preparation; but, a 1-μm diamondfinish is adequate for most work. Attack-polishinghas been used to prepare gold and its alloys andchemical polishing has been performed after me-chanical polishing of silver, but neither practice iscommonly performed. Alternate etch-polish cyclesmay be needed to remove fine polishing scratches,especially for annealed specimens.

A procedure for an automated system wasdeveloped after experimentation with a number ofprecious metal specimens. Most of thesemetals and alloys are quite soft, unless they havebeen cold worked, and they are susceptible to em-bedding of abrasives. In this method, only onesilicon carbide step is used. TexMet 1000 padsare used for the diamond steps, as it will hold theabrasive in its surface well, which minimizes em-

bedding. Only diamond paste is used, as slurrieswill be more prone to embedding, as we observedin high gold alloys. Use only a small amount ofdistilled water as the lubricant. Do not get the clothexcessively wet. Final polishing is with aChemoMet I cloth and MasterPrep alumina. Dueto their excellent corrosion resistance, colloidalsilica is not effective as an abrasive for preciousmetals. The ChemoMet pad has many fine poresto hold the abrasive. The cycle is given in Table28.

For 18 karat gold and higher (≥75% Au), it is nec-essary to use an attack polish agent in the finalstep. An aqueous solution of 5 g CrO3 in 100 mLwater works well. Mix 10 mL of the attack polishagent with 50 mL of MasterPrep suspension. Thiswill thicken, so add about 20-30 mL water to makeit thinner. A 3 to 6 minute attack- polish step willremove the fine polishing scratches. Wear protec-tive gloves as the chromium trioxide solution is astrong oxidizer.

Table 28. Five-Step Automated Procedure for Precious Metals

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 220- to 320- (P240 to 3 (13) 150-240 Untilabrasive discs P400) grit water cooled Comp. Plane

TexMet 1000 9-μm MetaDi II 3 (13) 150-240 5:00pad diamond paste* Comp.

TexMet 1000 3-μm MetaDi II 3 (13) 150-240 3:00pad diamond paste* Comp.

TexMet 1000 1-μm MetaDi II 3 (13) 150-240 2:00pad diamond paste* Comp.

ChemoMet MasterPrep 2 (9) 100-150 2:00pad alumina suspension Comp.

* Use water as the lubricant, but keep the pad relatively dryComp. = Complementary (platen and specimen holder both rotate in the same direction)

Equiaxed grain structure of cold rolled and annealed18-karat gold (Neyoro 28A: 75% Au – 22% Ag – 3%Ni) revealing annealing twins (equal parts 10%NaCN and 30% H2O2, 50X).

5 3

Thermally-Spray Coated Specimens

THERMALLY-SPRAY COATEDSPECIMENS

Thermally sprayed coatings (TSC) and thermalbarrier coatings (TBC) are widely used on manymetal substrates. Invariably, these coatings arenot 100% dense but contain several types of voids,such as porosity and linear detachments. Hot com-pression mounting is not recommended as themolding pressure can collapse the voids. Use alow-viscosity castable epoxy and use vacuum in-filtration to fill the connected voids with epoxy.Fluorescent dyes, such as FluorMet dye, may be

Table 30. Four-Step Procedure for TSC and TBC Specimens with Metallic Coatings

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

ApexHercules H 30-μm MetaDi 5 (22) 240-300 Untilrigid grinding disc Supreme diamond Contra Plane

suspension*

UltraPol 9-μm MetaDi 5 (22) 150-200 4:00cloth Supreme diamond Contra

suspension*

TriDent 3-μm MetaDi 6 (27) 120-150 3:00cloth Supreme diamond Comp.

suspension*

MicroCloth, MasterPrep 4 (18) 120-150 2:00NanoCloth or 0.05-μm alumina ContraChemoMet suspension orcloths MasterMet

colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Table 29. Four-Step Procedure for TSC and TBC Specimens with Metallic Coatings

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

ApexHercules H 30-μm MetaDi 5 (22) 240-300 Untilrigid grinding disc Supreme diamond Contra Plane

suspension*

ApexHercules H 6-μm MetaDi 5 (22) 150-200rigid grinding disc Supreme diamond Contra 4:00

suspension*

TriDent 3-μm MetaDi 6 (27) 120-150 3:00cloth Supreme diamond Comp.

suspension*

MicroCloth, MasterPrep 4 (18) 120-150 2:00NanoCloth or 0.05-μm alumina ContraChemoMet suspension orcloths MasterMet

colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

added to the epoxy. When viewedwith fluorescent illumination, the epoxy-filled voidsappear bright yellow-green. This makes it easy todiscriminate between dark holes and dark oxides,as would be seen with bright field illumination. Fill-ing the pores with epoxy also makes it easier tokeep the pore walls flat to the edge during prepa-ration. Aside from this mounting requirement, TSCand TBC specimens are prepared using all of thefactors needed for good edge retention (see pages13-15). A variety of procedures can be used. TheApexHercules H rigid grinding disc produces ex-ceptional edge flatness

5 4

for these specimens. Two four step proceduresfor TSC and TBC specimens using the Apex Her-cules H rigid grinding disc for specimens withmetallic coatings and one procedure for specimenswith ceramic coatings are given, see Tables 29-31.

The 30-μm resin bonded, or the 45-μm metalbonded UltraPrep diamond discs, can be substi-tuted for the planar grinding step. (Table 29 and30). UltraPol or UltraPad cloths can be used in thesecond step. (Table 29 and 31).

Table 31. Four-Step Procedure for TSC and TBC Specimens with Ceramic Coatings

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 45-μm disc 5 (22) 240-300 Untilmetal-bonded disc water cooled Contra Plane

ApexHercules H 9-μm 5 (22) 150-180 4:00rigid grinding disc MetaDi Contra

Supremediamond

suspension*

TexMet 1000 or 2000 3-μm 6 (27) 120-150 3:00or TriDent cloth MetaDi Comp.

Supremediamond

suspension*

MicroCloth, MasterPrep 8 (36) 120-150 2:30NanoCloth or 0.05-μm alumina ContraChemoMet suspension orcloths MasterMet

colloidal silica

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Thermally-Spray Coated Specimens

NiCrAlY thermally-spray coated steel specimenrevealing a small amount of porosity (black spots),linear detachments (elongated black lines), andinclusions (gray particles) (unetched, 100X).

Microstructure of a steel substrate covered by twothermally-sprayed layers, a NiAl bond coat andyittria-zirconia top coat (unetched, 100X). The bondcoat contains pores, linear detachments andinclusions while the top coat is quite porous.

5 5

SINTERED CARBIDES

Sintered carbides are very hard materials madeby the powder metallurgy process and may byreinforced with several types of MC-typecarbides besides the usual tungsten carbide (WC).The binder phase is normally cobaltalthough minor use is made of nickel. Modern cut-ting tools are frequently coated with avariety of very hard phases, such as alumina, ti-tanium carbide, titanium nitride and titaniumcarbonitride. Sectioning is normally performed witha precision saw, so surfaces are very good andrough abrasives are not usually required. Listedbelow are two alternate proceduresfor preparing sintered carbides using metal-bonded

UltraPrep discs or the ApexHerculesH rigid grind-ing discs. A further option would be the use of thetraditional metal-bonded diamond discs.

Table 32. Four-Step Procedure for Sintered Carbides using UltraPrep Discs

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 45-μm diamond 6 (27) 240-300 5:00**metal-bonded disc water cooled Contra

UltraPrep 9-μm diamond 6 (27) 240-300 4:00metal-bonded disc water cooled Contra

TriDent, UltraPol, 3-μm MetaDi 6 (27) 120-150 3:00TexMet 1000 Supreme diamond Contraor Nylon cloths suspension*

TriDent, ~0.05-μm MasterMet 6 (27) 120-150 2:00UltraPol colloidal silica or Contraor Nylon cloths MasterPrep alumina

suspensions

*Plus MetaDi fluid extender as desired

** Grind longer than “Until Plane”Comp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Table 33. Four-Step Procedure for Sintered Carbides using ApexHercules H Discs

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

ApexHercules H 30-μm MetaDi 6 (27) 240-300 5:00**rigid grinding disc Supreme diamond Contra

suspension*

ApexHercules H 9-μm MetaDi 6 (27) 240-300 4:00rigid grinding disc Supreme diamond Contra

suspension*

TriDent , UltraPol, 3-μm MetaDi 6 (27) 120-150 3:00TexMet 1000 Supreme diamond Comp.or Nylon cloths suspension*

MicroCloth, ~0.05-μm MasterMet 6 (27) 120-150 2:00NanoCloth or colloidal silica or ContraChemoMet cloths MasterPrep alumina

suspensions

*Plus MetaDi fluid extender as desired** Grind longer than “Until Plane”Comp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Surface of a WC – 11.5% Co cutting tool enrichedwith 1.9% Ta and 0.4% Nb (form complex carbides,the dark regions in the matrix) and coated withalumina (arrows) for enhanced tool life (Murakami’sreagent, 1000X).

Sintered Carbides

5 6

If a greater amount of material must be removedin the planar grinding step, use either the 30-μmresin bonded or the 45-μm metal-bonded UltraPrepdisc for step one, in Table 33.

The final step in these two procedures, Tables 32and 33, employs either MasterMet colloidal silicaor MasterPrep alumina suspensions as they willproduce the best definition of the structure, par-ticularly if the surfaces have a complex series ofcoatings for improved wear resistance, as usedin some coated carbide inserts. However, if sucha coating is not present, the matrix grain structurecan be observed quite clearly after the 3-μm dia-mond step. For routine examination there is littleneed to go beyond the third step. It is also pos-sible to use a 1-μm diamond step, similar to stepthree, but for 2 minutes, as an alternative fourthstep.

Sintered Carbides

Microstructure at the surface of a multilayered,CVD-coated WC - 8% Co cutting tool. The arrowspoint to the CVD layers of TiCN, TiN, TiC andalumina. The region below the coatings is madehigher in Co to improve crack resistance andcomplex carbide forming elements (Ta, Ti and Nb)are added to the matrix (dark spots at bottom) forwear resistance (Murakami’s reagent, 1000X).

HELPFUL HINTS FOR

SINTERED CARBIDES

The ability to see the bound-aries between the WC

particles and cobalt binder in the as-polishedcondition depends on the surface of the pol-ishing cloth used in the last step. To see noboundaries, which makes it easier to seegraphite or eta phase, use a nap-less sur-face, such as TriDent, TexMet 1000 orUltraPol pads. To see the phase boundariesbetween WC particles and the cobalt binder,use a medium nap surface, such asMicroCloth, MasterTex or NanoCloth pads.

Microstructure of cold isostatically pressed andage hardened Ferro-Titanit Nikro 128 (Fe - 13.5% Cr- 9% Co - 4% Ni - 30% TiC) cutting tool;a) etched with Murakami’s reagent at roomtemperature to darken the TiC; and b) electrolyticetched with aqueous 1% chromium trioxide at 2 Vdc, 30 s to reveal the martensitic matrix (1000X).

b

a

5 7

CERAMICS

Ceramic materials are extremely hard and brittleand may contain pores. Sectioning must be per-formed using diamond blades. If the specimen isto be thermally etched, then it must be mounted ina resin that permits easy demounting, and vacuuminfiltration of epoxy into the voids should not bedone. Deformation and smearing are not problemswith ceramics due to their inherent characteris-tics. But, it is possible to break out grains orproduce cracking during preparation.

Ceramics

Table 34. Four-Step Procedure Suitable for Most Ceramic Materials

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 45-μm diamond 8 (36) 120-150 Untilmetal-bonded disc MetaDi fluid Comp. Flat

for coolant

ApexHercules H 9-μm MetaDi 6 (27) 120-150 4:00rigid grinding disc Supreme diamond Comp.

suspension*

ApexHercules S 3-μm MetaDi 6 (27) 120-150 3:00rigid grinding disc Supreme diamond Comp.

suspension*

TexMet 1000, 1-μm MetaDi 10 (44) 120-150 5:00Nylon or Supreme diamond ContraUltraPol suspension* orcloths MasterPolish 2

suspension

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Table 35. Four-Step Procedure Suitable for Most Ceramic Materials

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 45-μm diamond 8 (36) 120-150 Untilmetal-bonded MetaDi fluid Contra Flatdisc for coolant

UltraPad 15-μm MetaDi 6 (27) 120-150 6:00cloth Supreme diamond Comp.

suspension*

UltraPad or 6-μm MetaDi 6 (27) 120-150 4:00TexMet 1000 or Supreme diamond Comp.TexMet 2000 cloths suspension*

TexMet 1000 or 3-μm MetaDi 8 (36) 120-150 4:00TexMet 2000, Supreme diamond ContraNylon or suspension andUltraPol MasterMetcloths colloidal silica

(50:50 simultaneously)

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Microstructure of hot-pressed silicon nitride (binderunknown) (unetched, 500X).

5 8

Ceramics

Pullouts are a major problem to control as theycan be misinterpreted as porosity. Mechanicalpreparation has been done successfully with laps,metal-bonded diamond discs, rigid grinding discs,or hard cloths. SiC paper is rather ineffective withmost ceramics, as they are nearly as hard or ashard as the SiC abrasive. Consequently, diamondis commonly used for nearly all preparation steps.Automated preparation is highly recommendedwhen preparing ceramic materials as very highforces arise between the specimen and the work-ing surface, often too high for manual preparation.Listed are two generic procedures for preparingceramics using different approaches, see Tables34 and 35.

MasterPolish 2 suspension is specifically formu-lated for final polishing ceramic materials and willyield a better surface finish than 1-μm diamond.

HELPFUL HINTS FOR

CERAMICS

Cut specimens as close aspossible to the desired exami-

nation plane using a saw and blade thatproduces the least possible amount of dam-age, producing the best possible surface forsubsequent preparation. Rough grinding isslow and tedious and does considerabledamage to the structure and should beavoided as much as possible.

Coarse abrasives for grinding may beneeded at times. However, severe grain pull-outs can result from use of coarse abrasivesand it is recommended to use the finest pos-sible coarse abrasive for grinding.

(top) Microstructure of an alumina (matrix), zirconia(gray), silica (white) refractory material (unetched,200X). (bottom) Silicon - SiC ceramic etchedelectrolytically with 10% oxalic acid (500X).

5 9

COMPOSITES

Composites cover a wide range of compositions,general grouped under the subcategoriesmetal-matrix composites (MMC), polymer-matrixcomposites (PMC) and ceramic-matrix compos-ites (CMC). Consequently, preparation schemesare quite difficult to generalize due to the extremerange of materials employed, differences in hard-ness and grinding/polishing characteristics; so,relief control is a big issue. Pull out problems arealso very common, especially with PMCs. Sec-tioning frequently produces considerable damagethat must be removed in the initial preparation step.Mounting with castable epoxy resin along withvacuum impregnation is frequently performed.

Composites

(top) Microstructure of a polymer-matrix composite,graphite fabric-reinforced polysulfone (unetched,polarized light, 100X). (bottom) Microstructure ofalumina fibers in an aluminum - lithium alloy matrix(500X, as-polished, Nomarski DIC).

(top) Microstructure of a metal-matrix composite,SiC in a CP titanium matrix (unetched, NomarskiDIC, 100X).

HELPFUL HINTS FOR

COMPOSITES

Sectioning damage canpropagate down the fibers and

can be difficult to remove. Prevent damageby using precision saw sectioning and dia-mond wafering blades with a very fineabrasive, specifically like 5 LC and 10 LCSeries. Thinner blades produce less damage.Section as close as possible to the plane ofinterest.

Composites can be damaged by compres-sion mounting. Castable mounting resins withlow peak temperature are preferred.

Avoid aggressive grinding steps by usinglonger times and finer abrasives.

(top) Synthetic foam made with hollow ceramicspheres added to 7075 aluminum (100X originalmagnification, magnification bar is 100 μm long).Etched with aqueous 0.5% HF. (bottom) Carbon-reinforced polymer composite (original at 100X;magnification bar is 100 μm long). Specimen is as-polished and viewed with Nomarski DICillumination.

6 0

Composites

Table 38. Four-Step Procedure for Ceramic-Matrix Composites

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 45-μm diamond MetaDi 6 (27) 240-300 Untilmetal-bonded disc fluid for coolant Contra PLane

UltraPol, UltraPad, 15-μm MetaDi 6 (27) 120-150 4:00or TexMet 1000 Supreme diamond Contra or 2000 cloths suspension*

UltraPol, TriDent, 6-μm MetaDi 6 (27) 120-150 3:00 or TexMet 1000 Supreme diamond Contraor 2000 cloths suspension*

UltraPol, 1-μm MetaDi 6 (27) 120-150 2:00ChemoMet Supreme diamond Contraor Nylon suspension andcloths ~0.05-μm MasterMet

colloidal silica(50:50 simultaneously)

*Plus MetaDi Fluid extender as desiredContra = Platen and specimen holder rotate in opposite directions

Table 37. Four-Step Procedure for Polymer-Matrix Composites

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet abrasive discs 320- (P400) grit 4 (18) 150-240 Until(waterproof paper) SiC water cooled Comp. Plane

UltraPol, or 6-μm MetaDi 4 (18) 120-150 4:00UltraPad Supreme diamond Comp.cloths suspension*

TexMet 1000, 3-μm MetaDi 5 (22) 120-150 4:00TriDent or Supreme diamond Comp.Nylon cloths suspension*

ChemoMet or ~0.05-μm MasterMet 6 (27) 120-150 2:00-4:00MicroCloth colloidal silica or Comp.pads MasterPrep

alumina suspension

*Plus MetaDi Fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)

Table 36. Four-Step Procedure for Metal-Matrix Composites

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep 30-μm diamond 5 (22) 240-300 Untilresin-bonded disc water cooled Contra Plane

UltraPol, or 9-μm 5 (22) 150-180 4:00UltraPad cloths MetaDi Supreme Contra

diamond suspension*

TriDent 3-μm 6 (27) 120-150 3:00cloth MetaDi Supreme Comp.

diamond suspension*

ChemoMet ~0.05-μm MasterMet 6 (27) 120-150 2:00cloth colloidal silica or Contra

MasterMet alumina suspension

*Plus MetaDi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

6 1

PRINTED CIRCUIT BOARDS

The vast majority of printed circuit boards (PCB’s)are of a rigid variety, the bulk of which are com-posed of layers of woven glass fiber cloth in apolymeric matrix. Flex circuits, which are becom-ing quite common, do not typically contain glassfiber, but instead, their bulk is often composed oflayers of polyimide. The circuitry in both types ofboards is composed of plated and/or foil metal.The metal used is generally copper, while in a fewcases, gold and/or nickel plating may be present.Furthermore, depending upon whether the boardshave undergone assembly or shock testing, sol-ders of various compositions might also bepresent.

Luckily for the metallographer, the variety of ma-terials present in PCB’s generally do notcomplicate the preparation methods due to the fact

that extremely hard and brittle materials are notcommonly found in the boards. This changes, how-ever, when ‘packed’ boards with ceramic orsemiconductor components must be sectioned.Please refer to methods of preparation forceramics or microelectronic materials for informa-tion on these cases. A generalized method fornonpacked PCB’s of both the rigid and flex variet-ies is presented in Table 39.

Quality control of manufactured PCB’s oftendemands statistical analysis based on platingthickness measurements taken from the centersof plated through-holes. However, to generateenough data for a representative sampling,numerous specimens must be prepared. This iseasiest when coupons are taken from PCB’s andare ganged together in precise alignment. In thismanner, the through holes in all of the couponscan be sectioned at one time, with one set ofconsumables. The Buehler NelsonZimmerSystemallows this type of automation with few adjustmentsto the method listed below.

Printed Circuit Boards

Table 39. Five-Step Procedure for Non-Packed Printed Circuit Boards

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet abrasive 180- (P180) grit SiC 5 (22) 120-150 Untildiscs (waterproof paper) water cooled Contra Plane

CarbiMet 320- (P400) grit SiC 5 (22) 120-150 1:30abrasive discs water cooled Contra

CarbiMet 600- (P1200) grit SiC 5 (22) 120-150 1:30abrasive discs water cooled Contra

TexMet 1000, 3-μm MetaDi II 5 (22) 150-240 1:30or TriDent cloth diamond paste* Contra

MicroCloth 0.05-μm MasterPrep 5 (22) 100-150 1:15cloth alumina suspension Contra

*Plus MetaDi Fluid extender as desiredContra = Platen and specimen holder rotate in opposite directions

Micrograph showing a plated-through hole in aprinted circuit board (resin and fibers) with copperfoil, electroless-plated copper and Pb-Sn solder(equal parts ammonium hydroxide and 3% hydrogenperoxide, 200X).

HELPFUL HINTS FOR

PRINTED CIRCUIT

BOARDS

To improve penetration ofacrylic mounting material into the through-holes of a PCB specimen, dip the “pinned”PCB coupons in liquid hardener immediatelyprior to mounting in acrylic. This will “wick”the acrylic into the through-holes and mini-mize the presence of problematic bubblesin these critical locations.

6 2

ELECTRONIC MATERIALS

The term, ‘microelectronic materials’ encompassesan extremely wide range of materials. This is dueto the fact that most microelectronic devices arecomposites, containing any number of individualcomponents. For example, present day micropro-cessor failure analysis might require themetallographer to precisely cross section througha silicon chip plated with multiple thin-film layersof oxides, polymers, ductile metals such as cop-per or aluminum, and refractory metals such astungsten and/or titanium-tungsten. In addition, thepackaging of such a device might contain materi-als of such varying mechanical properties astoughened aluminum oxide and solder. The soldermaterials may have compositions ranging up to97% lead. With such a vast number of materialsincorporated into a single device, and with thesematerials having such highly disparate mechani-cal properties, it is virtually impossible to developa general method for achieving perfect metallo-graphic results. Instead, we must focus on a fewindividual materials, and develop a philosophy of

preparation in which we give our attention specifi-cally to the materials of interest.

First and foremost in the class of ‘microelectronicmaterials’ is silicon. Silicon is a relatively hard,brittle material, which does not respond well togrinding with large silicon carbide abrasives. Sili-con carbide papers contain strongly bondedabrasive particles which, when they collide withthe leading edge of a piece of silicon, create sig-nificant impact damage. In addition, they createtensile stresses on the trailing edge of silicon,which results in deep and destructive cracking.Cutting close to the target area is preferable togrinding, but to accurately approach the target areawithin a silicon device, fine grinding is still neces-sary. This is the point at which silicon preparationdivides into two distinct classes. The first is prepa-ration through traditional metallographictechniques. The second is preparation of unen-capsulated silicon chips (or die), using specialfixturing and abrasives. This second category is aspecialized field of metallography, which is dis-cussed in other texts [8] and is beyond the scopeof this book. Therefore, we will focus on standardmetallographic techniques throughout this discus-sion.

Standard preparation of epoxy encapsulatedsilicon is similar to general metallographicpreparation, with the exception that only veryfine grades of silicon carbide paper are used.Table 40 illustrates a typical process of prepara-tion.

Electronic Materials

Table 40. Three-Step Procedure for Preparing Silicon in Microelectronic Devices

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

UltraPrep, Type A 15-μm diamond Manual 100-150 Untildiamond lapping film water cooled Planeon TriDent cloth

TexMet 1000 3-μm diamond MetaDi II 7 (31) 240-300 2:00pad diamond paste/ Comp

MasterPrep as lubricant

ChemoMet 0.02-μm 5 (22) 100-150 1:30pad MasterMet 2 Contra

colloidal silica

Comp. = Complementary(platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Note: The first step is manual so that you can monitor the progress of grinding relative to the feature of interest. The remaining steps areaccomplished using a system that can apply individual force to the back of the spceimen. The first step can be automated if there is asignificant amount of material to remove.

Aluminum circuitry on silicon adjacent to mineralfilled epoxy packaging, (1000x).

6 3

Using silicon carbide grinding paper with abrasivesizes coarser than 600- (P1200) grit may causeserious damage to the silicon. Therefore, thismethod often requires precision cutting or cleav-ing near to the target area of the specimen prior tothe initial grinding step. Care must be taken to as-sure that only precision saw blades designedespecially for microelectronic materials are usedwhen cutting.

Silicon responds very well to the chemo-mechani-cal polishing effects of colloidal silica, but not aswell to aluminum oxide. However, there are in-stances when an aluminum oxide product, suchas the MasterPrep suspension, should be usedfor final polishing. This relates back to the philoso-phy of preparation mentioned earlier. An exampleof the appropriate use of MasterPrep aluminawould be: when a silicon die is attached to a leadframe material such as nickel-plated copper, andthe lead frame and die attach materials are thematerials of interest. In this case, we are not so

concerned with the surface finish of the silicon,but instead, we want to be sure that the nickeldoesn’t smear, and that we are able to discern thetrue microstructure of the lead frame materials (re-fer to section listing preparation methods forcopper and nickel).

When preparing a silicon device for the purposeof inspecting the metalized, thin film circuitry,the techniques would be the same as those listedabove. Again, the choice of final polishing agentwould be determined entirely by the features ofinterest. For instance, aluminum circuitry respondsextremely well to the chemomechanical effects ofcolloidal silica, but tungsten and titanium-tungstenare not as effectively polished with colloidal silicaas are the materials by which they are surrounded.Therefore, colloidal silica often causes such re-fractory metals to stand in relief against a wellpolished background. This results in edge round-ing of the refractory metals and can make interfaceanalysis quite difficult. An alternative is to use anextremely fine diamond suspension for final pol-ishing in order to reduce these effects.

When lead alloy solders are the materials of inter-est in a microelectronic package, we can oftenseparate them into two general categories. The firstis the eutectic, or near-eutectic solders. These tendto be ductile, but they respond well to the samegeneral preparation listed in Table 40. (During the3-μm polishing step, the best results are obtainedby using diamond paste.

Electronic Materials

Table 41. Three-Step Procedure for High-Temperature Solder on Ceramic Devices

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

ApexHercules S 9-μm MetaDi II 7 (31) 100-150 Untilrigid grinding disc diamond paste/ Comp. Plane

MasterPrepas lubricant*

TexMet 3-μm MetaDi II 7 (31) 240-300 Until1000 diamond paste/ Comp. all embedded

MasterPrep 9-μm diamondas lubricant is removed

ChemoMet 0.02-μm 5 (22) 100-150 2:00pad MasterMet 2 Contra

colloidal silica

*Apply sparinglyComp. = Complementary(platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

High temperature (96% Pb) solder under flip chip onceramic substrate, (200x).

6 4

Pastes contain a waxy base that reduces diamondembedding in ductile metals. Limit the use of ex-tenders in such cases. A few drops of waterlubrication does not break down the waxy base,and therefore it retains the desired properties ofdiamond paste.)

The second category of lead solders is the hightemperature solders (in the range of 90% to 97%lead). These are difficult to prepare, and requirespecial consideration. This is especially true whenthey are incorporated into a device that containsa ceramic package. The mechanical properties ofthese toughened ceramics require aggressivegrinding techniques, which cause ceramic debris(and often, the abrasive used for the grinding op-eration) to became embedded in the solder. Siliconcarbide abrasives are a poor choice for grindingin such cases, as silicon carbide is only slightlyharder (and significantly more brittle) than the typi-cal packaging material they are intended to grind.As the silicon carbide abrasive fractures, it pro-duces elongated shards, which embed deeply intohigh-temperature solders. In addition, they do notgrind the ceramic package effectively, and there-fore produce extreme edge rounding at theinterface with the solder. Diamond grinding pro-duces a more desired result since diamondabrasives are blockier in shape, and are there-fore easier to remove after they have embeddedthemselves. In addition, they are capable of cre-ating a flat surface when ceramic packages arepresent.

High removal rates and good surface finisheshave been obtained on ceramic packages whenusing diamond paste on the ApexHercules Sdiscs.

Polishing of high-temperature solders attached toceramics often produces undesired edge round-ing. However, this cannot be completely avoided.The cushioning effects of polishing pads will tendto cause the abrasive to preferentially removematerial from ductile materials faster than from hard,brittle materials. By using stiff, flat polishing pads,one can reduce this effect, but at the expense ofhaving additional material embedded in the high-temperature solder. One effective preparationtechnique is to perform a grind-polish-grind-polishtype of procedure. In this technique, a polishingstep, which utilizes a napped polishing cloth, isused in conjunction with a step-down in abrasivesize from the previous grinding step. This processis used to polish out the embedded material. Thenext step is a grinding process, utilizing anotherstep-down in abrasive size, to flatten the speci-men again. This continues to a sufficiently finepolishing stage at which final polishing will pro-duce the desired result. Table 41 illustrates anexample of such a process.

Electronic Materials

HELPFUL HINTS FOR

ELECTRONIC

MATERIALS

Consider what area of the specimen is mostcritical when choosing a final polishingsuspension. For example, the chemo- me-chanical nature of colloidal silicas (such asMasterMet and MasterMet 2) make themideal for polishing silicon, glass, oxides,aluminum, and to some degree copper. How-ever, polishing nickel platings and gold ballbonds with these agents often results insmearing. For such non-reactive materials,use MasterPrep aluminum oxide final polish-ing suspension to produce a flat, scratch-freesurface.

Silicon flip chip, solder balls, on PC board, (200x).

6 5

POLYMERS

Plastics and polymers are normally quite soft.Many different sectioning methods have been used.A sharp razor blade or scalpel, or even a pair ofscissors, can be used. Microtomes have beenused, and a surface cut after refrigeration in liquidnitrogen requires very little subsequent prepara-tion. A jeweler’s saw may be used, although thesurface is somewhat rough. The precision sawproduces excellent surfaces, while an abrasivecut-off saw yields a bit more damage and a roughersurface. Blades and wheels for sectioning poly-mers are available from Buehler. Damage fromsectioning is quite easy to remove.

Surface quality can be degraded by abrasion fromthe debris produced during grinding and polishing.Mounted specimens are much easier to preparethan nonmounted specimens. Castable resins arepreferred as the heat from a mounting press maydamage or alter the structure of the specimen.

However, there may be a visibility problem if atransparent, clear polymeric specimen ismounted in a clear, transparent epoxy. In thiscase, EpoColor resin, with its deep red color, willproduce excellent color contrast between speci-men and mount in darkfield or polarized light. Dueto their light weight, a specimen may float in theepoxy filled mold. To prevent this, put doublesided adhesive tape on the bottom of the moldcup and set the specimen on top of the tape. Al-ways use practices that minimize the exothermduring polymerization.

Preparation of plastics and polymers for micro-structural examination follows the same basicprinciples as for other materials. Automated pol-ishing devices are preferred to manualpreparation. Rough grinding abrasives are un-necessary, even for the planar grinding step.Pressures should be lighter than used for mostmetals. Water is generally used as the coolant,

Table 42. Generic Procedure for Preparing Plastic and Polymeric Specimens

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

CarbiMet 320- (P400) grit SiC 4 (18) 200-250 Untilabrasive discs water cooled Contra Plane(waterproof paper)

CarbiMet 400- (P600) grit SiC 4 (18) 200-250 0:30abrasive discs water cooled Contra

CarbiMet 600- (P1200) grit SiC 4 (18) 200-250 0:30abrasive discs water cooled Contra

CarbiMet 800- (P1500) grit SiC 4 (18) 200-250 0:30abrasive discs water cooled Contra

TexMet 1000 or 3-μm MetaDi II 5 (22) 100-120 4:00TriDent diamond paste Comp.cloths and MetaDi fluid

MasterTex 0.05-μm MasterPrep 3 (13) 100-120 4:00cloth alumina suspension Contra

Comp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Polymers

Micrograph of high-density polyethylene containinga filler (white particles) (unetched, Nomarski DIC,100X).

Micrograph showing multi-layered polymer meatcasing mounted in epoxy (dark field, 100X).

6 6

although some plastics and polymers may reactwith water. In such cases, use a fluid that will notreact with the particular plastic or polymer. Em-bedding can be a problem with plastics andpolymers. ASTM E 2015 (Standard Guide forPreparation of Plastics and Polymeric Specimensfor Microstructural Examination) describesprocedures for preparing several types of plas-tics and polymers. Table 42 contains a genericprocedure for preparing many plastic and polymericmaterials. Depending upon the material and thesurface roughness after sectioning, it may be pos-sible to start grinding with 400- (P600) or even600- (P1200) grit SiC paper.

If flatness is critical, the final step can be altered.Use a TexMet pad with MasterPrep alumina abra-sive slurry at 10 lbs. pressure (44N) per specimen,120 rpm, contra rotation, for 3 minutes.

HELPFUL HINTS FOR

POLYMERS

Plastics often have little con-trast from the mounting

media when viewed under a microscope.This makes thickness measurement anddiscerning edges difficult. Mounting inEpoColor dye enhanced epoxy will resolvethis situation by producing excellent colorcontrast between the specimens and mount.

Polymers and plastics may react with fluidsduring sectioning and preparation. Alwayscheck your particular polymer and choosethe proper fluid, usually a water, or oil basedsolution to avoid a reaction.

Polymers

6 7

screen magnifications that may make detectionof fine structures easier. However, resolution isnot improved beyond the limit of 0.2-0.3-μm for thelight microscope. Microscopic examination of aproperly prepared specimen will clearly revealstructural characteristics such as grain size, seg-regation, and the shape, size, and distribution ofthe phases and inclusions that are present. Ex-amination of the microstructure will reveal priormechanical and thermal treatments give the metal.Many of these microstructural features are mea-sured either according to established imageanalysis procedures, e.g., ASTM standards, or in-ternally developed methods.

Etching is done by immersion or by swabbing (orelectrolytically) with a suitable chemical solutionthat essentially produces selective corrosion.Swabbing is preferred for those metals andalloys that form a tenacious oxide surface layerwith atmospheric exposure such as stainlesssteels, aluminum, nickel, niobium, and titanium andtheir alloys. It is best to use surgical grade cottonthat will not scratch the polished surface. Etch timevaries with etch strength and can only be deter-mined by experience. In general, for highmagnification examination the etch depth shouldbe shallow; while for low magnification examina-tion a deeper etch yields better image contrast.Some etchants produce selective results in thatonly one phase will be attacked or colored. A vastnumber of etchants have been developed; thereader is directed to references 1-3, 9 and ASTME 407. Table 43 lists some of the most commonlyused etchants for the materials described in thisbook.

Etchants that reveal grain boundaries are very im-portant for successful determination of the grainsize. Grain boundary etchants are given in [1-3,9]. Problems associated with grain boundary etch-ing, particularly prior austenite grain boundaryetching, are given in [2, 10 and 11]. Measurementof grain size in austenitic or face-centered cubicmetals that exhibit annealing twins is a commonlyencountered problem. Etchants that will revealgrain boundaries, but not twin boundaries, are re-viewed in [2].

ETCHING

Metallographic etching encompasses all pro-cesses used to reveal particular structuralcharacteristics of a metal that are not evident inthe as-polished condition. Examination of a prop-erly polished specimen before etching may revealstructural aspects such as porosity, cracks, andnonmetallic inclusions. Indeed, certain constitu-ents are best measured by image analysis withoutetching, because etching will reveal additional,unwanted detail and make detection difficult orimpossible. The classic examples are the mea-surement of inclusions in steels and graphite incast iron. Of course, inclusions are present in allmetals, not just steels. Many intermetallic precipi-tates and nitrides can be measured effectively inthe as-polished condition.

In certain nonferrous alloys that have non-cubiccrystallographic structures (such as beryllium,hafnium, magnesium, titanium, uranium and zir-conium), grain size can be revealed adequately inthe as polished condition using polarized light.Figure 30 shows the microstructure of cold-drawnzirconium viewed in cross-polarized light. This pro-duces grain coloration, rather than a “flat etched”appearance where only the grain boundaries aredark.

Etching ProceduresMicroscopic examination is usually limited to amaximum magnification of 1000X — the approxi-mate useful limit of the light microscope, unlessoil immersion objectives are used. Many imageanalysis systems use relay lenses that yield higher

Etching

Figure 30. Mechanical twins at the surface of hotworked and cold drawn high-purity zirconiumviewed with polarized light (200X).

6 8

either the anodic or cathodic constituent in amicrostructure. Tint etchants are listed and illus-trated in several publications [1-3, 16-21].

A classic example of the different behavior ofetchants is given in Figure 31 where low-carbonsheet steel has been etched with the standard nitaland picral etchants, and also a color tint etch. Etch-ing with 2% nital reveals the ferrite grain boundariesand cementite. Note that many of the ferrite grainboundaries are missing or quite faint; a problemthat degrades the accuracy of grain size ratings.Etching with 4% picral reveals the cementite ag-gregates (one could not call this pearlite as it istoo nonlamellar in appearance and some of thecementite exists as simple grain boundary films)but no ferrite grain boundaries. If one is interestedin the amount and nature of the cementite (whichcan influence formability), then the picral etch isfar superior to the nital etch as picral revealed onlythe cementite. Tint etching with Beraha’s solution(Klemm’s I could also be used) colored the grainsaccording to their crystallographic orientation. Withthe development of color image analyzers, thisimage can now be used quite effectively to pro-vide accurate grain size measurements since allof the grains are colored.

Figure 32 shows a somewhat more complex ex-ample of selective etching. The micrograph showsthe ferrite-cementite-iron phosphide ternary eutec-tic in gray iron. Etching sequentially with picral andnital revealed the eutectic, Figure 32a, surroundedby pearlite. Etching with boiling alkaline sodium

picrate, Figure 32b, colored the cementite phaseonly, including in the surrounding pearlite (a highermagnification is required to see the very finelyspaced cementite that is more lightly colored).Etching with boiling Murakami’s reagent, Figure32c, colors the iron phosphide darkly and will

Selective EtchingImage analysis work is facilitated if the etchantchosen improves the contrast between the fea-ture of interest and everything else. Thousands ofetchants have been developed over the years, butonly a small number of these are selective in na-ture. Although the selection of the best etchant,and its proper use, is a very critical phase of theimage analysis process, only a few publicationshave addressed this problem [12-14]. Selectiveetchants, that is, etchants that preferentiallyattack or color a specific phase, are listed in[1-3, 9, 13 and 14] and illustrated in 13 and 14.Stansbury [15] has described how potentiostaticetching works and has listed many preferentialpotentiostatic etching methods. The potentiostatoffers the ultimate in control over the etching pro-cess and is an outstanding tool for this purpose.Many tint etchants act selectively in that they color

Figure 32a. The ternary eutectic (ααααα-Fe3C-Fe3P) ingray cast iron revealed by etching (a, top) in picraland nital to “outline” the phases

b

a

a

Etching

Figure 31. Microstructure of low-carbon sheet steeletched with (a, top) 2% nital, (b, middle) 4% picral;and (c, bottom) Beraha’s reagent (100 mL water, 10g Na2S2O3 and 3 g K2S2O5) at 100X.

c

6 9

lightly color the cementite after prolonged etching.The ferrite could be colored preferentially if Klemm’sI was used.

Figure 33. Microstructure of a duplex stainlesssteel revealed by electrolytic etching with (c)aqueous 60% HNO3 (1 V dc, 20 s); with (d) aqueous10% oxalic acid (6 V dc, 75 s); with (e) aqueous10% CrO3 (6 V dc, 30 s); and, with (f) aqueous 2%H2SO4 (5 V dc, 30 s) at 200X.

Figure 33. Microstructure of a duplex stainlesssteel revealed using (a, top) alcoholic 15% HCl byimmersion (30 min.); and, with (b, bottom)glyceregia by swabbing (2 min.) at 200X.

Selective etching has been commonly applied tostainless steels for detection, identification andmeasurement of delta ferrite, ferrite in dual phasegrades, and sigma phase. Figure 33 shows ex-amples of the use of a number of popular etchants

to reveal the microstructure of a dual phase stain-less steel in the hot rolled and annealed condition.Figure 33a shows a well delineated structure whenthe specimen was immersed in ethanolic 15% HClfor 30 minutes. All of the phase boundaries areclearly revealed, but there is no discrimination be-tween ferrite and austenite. Twin boundaries in theaustenite are not revealed. Glyceregia, a popular

Figure 32b. (top) boiling alkaline sodium picrate tocolor the cementite. Figure 32c. (bottom) boilingMurakami’s reagent to color the phosphide (200X).

b

a

e

d

c

f

b

c

Etching

7 0

Figure 33. Selective coloring of ferrite in a duplexstainless steel by electrolytic etching with (g)aqueous 20% NaOH (4 V dc, 10 s) and with (h)aqueous 10 N KOH, (3 V dc, 4 s); and by immersioncolor etching (i) with Beraha’s reagent (100 mLwater, 10 mL HCl and 1 g K2S2O5) at 200X.

etchant for stainless steels, was not suitable forthis grade, Figure 33b, as it appears to be ratherorientation sensitive. Many electrolytic etchantshave been used for etching stainless steels, butonly a few have selective characteristics. Of thefour shown in Figures 33c to f, only aqueous 60%nitric acid produced any gray level discriminationbetween the phases, and that was weak. All re-vealed the phase boundaries nicely, however. Twoelectrolytic reagents, shown in Figures 33g andh, are commonly used to color ferrite in dual phasegrades and delta ferrite in martensitic grades. Ofthese, aqueous 20% sodium hydroxide, Figure33g, usually gives more uniform coloring of theferrite. Murakami’s and Groesbeck’s reagents havealso been used for this purpose. Tint etchants havebeen developed by Beraha that color the ferritephase nicely, as demonstrated in Figure 33i.

Selective etching techniques are not limited to iron-based alloys, although these have more thoroughlydeveloped than for any other alloy system. Selec-tive etching of beta phase in alpha-beta copperalloys has been a popular subject. Figure 34 illus-trates coloring of beta phase in Naval Brass (UNS46400) using Klemm’s I reagent. Selective etch-ing has a long historical record for identification of

Figure 34. Beta phase colored in Naval Brass (Cu-39.7% Zn-0.8% Sn) by immersion color etching withKlemm’s I reagent (200X).

intermetallic phases in aluminum alloys. Thismethod was used for many years before the de-velopment of energy-dispersive spectroscopy.Today, it is still useful for image analysis work. Fig-ure 35 shows selective coloration of theta phase,CuAl2, in the Al-33% Cu eutectic alloy. As a finalexample, Figure 36 illustrates the structure of a

simple WC-Co sintered carbide, cutting tool. In theas-polished condition, Figure 36a, the cobalt bindercan be seen faintly against the more grayish tung-sten carbide grains. A few particles of graphiteare visible. In Figure 36b, light relief polishing hasbrought out the outlines of the cobalt binder phase,but this image is not particularly useful for imageanalysis. Etching in a solution of hydrochloric acidsaturated with ferric chloride, Figure 36c, attacksthe cobalt and provides good uniform contrast formeasurement of the cobalt binder phase. Follow

Figure 35. Theta phase, CuAl2, colored in the α-Al/CuAl2 eutectic in an as-cast Al-33% Cu specimen byimmersion color etching with the Lienard andPacque etch (200 mL water, 1 g ammoniummolybdate, 6 g ammonium chloride) at 1000X.

h

g

i

Etching

7 1

Electrolytic Etching and AnodizingThe procedure for electrolytic etching is basicallythe same as for electropolishing except that volt-age and current densities are considerably lower.The specimen is made the anode, and some rela-tively insoluble but conductive material such asstainless steel, graphite, or platinum is used forthe cathode. Direct current electrolysis is used formost electrolytic etching. Electrolytic etching iscommonly used with stainless steels, either toreveal grain boundaries without twin boundaries,or for coloring ferrite (as illustrated) in Figure 33,delta ferrite, sigma or chi phases. Anodizing is aterm applied to electrolytic etchants that developgrain coloration when viewed with crossed-polar-ized light, as in the case of aluminum, tantalum,titanium, tungsten, uranium, vanadium and zirco-nium [2]. Figure 37 shows the grain structure of5754 aluminum revealed by anodizing withBarker’s reagent, viewed with crossed-polarizedlight. Again, color image analysis makes thisimage useful now for grain size measurements.

Heat TintingAlthough not commonly utilized, heat tinting[2] is an excellent method for obtaining colorcontrast between constituents or grains. Anunmounted polished specimen is placed face upin an air-fired furnace at a set temperature andheld there as an oxide film grows on the surface.Interference effects, as in tint etching, create col-oration for film thicknesses within a certain range,about 30-500 nm. The observed color is a func

Figure 36. Microstructure of WC-Co cutting tool: a)as-polished revealing graphite particles; b) reliefpolished revealing the cobalt binder phase; c) afterimmersion in Chaporova’s etch (HCl saturated withFeCl3) to attack and “darken” the cobalt; and, d)after (c) plus Murakami’s reagent to outline the WCgrains (1000X).

ing this with Murakami’s reagent at room tempera-ture reveals the edges of the tungsten carbidegrains, useful for evaluation of the WC grain size,Figure 36d.

b

a

c

d

Etching

Figure 37. Grain structure of annealed 5754aluminum sheet revealed by anodizing with Barker’sreagent (30 V dc, 2 min.) and viewing with polarizedlight plus sensitive tint (100X).

7 2

HELPFUL HINTS FOR

ETCHING

Many etchants can be usedby swabbing or by immersion. Swabbing ispreferred for those specimens that form atight protective oxide on the surface in air,such as Al, Ni, Cr, stainless steels, Nb (Cb),Ti and Zr. However, if the etchant forms afilm, as in tint etchants, then immersion mustbe used as swabbing will keep the film fromforming. Keller’s reagent reveals the grainsize of certain aluminum alloys by forminga film. This will not occur if the etch is usedby swabbing.

Many etchants, and their ingredients, dopresent potential health hazards to the user.ASTM E 2014, Standard Guide on Metal-lography Laboratory Safety, describes manyof the common problems and how to avoidthem.

tion of the film thickness. Naturally, the thermalexposure cannot alter the microstructure. The cor-rect temperature must be determined by thetrial-and-error approach, but the procedure is re-producible and reliable. Figure 38 shows the grainstructure of CP titanium revealed by heat tinting.

Interference Layer MethodThe interference layer method [2], introduced byPepperhoff in 1960, is another procedure forobtaining a film over the microstructure thatgenerates color by interference effects. In thismethod, a suitable material is deposited on the pol-ished specimen face by vapor deposition toproduce a low-absorption, dielectric film with a highrefractive index at a thickness within the range forinterference. Very small differences in the naturalreflectivity between constituents and the matrix

Figure 38. Grain structure of annealed CP titaniumrevealed by heat tinting on a hot plate (100X,polarized light plus sensitive tint).

can be dramatically enhanced by this method.Suitable materials for the production of evapora-tion layers have been summarized in [22,23]. Thetechnique is universally applicable, but does re-quire a vacuum evaporator. Its main weakness isdifficulty in obtaining a uniformly coated large sur-face area for image analysis measurements.

Etching

7 3

Table 43. Commonly Used Etchants for Metals and Alloys3.

LIGHT METALS - Aluminum and AlloysComposition Comments

1. 95 mL water Keller’s reagent, very popular general purpose reagent for Al and Al2.5 mL HNO3 alloys, except high-Si alloys. Immerse sample 10-20 seconds, wash in1.5 mL HCI warm water. Can follow with a dip in conc. HNO3. Outlines all common1.0 mL HF constituents, reveals grain structure in certain alloys when used by

immersion.

2. 90-100 mL water General-purpose reagent. Attacks FeAl3, other constituents outlined. The0.1-10 mL HF 0.5% concentration of HF is very popular.

3. 84 mL water Graff and Sargent’s etchant, for grain size of 2XXX, 3XXX, 6XXX, and15.5 mL HNO3 7XXX wrought alloys. Immerse specimen 20-60 seconds with mild0.5 mL HF agitation.3 g CrO3

4. 1.8% fluoboric acid in water Barker’s anodizing method for grain structure. Use 0.5-1.5 A/in2, 30-45 Vdc. For most alloys and tempers, 20 seconds at 1 A/in2 and 30 V dc at 20 °Cis sufficient. Stirring not needed. Rinse in warm water, dry. Use polarizedlight; sensitive tint helpful.

Magnesium and Alloys

Composition Comments

5. 25 mL water Glycol etch, general purpose etch for pure Mg and alloys. Swab specimen75 mL 3-5 ethylene glycol seconds for F and T6 temper alloys, 1-2 minutes for T4 and 0 temper1 mL HNO3 alloys.

6. 19 mL water Acetic glycol etchant for pure Mg and alloys. Swab specimen 1-3 seconds60 mL ethylene glycol for F and T6 temper alloys, 10 seconds for T4 and 0 temper alloys. Reveals20 mL acetic acid grain boundaries in solution-treated castings and most wrought alloys.1 mL HNO3

7. 100 mL ethanol For Mg and alloys. Use fresh. Immerse specimen for 15-30 seconds.10 mL water Produces grain contrast.5 g picric acid

LOW MELTING POINT METALS - Sb, Bi, Cd, Pb, Sn and ZnComposition Comments

8. 100 mL water For Sb, Bi and alloys. Immerse specimen up to a few minutes.30 mL HCI2 g FeCI3

9. 100 mL water For Sb-Pb, Bi-Sn, Cd-Sn, Cd-Zn, and Bi-Cd alloys. Immerse specimen up25 mL HCI to a few minutes.8 g FeCI3

10. 95-99 mL ethanol For Cd, Cd alloys, Sn, Zn alloys, Pb and alloys, Bi-Sn eutectic alloy and Bi-1-5 mL HNO3 Cd alloys. Can add a few drops of zephiran chloride. Immerse sample. For

Pb and alloys, if a stain forms, wash in 10% alcoholic HCI.

11. 100 mL water Pollack’s reagent for Pb and alloys. Immerse specimen 15-30 seconds.10 g ammonium molybdate Other compositions used are: 100 mL: 9 g: 15 g and 100 mL: 10 g: 25 g.10 g citric acid

12. 100 mL water For Sn-based babbitt metal. Immerse specimen up to 5 minutes.2 mL HCI10 g FeCI3

13. 200 mL water Palmerton reagent for pure Zn and alloys. Immerse specimen up to 340 g CrO3 minutes. Rinse in 20% aqueous CrO3.3 g Na2SO4

14. 200 mL water Modified Palmerton reagent for Zn die-casting alloys. Immerse specimen10 g CrO3 for several seconds, rinse in 20% aqueous CrO3.1 g Na2SO4

3 When water is specified, always use distilled water. It is best to use reagent grade chemicals. Etchants can be quite dangerous and itis advisable to consult with a chemist when dealing with new etchants that may be potentially dangerous. ASTM E 2014, StandardGuide on Metallographic Laboratory Safety, is a valuable reference.

Etching

7 4

Table 43. Commonly Used Etchants for Metals and Alloys3.

REFRACTORY METALS: Ti, Zr, Hf, Cr, Mo, Re, Nb, Ta, W and VComposition Comments

15. 100 mL water Kroll’s reagent for Ti alloys. Swab specimen 3-10 seconds or immerse1-3 mL HF specimen 10-30 seconds.2-6 mL HNO3

16. 200 mL water For Ti, Zr and alloys. Swab or immerse specimen. Higher concentrations1 mL HF can be used but are prone to staining problems.

17. 30 mL lactic acid For Ti alloys. Swab specimen up to 30 seconds. Decomposes, do not store.15 mL HNO3 Good for alpha-beta alloys.30 mL HF

18. 30 mL HCI For Zr, Hf, and alloys. Swab specimen 3-10 seconds, or immerse specimen15 mL HNO3 up to 120 seconds.30 mL HF

19. 45 mL H2O (H2O2 or glycerol) Cain’s chemical polish and etch for Hf, Zr, and alloys. Can dilute aqueous45 mL HNO3 solution with 3-5 parts water to stain the structure (swab specimen) after8-10 mL HF chemical polishing. Chemically polish and etch specimen by swabbing 5-

20seconds. Use polarized light.

20. 60 mL HCI Aqua regia. For Cr and alloys. Immerse or swab specimen up to 1 minute.20 mL HNO3 Use under a hood with care, do not store.

21. 30 mL HCI Modified “Glyceregia”. For Cr and alloys. Immerse specimen up to a few45 mL glycerol minutes.15 mL HNO3

22. 100 mL water Murakami’s reagent. For Cr, Mo, Re, Ta-Mo, W, V and alloys. Use fresh, can10 g KOH or NaOH immerse sample for up to 1 minute.10 g K3Fe(CN)6

23. 70 mL water For Mo alloys. Immerse specimen 2 minutes. Wash with water and dry;20 mL H2O2 (30%) immersion produces colors, swabbing produces grain-boundary etch.10 mL H2SO4

24. 10-20 mL glycerol For Mo and Mo-Ti alloys. Immerse specimen for up to 5 minutes.10 mL HNO3

10 mL HF

25. 100 mL water For Mo-Re alloys. Use at 20 °C by immersion.5 g K3Fe(CN)6

2 g KOH

26. 50 mL acetic acid For Nb, Ta and alloys. Swab specimen 10-30 seconds.20 mL HNO3

5 mL HF

27. 50 mL water DuPont Nb reagent. For Nb-Hf and Nb alloys.14 mL H2SO4

5 mL HNO3

28. 50 mL water For Nb-Zr and Nb-Zr-Re alloys. Swab specimen.50 mL HNO3

1 mL HF

29. 30 mL lactic acid For Re and W-Re alloys. Swab specimen.10 mL HNO3

5 mL HF

30. 10 mL HF For V and alloys; grain-boundary etch for Ta alloys. Swab specimen. Equal10 mL HNO3 parts used for Ta and high Ta alloys.10-30 mL glycerol

IRON and STEELComposition Comments

31. 90-99 mL methanol or ethanol Nital. Most common etchant for Fe, carbon and alloy steels, cast iron.1-10 mL HNO3 Reveals alpha grain boundaries and constituents. Excellent for martensitic

structures. The 2% solution is most common, 5-10% used for high alloysteels (do not store). Use by immersion or swabbing of sample for up toabout 60 seconds.

Etching

7 5

32. 100 mL ethanol Picral. Recommended for structures consisting of ferrite and carbide. Does4 g picric acid not reveal ferrite grain boundaries. Addition of about 0.5-1% zephiran

chloride improves etch rate and uniformity.

33. 100 mL ethanol Vilella’s reagent. Good for ferrite-carbide structures. Produces grain5 mL HCI contrast for estimating prior austenite grain size. Results best on1 g picric acid martensite tempered at 572-932 °F (300-500 °C). Occasionally reveals

prior-austenite grain boundaries in high alloy steels. Outlines constituentsin stainless steels. Good for tool steels and martensitic stainless steels.

34. Saturated aqueous picric acid Bechet and Beaujard’s etch, most successful etchant for prior-austenitesolution grain plus small amount boundaries. Good for martensitic and bainitic steels. Many wetting agentsof a wetting agent have been used, sodium tridecylbenzene sulfonate is one of most

successful (the dodecyl version is easier to obtain and works as well). Useat 20-100 °C. Swab or immerse sample for 2-60 minutes. Etch in ultrasoniccleaner (see ref.2, pg. 219-223). Additions of 0.5g CuCl2 per 100mLsolution or about 1% HCI have been used for higher alloy steels toproduce etching. Room temperature etching most common. Lightly backpolish to remove surface smut.

35. 150 mL water Modified Fry’s reagent. Used for 18% Ni maraging steels, martensitic and50 mL HCI PH stainless steels.25 mL HNO3

1 g CuCl2

36. 100 mL water Alkaline sodium picrate. Best etch for McQuaid-Ehn carburized samples.25 g NaOH Darkens cementite. Use boiling for 1-15 minutes or electrolytic at 6 V dc,2 g picric acid 0.5 A/in2, 30-120 seconds. May reveal prior-austenite grain boundaries in

high carbon steels when no apparent grain boundary film is present.

37. 3 parts HCI “Glyceregia”. For austenitic stainless steels. Reveals grain structure,2 parts glycerol outlines sigma and carbides. Mix fresh, do not store. Use by swabbing.1 part HNO3

38. 100 mL ethanol Kalling’s no. 2 (“waterless” Kalling’s) etch for austenitic and duplex100 mL HCI stainless steels. Ferrite attacked readily, carbides unattacked, austenite

5 g CuCl2 slightly attacked. Use at 20 °C by immersion or swabbing. Can be stored.

39. 15 mL HCI Acetic glyceregia. Mix fresh; do not store. Use for high alloy stainless10 mL acetic acid steels.5 mL HNO3

2 drops glycerol

40. 100 mL water Murakami’s reagent. Usually works better on ferritic stainless grades than10 g K2Fe(CN)6 onaustenitic grades. Use at 20 °C for 7-60 seconds: reveals carbides si10 g KOH or NaOH sigma faintly attacked with etching up to 3 minutes. Use at 80°C (176°F) to

boiling for 2-60 minutes: carbides dark, sigma blue (not always attacked),ferrite yellow to yellow-brown, austenite unattacked. Do not always getuniform etching.

41. 100 mL water Use for stainless steels at 6 V dc. Carbides revealed by etching for10 g oxalic acid 15-30 seconds, grain boundaries after 45-60 seconds, sigma

outlined after 6 seconds. 1-3 V also used. Dissolves carbides, sigmastrongly attacked, austenite moderately attacked, ferrite unattacked.

42. 100 mL water Used to color ferrite in martensitic, PH or dual-phase stainless steels.20 g NaOH Use at 3-5 V dc, 20 °C, 5 seconds, stainless steel cathode. Ferrite outlined

and colored tan.

43. 40 mL water Electrolytic etch to reveal austenite boundaries but not twin boundaries in60 mL HNO3 austenitic stainless steels (304, 316, etc.). Voltage is critical. Pt cathode

preferred to stainless steel. Use at 1.4 V dc, 2 minutes (see ref. 2, pgs. 235,238 and 239).

COPPER, NICKEL and COBALT: Copper and AlloysComposition Comments

44. 25 mL NH4OH General purpose grain contrast etch for Cu and alloys (produces a flat etch25 mL water (optional) for some alloys). Use fresh, add peroxide last. Use under a hood. Swab25-50 mL H2O2 (3%) specimen 5-45 seconds.

45. 100 mL water General purpose etch for Cu and alloys. Immerse or swab for 3-6010 g ammonium persulfate seconds. Reveals grain boundaries but is sensitive to crystallographic

orientation.

Table 43. Commonly Used Etchants for Metals and Alloys3.

Etching

7 6

Table 43. Commonly Used Etchants for Metals and Alloys3.

46. 100 mL water General purpose etch for Cu and alloys, particularly Cu-Be alloys.3 g ammonium persulfate1 mL NH4OH

47. 70 mL water Excellent general purpose etch, reveals grain boundaries well. Immerse5 g Fe(NO3)3 specimen 10-30 seconds.25 mL HCI

Nickel and Alloys

Composition Comments

48. 5 g FeCl3 Carapella’s etch for Ni and Ni-Cu (Monel) alloys. Use by immersion2 mL HCI or swabbing.99 mL ethanol

49. 40-80 mL ethanol Kalling’s no.2 etch (“waterless” Kalling’s) for Ni-Cu alloys and superalloys.40 mL HCI Immerse or swab specimen up to a few minutes.2 g CuCI2

50. 50 mL water Marble’s reagent for Ni, Ni-Cu, and Ni-Fe alloys and superalloys. Immerse50 mL HCI or swab sample 5-60 seconds. Reveals grain structure of superalloys.10 g CuSO4

51. 15 mL HCI “Glyceregia”, for superalloys and Ni-Cr alloys. Swab specimen for 5-6010 mL glycerol seconds. Mix fresh. Do not store. Use under a hood.5 mL HNO3

52. 60 mL glycerol Modified Glyceregia for superalloys. Reveals precipitates. Use under hood;50 mL HCI do not store. Add HNO3 last. Discard when dark yellow. Immerse or swab10 mL HNO3 specimen 10-60 seconds.

Cobalt and AlloysComposition Comments

53. 60 mL HCI For Co and alloys. Mix fresh and age 1 hour before use. Immerse spe15 mL water specimen for up to 30 seconds. Do not store.15 mL acetic acid15 mL HNO3

54. 200 mL ethanol General etch for Co and alloys. Immerse specimen 2-4 minutes.7.5 mL HF2.5 mL HNO3

55. 50 mL water Marble’s reagent, for Co high temperature alloys. Immerse or swab spe50 mL HCI specimen for up to 1 minute.10 g CuSO4

56. 80 mL lactic acid For Co alloys. Use by swabbing.10 mL H2O2 (30%)10 mL HNO3

PRECIOUS METALS: Au, Ag, Ir, Os, Pd, Pt, Rh and RuComposition Comments

57. 60 mL HCI For gold, silver, palladium and high noble metal alloys. Use under hood.40 mL HNO3 Immerse specimen up to 60 seconds. Equal parts of each acid also used.

58. 60 mL HCI Aqua regia for pure gold, Pt and alloys, some Rh alloys. Use boiling for up20 mL HNO3 to 30 minutes.

59. 1-5 g CrO3 For Au, Ag, Pd and alloys. Swab or immerse specimen up to 60 seconds.100 mL HCI

60. 30 mL water For pure Pt. Use hot, immerse specimen up to 5 minutes.25 mL HCI5 mL HNO3

61. Conc. HCI For Rh and alloys. Use at 5 V ac, 1-2 minutes, graphite cathode,Pt lead wires.

Etching

7 7

Table 43. Commonly Used Etchants for Metals and Alloys3.

62. Solution a For Ru. Mix 4 parts solution a to 1 part solution b, use 5-20 V ac, 1-2 100 mL water minutes, graphite cathode, Pt-lead wires. 40 g NaCISolution b Conc. HCI

63. 50 mL NH4OH For pure Ag, Ag solders, and Ag-Pd alloys. Mix fresh. Swab specimen up to20 mL H2O2 (3%) 60seconds. Discard after etching. Use a 50:50 mix for sterling silver; for

fine silver, use 30% conc. hydrogen peroxide.

64. Solution a For gold alloys up to 18 karat. Mix equal amounts of a and b directly on 10 g NaCN the specimen using eye droppers. Swab and replenish the etchants 100 mL water until the desired etch level is obtained. If a brown stain forms, swab withSolution b a to remove it. H2O2 (30%)

SINTERED CARBIDESComposition Comments

65. 100 mL water Murakami’s reagent, for WC-Co and complex sintered carbides. Immerse10 g KOH or NaOH specimen seconds to minutes. Use 2-10 seconds to identify eta phase10 g K3Fe(CN)6 (colored). Longer times attack eta. Reveals phase and grain boundaries.

Normally used at 20 °C.

66. 97 mL water For WC, MO2C, TiC, or Ni in sintered carbides. Use boiling for up to 603 mL H2O2 (30%) seconds. For coarse carbides or high Co content, use short etch time.

67. 15 mL water For WC, TiC, TaC, and Co in sintered carbides. Use at 20 °C for 5-3030 mL HCI seconds.

15 mL HNO3

15 mL acetic acid

68. 100 mL H2O To darken Co (or Ni) binder phase. Mix fresh. Swab 10 seconds.3 g FeCl3

CERAMICS and NITRIDESComposition Comments

69. Phosphoric acid For alumina, Al2O3, and silicon nitride, Si3N4. Use by immersion at 250 °Cfor up to a few minutes for Al2O3 or up to 15 minutes for Si3N4.

70. 100 mL water For magnesia, MgO. Use by immersion at 25-60 °C for several minutes.15 mL HNO3

71. 100 mL water For titania, TiO2. Immerse up to a few minutes.5 g NH4 FHF4 mL HCI

72. 50 mL water For zirconia, ZrO2. Use by immersion in boiling solution for up to 5 minutes.50 mL H2SO4

73. HCI For calcia, CaO, or MgO. Immerse for up to 6 minutes.

74. HF For Si3N4, BeO, BaO, MgO, ZrO2 and Zr2O3. Immerse for up to 6 minutes.

PLASTICS and POLYMERSComposition Comments

75. 100 mL water For polypropylene (PP). Immerse for up to several hours at 70°C.60 g CrO3

76. HNO3 For polyethylene (PE). Immerse for up to a few minutes.

77. Xylol Reveals spherolites in polyethene (PE). Immerse for up to a few days at70 °C.

78. 70 mL water For polyoxymethylene (POM). Immerse up to 20 seconds.30 mL HCI

79. Xylene For polyamid (PA) and polyethylene (PE). Use at 70 °C for 60 seconds.For Nylon 6, use at 65-70 °C for 2-3 minutes. For Nylon 6, use at 75 °C for3-4 minutes.

Etching

7 8

LIGHT OPTICAL MICROSCOPY

Metallurgical microscopes differ from biologicalmicroscopes primarily in the manner by which thespecimen is illuminated due to the opacity of met-als. Unlike biological microscopes, metallurgicalmicroscopes must use reflected light, rather thantransmitted light. Figure 39 is a simplified light raydiagram of a metallurgical microscope. The pre-pared specimen is placed on the stage with thesurface perpendicular to the optical axis of themicroscope and is illuminated through the objec-tive lens by light from a lamp or arc source. Thislight is focused by the condenser lens into a beamthat is made approximately parallel to the opticalaxis of the microscope by the half silvered mirror.The light then passes through the objective ontothe specimen. It is then reflected from the surfaceof the specimen, back through the objective, thehalf silvered mirror, and then to the eyepiece tothe observer’s eye, or to a camera port or a filmplane.

After the specimen has been properly sectionedand polished, it is ready for examination. However,in many cases, it is important to begin the exami-nation with an as-polished specimen,particularly for quality control or failure analysiswork. Certain microstructural constituents, such

as nonmetallic inclusions, graphite, nitrides,cracks or porosity, are best evaluated in theas-polished condition as etching reveals other mi-crostructural features that may obscure suchdetails. In the as-polished condition, there is a mini-mum of extra information for the observer to dealwith which makes examination of these featuresmost efficient. Some materials do not require etch-ing and, in a few cases, examination is lesssatisfactory after etching.

For most metals, etching must be used to fullyreveal the microstructure. A wide range ofprocesses are used, the most common being etch-ing with either acidic or basic solutions. For mostmetals and alloys, there are a number of general-purpose etchants that should be used first toreveal the microstructure. Subsequentexamination may indicate that other morespecialized techniques for revealing the micro-structure may be useful. For example, theexamination may reveal certain constituents thatthe metallographer may wish to measure.Measurements, particularly those performed withautomatic image analyzers, will be simpler to per-form and more reliable if the desired constituentcan be revealed selectively, and numerous pro-cedures are available to produce selective phasecontrasting.

Light Optical Microscopy

Figure 39. Schematic diagram showing the light path through an upright reflected light microscope operatingwith bright field illumination.

7 9

The Light MicroscopeWhen Henry Clifton Sorby made his first exami-nation of the microstructure of iron on 28 July 1863,he was using a transmitted light petrographic mi-croscope made by Smith, Beck and Beck ofLondon that he had purchased in 1861. Althoughcapable of magnifications up to 400X, most of hisexaminations were conducted at 30, 60 or 100Xand his first micrographs were produced at only9X. The objective lenses of this microscope wereequipped with Lieberkühns silvered concave re-flectors for focusing light on opaque specimens.

Sorby quickly realized that reflected lightproduced with the Lieberkühns reflectors wasinadequate and he developed two alternatemethods for this purpose. Subsequently, othersdeveloped vertical illuminators using prisms orplane glass reflectors and Sorby’s systems werenot further developed. In 1886, Sorby reported onthe use of “very high power” (650X) for the studyof pearlite. This was accomplished using a 45ºcover glass vertical illuminator made for him byBeck.

For many years, photomicroscopy was conductedusing specially built reflected microscopes knownas metallographs. These devices represented the“top-of-the-line” in metallurgical microscopes andwere essential for quality work. In the late 1960’sand early 1970’s, manufacturers developedmetallographs that were easier to use from theeyepiece viewing position. The temperamentalcarbon arc light sources were replaced by xenonarc lamps. The unwieldy bellows systems for al-tering magnification were replaced by zoomsystems. Vertical illuminators, previously usingsingle position objectives, were equipped with fourto six position rotatable nosepiece turrets to mini-mize objective handling. The light path wasdeflected so that the film plane was convenientlynear at hand. Universal type vertical illuminatorsand objective lenses were introduced so that theillumination mode could be readily switched frombright field to darkfield, polarized light or differen-tial interference contrast. Such systems were veryattractive in that separate vertical illuminators andobjectives were no longer needed for each illumi-nation mode and handling was eliminated.Exposure meters were also added at this time,chiefly as a result of the rising popularity of instantfilms where such devices are needed to minimize

film wastage. But, by the late 1970s, these largemetallographs had become very expensive, tooexpensive for most laboratories.

In 1979, microscope manufacturers began tointroduce very high quality, reasonably priced,compact metallographs. These microscopescan be obtained with a wide variety of objectivelens types and auxiliary accessories to meet themetallographer’s needs. They are available withuniversal vertical illuminators that permit easyswitching from one illumination mode to anotherusing the same set of objective lenses. Further-more, the manufacturers have introduced newglass compositions and lens formulations,generally by computer-aided design, for improvedimage contrast and brightness. Tungsten- halogenfilament lamps have largely replacedxenon arc lamps as the preferred light source.

Microscope ComponentsLight Sources The amount of light lost duringpassage from the source through a reflecting typemicroscope is appreciable because of the intri-cate path the light follows. For this reason, it isgenerally preferable that the intensity of the sourcebe high, especially for photomicroscopy. Severaltypes of light sources are used including tungsten-fi lament lamps, tungsten-halogen lamps,quartz-halogen lamps, and xenon arc bulbs. Tung-sten-filament lamps generally operate at lowvoltage and high current. They are widely used forvisual examination because of their low cost andease of operation.

Tungsten-halogen lamps are the most popular lightsource today due to their high light intensity. Theyproduce good color micrographs when tungsten-corrected fi lms are employed. Lightintensity can be varied easily to suit the viewingconditions by adjusting a rheostat, a distinctadvantage over arc lamps.

Xenon arc lamps produce extremely high inten-sity light, and their uniform spectra and daylightcolor temperature makes them suitable for colorphotomicrography. The first xenon lamps producedozone, but modern units have overcome this prob-lem. Light output is constant and can only bereduced using neutral density filters.

Light Optical Microscopy

8 0

Condenser. An adjustable lens free of sphericalaberration and coma is placed in front of the lightsource to focus the light at the desired point in theoptical path. A field diaphragm is placed in front ofthis lens to minimize internal glare and reflectionswithin the microscope. The field diaphragm isstopped down to the edge of the field of view.

A second adjustable iris diaphragm, the aperturediaphragm, is placed in the light path before thevertical illuminator. Opening or closing thisdiaphragm alters the amount of light and the angleof the cone of light entering the objective lens. Theoptimum setting for the aperture varies with eachobjective lens and is a compromise among imagecontrast, sharpness, and depth of field. As magni-fication increases, the aperture diaphragm isstopped down. Opening this aperture increasesimage sharpness, but reduces contrast; closingthe aperture increases contrast, but impairs im-age sharpness. The aperture diaphragm shouldnot be used for reducing light intensity. It shouldbe adjusted only for contrast and sharpness.

Filters. Light filters are used to modify the light forease of observation, for improved photo-microscopy, or to alter contrast. Neutral densityfilters are used to reduce the light intensity uni-formly across the visible spectrum. Various neutraldensity filters, with a range of approximately 85 to0.01% transmittance, are available. They are anecessity when using an arc-lamp but virtuallyunnecessary when using a filament lamp.

Selective filters are used to balance the colortemperature of the light source to that of the film.They may be needed for faithful reproduction ofcolor images, depending on the light source usedand the film type. A green or yellow-green filter iswidely used in black and white photography toreduce the effect of lens defects on image quality.Most objectives, particularly the lower cost ach-romats, require such filtering for best results.

Polarizing filters are used to produce planepolarized light (one filter) or crossed-polarized light(two filters rotated 90° to each other to produceextinction) for examinations of anisotropicnoncubic (crystallographic) materials.

Objectives. The objective lens forms the primaryimage of the microstructure and is the mostimportant component of the optical microscope.The objective lens collects as much light aspossible from the specimen and combines this lightto produce the image. The numerical aperture (NA)of the objective, a measure of the light-collectingability of the lens, is defined as:

NA = n sin α

where n is the minimum refraction index of thematerial (air or oil) between the specimen and thelens, and α is the half-angle of the mostoblique light rays that enter the front lens of theobjective. Light-collecting ability increases with α.The setting of the aperture diaphragm willalter the NA of the condenser and therefore theNA of the system.

The most commonly used objective is theachromat, which is corrected spherically for onecolor (usually yellow green) and for longitudinalchromatic aberration for two colors (usually redand green). Therefore, achromats are not suitablefor color photomicroscopy, particularly at highermagnifications. Use of a yellow-greenfilter yields optimum results. However, achromatsdo provide a relatively long working distance, thatis, the distance from the front lens of the objectiveto the specimen surface when in focus. The work-ing distance decreases as the objectivemagnification increases. Most manufacturers makelong-working distance objectives for special ap-plications, for example, in hot stage microscopy.Achromats are usually strain free, which is impor-tant for polarized light examination. Because theycontain fewer lenses than other more highly cor-rected lenses, internal reflection losses areminimized.

Semiapochromatic or fluorite objectives provide ahigher degree of correction of spherical and chro-matic aberration. Therefore, they produce higherquality color images than achromats. The apochro-matic objectives have the highest degree ofcorrection, produce the best results, and are moreexpensive. Plano objectives have extensive cor-rection for flatness of field, which reduceseyestrain, and are usually found on modern mi-croscopes.

Light Optical Microscopy

8 1

With parfocal lens systems, each objective on thenosepiece turret will be nearly in focus when theturret is rotated, preventing the objective front lensfrom striking the specimen when lenses areswitched. Many objectives also are spring loaded,which helps prevent damage to the lens. This ismore of a problem with high magnificationobjectives, because the working distance can bevery small.

Certain objectives are designed for use with oilbetween the specimen and the front lens of theobjective. However, oil-immersion lenses are rarelyused in metallography, because the specimen andlens must be cleaned after use. But, they do pro-vide higher resolution than can be achieved whenair is between the lens and specimen. In the lattercase, the maximum possible NA is 0.95; oil-im-mersion lenses produce a 1.3 to 1.45 NA,depending on the lens and the oil-immersion used.Objective magnifications from about 25 to 200Xare available, depending upon the manufacturer.Use of oil also improves image contrast, which isvaluable when examining low reflectivity speci-mens, such as coal or ceramics.

Eyepieces. The eyepiece, or ocular, magnifies theprimary image produced by the objective; the eyecan then use the full resolution capability of theobjective. The microscope produces a virtual im-age of the specimen at the point of most distinctvision, generally 250 mm (10-inch) from the eye.The eyepiece magnifies this image, permittingachievement of useful magnifications. The stan-dard eyepiece has a 24 mm diameter field of view;wide field eyepieces for plano-type objectives havea 30 mm diameter field of view, which increasesthe usable area of the primary image. Today, thewide field plano-objective is standard on nearly allmetallurgical microscopes.

The simplest eyepiece is the Huygenian, which issatisfactory for use with low and medium powerachromat objectives. Compensating eyepieces areused with high NA achromats and the more highlycorrected objectives. Because some lens correc-tions are performed using these eyepieces, theeyepiece must be matched with the type of objec-tive used. The newer, infinity-correctedmicroscopes do not perform corrections in the eye-pieces, but in the tube lens. Eyepieces,therefore, are simpler in infinity-correctedmicroscopes.

Eye clearance is the distance between the eyelens of the ocular and the eye. For most eyepieces,the eye clearance is 10 mm or less – inadequateif the microscopist wears glasses. Simple visionproblems, such as near sightedness, can be ac-commodated using the fine focus adjustment. Themicroscope cannot correct vision problems suchas astigmatism, and glasses must be worn. Higheyepoint eyepieces are available to provide an eyeclearance of approximately 20 mm, necessary foreyeglasses.

Eyepieces are commonly equipped with variousreticles or graticules for locating, measuring,counting or comparing microstructures. The eye-piece enlarges the reticle or graticule image andthe primary image. Both images must be infocus simultaneously. Special eyepieces are alsoproduced to permit more accurate measure-mentsthan can be made with a graticule scale.Examples are the filar-micrometer ocular or screw-micrometer ocular. Such devices can be automatedto produce a direct digital readout of the measure-ment, which is accurate to approximately 1-μm.

A 10X magnification eyepiece is usually used;to obtain standard magnifications, some systemsrequire other magnifications, such as 6.3X.Higher power eyepieces, such as 12, 15, 20, or25X, are also useful in certain situations. The over-all magnification is found by multiplying theobjective magnification, Mo, by the eyepiecemagnification, Me. If a zoom system or bellowsis also used, the magnification should be alteredaccordingly.

Stage. A mechanical stage is provided for focus-ing and moving the specimen, which is placed onthe stage and secured using clips. The stage ofan inver ted microscope has replaceablecenter stage plates with different size holes. Thepolished surface is placed over the hole forviewing. However, the entire surface cannot beviewed, unless the specimen is smaller than thehole and it is mounted. At high magnificationsit may not be possible to focus the objectivenear the edge of the hole due to the restrictedworking distance.

Using the upright microscope, the specimen isplaced on a slide on the stage. Because the

Light Optical Microscopy

8 2

polished surface must be perpendicular to the lightbeam, clay is placed between the specimen bot-tom and the slide. A piece of lens tissue is placedover the polished surface, and the specimen ispressed into the clay using a leveling press. How-ever, pieces of tissue may adhere to the specimensurface. An alternative, particularly useful withmounted specimens, is to use a ring instead oftissue to flatten the specimen. Aluminum or stain-less steel ring forms of the same size as themounts (flattened slightly in a vise) will seat onthe mount rather than the specimen.

The upright microscope allows viewing of the en-tire surface with any objective, and theoperator can see which section of the specimenis being viewed – a useful feature when examin-ing specific areas on coated specimens, welds,and other specimens where specific areas are tobe examined. For mounted specimens, anautoleveling stage holder can eliminate levelingspecimens with clay.

The stage must be rigid to eliminate vibrations.Stage movement, controlled by x and y microme-ters, must be smooth and precise; rack and piniongearing is normally used. Many stages havescales for measuring distances in the x and y di-rections. The focusing controls often containrulings for estimating vertical movement. Someunits have motorized stages and focus controls.

A circular, rotatable stage plate may facilitate po-larized light examination. Such stages, commonfor mineralogical or petrographic studies, aregraduated to permit measuring the angle of rota-tion. A rectilinear stage is generally placed on topof the circular stage.

Stand. Bench microscopes require a rigid stand,particularly if photomicroscopy is performed onthe unit. The various pieces of the microscope areattached to the stand when assembled.In some cases, the bench microscope is placedon a separate stand that also holds the photo-graphic system.

ResolutionTo see microstructural detail, the optical systemmust produce adequate resolution, or resolvingpower, and adequate image contrast. If resolutionis acceptable but contrast is lacking, detail can-not be observed. In general, the ability toresolve two points or lines separated by distanced is a function of the wavelength, λ, of theincident light and the numerical aperture, NA, ofthat objective.

where k is 0.5 or 0.61. Figure 40 illustrates thisrelationship for k = 0.61 and four light wavelengths.Other formulas have also been reported. Thisequation does not include other factors that influ-ence resolution, such as the degree ofcorrection of the objectives and the visual acuityof the microscopist. It was based on the work ofAbbe under conditions not present in metallogra-phy, such as self-luminous points, perfectblack-white contrast, transmitted light examination,an ideal point-light source, and absence of lensdefects.

Using the above equation, the limit of resolutionfor an objective with an NA of 1.3 is approximately0.2-μm. To see lines or points spaced 0.2-μm apart,the required magnification is determined by divid-ing by the resolving power of the human eye, whichis difficult to determine under observation condi-tions. Abbe used a value of 0.3 mm at a distanceof 250 mm – the distance from the eye for opti-mum vision. This gives 1500x. For light with a meanwavelength of 0.55-μm, the required magnificationis 1100 times the NA of the objective. This is theorigin of the 1000·NA rule for the maximum usefulmagnification. Any magnification above 1000·NAis termed “empty”, or useless.

Strict adherence to the 1000·NA rule should bequestioned, considering the conditions underwhich it was developed, certainly far different fromthose encountered in metallography. According tothe Abbe analysis, for a microscopist with opti-mum 20/20 vision and for optimum contrastconditions and a mean light wavelength of 550 nm,the lowest magnification that takes full advantageof the NA of the objective is 500 times the NA.

kλ

NAd =

Light Optical Microscopy

8 3

of field increases as the NA decreases and whenlonger wavelength light is used, as shown in Fig-ure 41.

Imaging ModesMost microscopical studies of metals are madeusing bright field illumination. In addition to this typeof illumination, several special techniques (obliqueillumination, dark field i l lumination,differential interference contrast microscopy andpolarized light microscopy) have particularapplications for metallographic studies.

Nearly all microscopes using reflected or trans-mitted light employ Köhler illumination because itprovides the most intense, most even illuminationpossible with standard light sources. The reflectedlight microscope has two adjustablediaphragms: the aperture diaphragm and the fielddiaphragm, located between the lamp housing andthe objective. Both are adjusted to improve illumi-nation and the image. To obtain Köhler illumination,the image of the field diaphragm must be broughtinto focus on the specimen plane.This normally occurs automatically when the

This establishes a useful minimum magnificationto use with a given objective. It has been sug-gested that the upper limit of useful magnificationfor the average microscopist is 2200·NA, not1000·NA.

Depth of FieldDepth of field is the distance along the optical axisover which image details are observed with ac-ceptable clarity. Those factors that influenceresolution also affect depth of field, but in the op-posite direction. Therefore, a compromise must bereached between these two parameters, which ismore difficult as magnification increases. This isone reason light etching is preferred for high-mag-nification examination. The depth of field, df, canbe estimated from:

where n is the refractive index of the medium be-tween the specimen and the objective (n ~ 1.0 forair), λ is the wavelength of light, and NA is thenumerical aperture. This equation shows that depth

Figure 40. Influence of objective numerical aperture and light wavelength on the resolution of the lightmicroscope.

λ (n2 −NA2)½

NA2df =

Light Optical Microscopy

RESOLUTION CHART

Sodium (589 nm)

Green (546 nm)

Blue (436 nm)

Ultraviolet(365 nm)

8 4

microstructural image is brought into focus. Thefilament image must also be focused onthe aperture diaphragm plane. This producesuniform illumination of the specimen imaged at theintermediate image plane and magnified by the eye-piece.

Bright Field. In bright field illumination, the sur-face of the specimen is normal to the optical axisof the microscope, and white light is used. A raydiagram for bright-field illumination is illustrated inFigure 42. Light that passes through theobjective and strikes a region of the specimen sur-face that is perpendicular to the beamwill be reflected back up the objective through theeyepieces to the eyes where it will appear to bebright or white. Light that strikes grainboundaries, phase boundaries, and otherfeatures not perpendicular to the optical axis willbe scattered at an angle and will not be collectedby the objective. These regions will appear tobe dark or black in the image. Bright field isthe most common mode of illumination usedby metallographers.

Oblique Illumination. The surface relief of ametallographic specimen can be revealed usingoblique illumination. This involves offsetting thecondenser lens system or, as is more usuallydone, moving the condenser aperture to a posi-tion slightly off the optical axis. Although it shouldbe possible to continually increase the contrastachieved by oblique illumination by moving thecondenser farther and farther from the optical axis,the numerical aperture of a lens is reduced whenthis happens because only a

Figure 41. Influence of the objective numerical aperture and light wavelength on the depth of field of the lightmicroscope.

Sodium (589 nm)

Green (546 nm)

Blue (436 nm)

Index of Refraction = 1.0

DEPTH OF FIELD CHART

Light Optical Microscopy

Figure 42. Schematic diagram showing the lightpath in bright field illumination.

8 5

produces plane polarized light that strikes the sur-face and is reflected through the analyzer to theeyepieces. If an anisotropic metal is examined withthe analyzer set 90° to the polarizer, the grainstructure will be visible. However, viewing of anisotropic metal (cubic metals) under such condi-tions will produce a dark, “extinguished” condition(complete darkness is not possible using Polaroidfilters). Polarized light is particularly useful in met-allography for revealing grain structure andtwinning in anisotropic metals and alloys (see theappendix for crystal structure information) and foridentifying anisotropic phases and inclusions.

Differential Interference Contrast (DIC). Whencrossed polarized light is used along with a doublequartz prism (Wollaston prism) placed betweenthe objective and the vertical illuminator, Figure45, two light beams are produced which exhibitcoherent interference in the image plane. This leadsto two slightly displaced (laterally) images differ-ing in phase (λ/2) that produces height contrast.The image produced reveals topographic detailsomewhat similar to that produced by oblique illu-mination but without the loss of resolution. Imagescan be viewed with natural colors similar to thoseobserved in bright field, or artificial coloring canbe introduced by adding a sensitive tint plate.

As an example of the use of these differentimaging modes, Figure 46 shows the micro-

portion of the lens is used. For this reason, thereis a practical limit to the amount of contrast thatcan be achieved. Illumination also becomes un-even as the degree of “obliqueness” increases.Since differential interference contrast systemshave been available, oblique illumination is rarelyoffered as an option on new microscopes.

Dark Field. Another method that can be used todistinguish features not in the plane of thepolished and etched surface of a metallographicspecimen is dark field (also called dark ground)illumination. This type of illumination (seeFigure 43 for a ray diagram) gives contrastcompletely reversed from that obtained with bright-field illumination — the features that are light inbright field will be dark in dark field, and those thatare dark in bright field will be light, appearing to beself luminous in dark field. This highlighting ofangled surfaces (pits, cracks, or etched grainboundaries) allows more positive identification oftheir nature than can be derived from a black im-age with bright field illumination. Due to the highimage contrast obtained and the brightness asso-ciated with features at an angle to the optical axis,it is often possible to see details not observed withbright field illumination.

Polarized Light. Because many metals and me-tallic and nonmetallic phases are opticallyanisotropic, polarized light is particularly useful inmetallography. Polarized light is obtained by plac-ing a polarizer (usually a Polaroid filter) in front ofthe condenser lens of the microscope and plac-ing an analyzer (another Polaroid filter) before theeyepiece, as illustrated in Figure 44. The polarizer

Light Optical Microscopy

Figure 44. Schematic diagram showing the lightpath in polarized light with an optional lambda plate(sensitive tint plate).

Figure 43. Schematic diagram showing the lightpath in dark field illumination.

8 6

structure of an aluminum bronze specimen thatwas water quenched from the beta field formingmartensite. The specimen was prepared but notetched. A minor amount of relief was introduced infinal polishing. Figure 46a shows the surface inbright field illumination. Due to the relief present, afaint image of the structure is visible. Dark field,Figure 46c, and differential interference contrast,

Light Optical Microscopy

Figure 46d, make use of this relief to reveal muchmore detail than in bright field. However, as themartensite is noncubic in crystal structure, it re-sponds well to crossed- polarized light, Figure 46b.Of course, not every material can be examinedusing all four illumination modes, but there aremany cases where two or more modes can beeffectively utilized to reveal more information thangiven by bright field alone.

Figure 46. Martensitic microstructure of heattreated, eutectoid aluminum bronze (Cu–11.8% Al) inthe unetched condition viewed using: a) bright fieldillumination, b) polarized light

Figure 46. Martensitic microstructure of heattreated, eutectoid aluminum bronze (Cu–11.8% Al)in the unetched condition viewed using: c) dark fieldillumination; and, d) differential interferencecontrast illumination (200X).

a

b

c

d

HELPFUL HINTS FOR

LIGHT OPTICAL

MICROSCOPY

It can be difficult to deter-mine which objective you are using with aninverted microscope due to the limited viewfrom above, as the magnification value maynot be easily observed. However, each ob-jective has a color-coded ring on it that isusually visible. The colors and correspond-ing magnifications are: red – 5X, yellow – 10X,green – 20X, blue – 50X, and white – 100X.

Figure 45. Schematic diagram showing the lightpath for differential interference contrast (DIC)illumination.

8 7

MICROINDENTATIONHARDNESS TESTING

Microindentation hardness testing, more com-monly (but incorrectly) called microhardnesstesting, is widely used to study fine scale changesin hardness, either intentional or accidental. Heattreaters have utilized the technique for many yearsto evaluate the success of surface hardening treat-ments or to detect and assess decarburization.Metallographers and failure analysts use themethod for a host of purposes including evalua-tion of homogeneity, characterization ofweldments, as an aid to phase identification, orsimply to determine the hardness of specimenstoo small for traditional bulk indentation tests.

Metallurgists and metallographers tend todevelop their own jargon, often as a matter of lin-guistic simplicity, which is not always as rigorouslycorrect as it could be. Although the term“microhardness” is generally understood by itsusers, the word implies that the hardness isextremely low, which is not the case. The appliedload and the resulting indent size are small rela-tive to bulk tests, but the same hardnessnumber is obtained. Consequently, ASTM Com-mittee E-4 on Metallography recommends use ofthe term “microindentation hardness testing” whichcould be given the acronym MHT. ASTM StandardE 384 fully describes the two most commonmicroindentation tests - the Vickers and the Knooptests.

The Vickers TestIn 1925, Smith and Sandland of the UK developeda new indentation test for metals that were too hardto evaluate using the Brinell test. The hardenedsteel ball of the Brinell test l imitedthe test to steels with hardnesses below ~450 HBS(~48 HRC). (The harder tungsten carbide ball wasnot available in 1925. The WC indenterextends the Brinell test to metals up to 615 HBW(~58 HRC). The WC ball has now replacedthe steel ball for the Brinell test.) In designing thenew indenter, a square-based diamondpyramid (see Figure 47), they chose a geometrythat would produce hardness numbers nearly iden-tical to Brinell numbers in the range where both

tests could be used. This was a very wisedecision as it made the Vickers test very easyto adopt. The ideal d/D ratio (d = impressiondiameter, D = ball diameter) for a sphericalindenter is 0.375. If tangents are drawn to the ballat the impression edges for d/D = 0.375,they meet below the center of the impressionat an angle of 136°, the angle chosen for theVickers indenter.

Use of diamond allowed the Vickers test to be usedto evaluate any material (except diamond) and,furthermore, had the very important advantage ofplacing the hardness of all materials on one con-tinuous scale. This is a major disadvantage ofRockwell type tests where different scales (15standard and 15 superficial) were developed toevaluate materials. Not one of these scales cancover the full hardness range. The HRA scale cov-ers the broadest hardness range, but it is notcommonly used.

In the Vickers test, the load is applied smoothly,without impact, forcing the indenter into the testpiece. The indenter is held in place for 10 or 15seconds. The physical quality of the indenter andthe accuracy of the applied load (defined in E 384)must be controlled in order to get thecorrect results. After the load is removed, the twoimpression diagonals are measured, usually to thenearest 0.1-μm with a filar micrometer, and aver-aged. The Vickers hardness (HV) is calculatedusing:

Microindentation Hardness Testing

Figure 47. Schematic of the Vickers indenter andthe shape of an impression.

136°

136°

1854.4L

d2HV =

8 8

where the load L is in gf and the average diagonald is in μm (this produces hardness number unitsof gf/μm2 although the equivalent units kgf/mm2 arepreferred; in practice the numbers are reportedwithout indication of the units). The original Vickerstesters were developed for test loads of 1 to120kgf that produce rather large indents. Recog-nizing the need for lower test loads, the NationalPhysical Laboratory (UK) reported on use of lowertest loads in 1932. Lips and Sack developed thefirst low-load Vickers tester in 1936.

Because the shape of the Vickers indentation isgeometrically similar at all test loads, the HV valueis constant, within statistical precision, over a verywide test load range as long as the test specimenis reasonably homogeneous. Numerous studiesof microindentation hardness test results con-ducted over a wide range of test loads have shownthat test results are not constant at very low loads.This problem, called the “indentation size effect”,or ISE, has been attributed to fundamental char-acteristics of the material. However, the sameeffect is observed at the low load test range (1-10kgf) of bulk Vickers testers [2] and an ASTMinterlaboratory “round robin” of indents made byone laboratory but measured by twelve differentpeople, reported all three possible ISE responses[24,25] for the same indents!

Since the 1960s, the standard symbol for Vickershardness per ASTM E 92 and E 384, has beenHV. This should be used in preference to the older,obsolete symbols DPN or VPN. The hardness isexpressed in a standard format. For example, if a300 gf load is used and the test reveals a hard-ness of 375 HV, the hardness is expressed as375 HV300. Rigorous application of the SI systemresults in hardness units expressed not in thestandard, understandable kgf/mm2 values but inGPa units that are meaningless to most engineersand technicians. ASTM recommends a ‘soft” met-ric approach in this case.

In the Vickers test, it is assumed that elasticrecovery does not occur once the load isremoved. However, elastic recovery does occur,and sometimes its influence is quite pronounced.Generally, the impression (Figure 48) appears tobe square, and the two diagonals have similarlengths. As with the Brinell test, the Vickers hard-ness number is calculated based on the surface

area of the indent rather than the projected area. Ifthe impression shape is distorted due to elasticrecovery (very common in anisotropic materials),Figure 49, should the hardness be based on theaverage of the two diagonals? It is possible to cal-culate the Vickers hardness based on theprojected area of the impression, which can bemeasured by image analysis. While rigorous stud-ies of this problem are scant in the literature, thediagonal measurement is the preferred approacheven for distorted indents, at this time.

The Knoop TestAs an alternative to the Vickers test, particularlyto test very thin layers, Frederick Knoop and hisassociates at the former National Bureau ofStandards developed a low-load test using a rhom-bohedral-shaped diamond indenter, Figure 50. The

long diagonal is seven times (7.114 actually) aslong as the short diagonal. With this indentershape, elastic recovery can be held to a minimum.Some investigators claim there is no elastic re-covery with the Knoop indent, but this cannot betrue as measurements of the ratio of long to shortdiagonal often reveal results substantially differ-ent than the ideal 7.114 value.

Microindentation Hardness Testing

Figure 48. (top) Example of a well-formed Vickersindentation (400X). Figure 49. (bottom) Example of adistorted Vickers indentation (400X).

8 9

The Knoop test is conducted in the same manner,and with the same tester, as the Vickers test. How-ever, only the long diagonal is measured. This, ofcourse, saves some time. The Knoop hardness iscalculated from

HK = 14229L d2

where the load L is in gf and the long diagonal d isin μm. Again, the symbol HK was adopted in theearly 1960’s while other terms; e.g., HKN or KHN,are obsolete and should not be used. The Knoophardness is expressed in the same manner asthe Vickers hardness; i.e., 375 HK300 means that a300 gf load produced a Knoop hardness of 375.(The kgf/mm2 unit information is no longer re-ported).

Aside from a minor savings of time, one chief meritof the Knoop test is the ability to test thin layersmore easily. For surfaces with varying hardness,such as case hardened parts, Knoop indents canbe spaced closer together than Vickers indents.Thus, a single Knoop traverse can define a hard-ness gradient more simply than a series of two orthree parallel Vickers traverses where each indentis made at different depths. Furthermore, if thehardness varies strongly with the depth, theVickers indent will be distorted by this change;that is, the diagonal parallel to the hardness changewill be affected by the hardness gradient (i.e., thereis a substantial difference in the lengths of the two

halves of the diagonal), while the diagonal perpen-dicular to the hardness gradient will be unaffected(both halves of this diagonal of the same approxi-mate length).

The down side of the Knoop indent is that the threedimensional indent shape will change with test loadand, consequently, HK varies with load. At highloads, this variation is not substantial. Conversionof HK values to other test scales can only be donereliably for HK values performed at the standardload, generally 500gf, used to develop the corre-lations. All hardness scale conversions are basedon empirical data. Conversions are not precise butare estimates.

Factors Affecting Accuracy, Precisionand BiasMany factors (see Table 44) can influence the qual-ity of microindentation test results [26]. In the earlydays of low-load (< 100gf) hardness testing, it wasquickly recognized that improper specimen prepa-ration may influence hardness test results. Mosttext books state that improper preparation will yieldhigher test results because the surface containsexcessive preparation induced deformation. Whilethis is certainly true, there are other cases whereimproper preparation can create excessive heatthat will lower the hardness and strength of manymetals and alloys. So either problem may be en-countered due to faulty preparation. For manyyears, it was considered necessary to electrolyti-cally polish specimens so that the pre-paration-induced damage could be removed thuspermitting bias-free low-load testing. However, thescience behind mechanical specimen preparation,chiefly due to the work of Len Samuels [3], hasled to development of excellent mechanical speci-men preparation procedures, and electropolishingis no longer required.

There are several operational factors that mustbe controlled in order to obtain optimum testresults. First, it is a good practice to inspect theindenter periodically for damage, for example,cracking or chipping of the diamond. If you havemetrology equipment, you can measure the faceangles and the sharpness of the tip. Specificationsfor Vickers and Knoop indenter geometries aregiven in E 384.

Microindentation Hardness Testing

Figure 50. Schematic of the Knoop indenter and theshape of an impression

130°

172°30´

9 0

A prime source of error in the tests is the align-ment of the specimen surface relative to theindenter. The indenter itself must be properlyaligned perpendicular (±1°) to the stage plate.Next, the specimen surface must be perpendicu-lar to the indenter. Most testers provide holdersthat align the polished face perpendicular to theindenter (parallel to the stage). If a specimen issimply placed on the stage surface, its backsurface must be parallel to its polished surface.Tilting the surface more than 1° from perpendicu-lar results in nonsymmetrical impressions and canproduce lateral movement between specimen andindenter. The occurrence of non- symmetrical in-dents is generally easily detected duringmeasurement.

In most cases, errors in indenting with moderntesters are not the major source of error, althoughthis can occur [26]. It is important to check theperformance of your tester regularly using acertified test block. It is safest to use a test blockmanufactured for microindentation testing andcertified for the test (Vickers or Knoop) and theload that you intend to use. Strictly speaking, ablock certified for Vickers testing at 300 or500 gf (commonly chosen loads) should yield es-sentially the same hardness with loads from about50 to 1000 gf. That is, if you take the average ofabout five indents and compare the average atyour load to the average at the calibrated load(knowing the standard deviation of the test results),statistical tests can tell you (at any desired confi-dence level) if the difference between the meanvalues of the tests at the two loads is statisticallysignificant or not.

Because of the method of defining HV and HK(equations given above) where we divide by d2,measurement errors become more critical as dget smaller; that is, as L decreases and thematerial’s hardness increases (discussed later).So departure from a constant hardness for theVickers or Knoop tests as a function of load willbe a greater problem as the hardness increases.For the Knoop test, HK increases as L decreasesbecause the indent geometry changes with indentdepth and width. The degree of the change in HKalso varies with test load being greater as L de-creases.

The greatest source of error is in measuring theindent as has been documented in an ASTMinterlaboratory test [24,25]. Place the indent in thecenter of the measuring field, if it is not alreadythere, as lens image quality is best in the center.The light source should provide adequate, evenillumination to provide maximum contrast and reso-lution. The accuracy of the filar micrometer, or othermeasuring device, should be verified using a stagemicrometer.

Specimen preparation quality becomes more im-portant as the load decreases, and it must be atan acceptable level. Specimen thickness must beat least 2.5 times the Vickers diagonal length. Be-cause the Knoop indent is shallower than theVickers at the same load, somewhat thinner speci-mens can be tested. Spacing of indents isimportant because indenting produces plasticdeformation and a strain field around the indent. Ifthe spacing is too small, the new indent will be

Table 44. Factors Affecting Precision and Bias in Microindentation Hardness Testing

Instrument Factors Measurement Factors Material Factors

Accuracy of the applied Calibration of the measurement Heterogenity of theload system specimen

Inertia effects, Numerical aperture Strength of crystallographicspeed of loading of the objective texture, if present

Lateral movement of the Magnification Quality of specimenindenter or specimen preparation

Indentation time Inadequate image quality Low reflectivity or transparency

Indenter shape deviations Uniformity of illumination Creep during indentation

Damage to the indenter Distortion in optics Fracture during indentation

Inadequate spacing between Operator’s visual acuity Oil, grease or dirt onindents or from edges indenter or specimen

Angle of indentation Focusing of the image

Microindentation Hardness Testing

9 1

affected by the strain field around the last indent.ASTM recommends a minimum spacing (centerto edge of adjacent indent) of 2.5 times the Vickersdiagonal. For the Knoop test where the long di-agonals are parallel, the spacing is 2.5 times theshort diagonal. The minimum recommended spac-ing between the edge of the specimen and thecenter of the indent should be 2.5 times. Again,Knoop indents can be placed closer to the sur-face than Vickers indents.

Several studies have stated that the resolution ofthe optical system is the major factor limiting mea-surement precision. They state that this causesindents to be undersized by a constant amount.However, as undersizing would increase the mea-sured hardness, not decrease it,resolution limits per se cannot explain thisproblem. For the Knoop indent, the chiefproblem is image contrast at the indent tips thatresults in undersized indents. This problem, plusthe variable indent shape, both result inincreasing HK with decreasing test load. For theVickers test, one would assume that undersizingor oversizing would be equally likely; but, experi-ence suggests that oversizing is much morecommonly encountered for low loads and smallindents.

AutomationMicroindentation hardness testing is tedious, soanything that can be done to simplify testing isvaluable, especially for laboratories that do sub-stantial testing. Many adjuncts to indentmeasurement have been tried and a variety ofsuch systems are available. There has been con-siderable interest in applying image analyzers tothe indent measurement task. Further, with stageautomation, it is possible to automate the indent-ing process itself, and with the same equipment.Figure 51 shows the Buehler OmniMet MHT auto-mated microindentation hardness testing system.This system can be used in the fully automatic,semiautomatic or manual mode, depending uponthe nature of the testing.

An automated system can be programmed tomake any number of indents in either a definedpattern (x number of equally spaced indents, or xnumber between two chosen points) or arandom pattern (locations selected at random witha mouse). Curved patterns may be used, not just

straight-line patterns. The system then makes theindents at the requested load and locations, mea-sures each indent, calculates the hardness (anddesired conversions to other scales), and prints/plots the results. Statistical analysis can also bepreformed.

In general, Vickers indents are easier to measureby image analysis than Knoop indents due to thelower contrast at the tips of the Knoopindents (leads to undersizing and higher HKvalues). Metal flow (plastic deformation) aroundthe indent edges can interfere with measurementprecision. Vickers indents, like the ones shown inFigure 48, exhibit excellent contrast and shapeand are easily measured. If the magnification istoo high (and this may be influenced by thenumerical aperture of the objective) for a givenindent size, image contrast suffers and correctdetection of the indent will be very difficult. On theother hand, if the indent is very small on the screen,it will be hard for the system to detect it automati-cally. In this case, use a higher magnification, oruse the semiautomatic measurement mode if thisis not possible. The operator fits a box around theindent with a mouse to measure the indent.

Microindentation hardness testing is a veryvaluable tool for the materials engineer but itmust be used with care and a full understandingof the potential problems that can occur. Alwaystry to use the highest possible load depending uponthe indent spacing or closeness to edges. Try tokeep indents greater than 20-μm in diameter, asrecommended in ASTM standard E 384. If youhave not read E 384, or it has been some time

Microindentation Hardness Testing

Figure 51. OmniMet MHT

9 2

since you read it (all ASTM standards must bereviewed and revised, if necessary, every fiveyears), get the latest copy and go over it. The Pre-cision and Bias section (and appendix X1 of E 384)contains a wealth of practical advice. Automationof indentation and measurement is possible and,for laboratories that do substantial testing, greatlyreduces operator fatigue while reducing testingtimes.

HELPFUL HINTS FOR

MICROINDENTATION

HARDNESS TESTING

If you suspect that the testresults obtained on a specimen are ques-tionable, verify the tester using a certifiedtest block. When testing thin coatings, usethe Knoop indenter with the long axis paral-lel to the surface. If you have a specimenwhere the hardness changes rapidly,Vickers indents may be distorted badly inthe direction of the hardness gradient.Switch to a Knoop indenter and place thelong axis perpendicular to the hardness gra-dient.

Microindentation Hardness Testing

9 3

Image Capture and Analysis

IMAGE CAPTURE ANDANALYSIS

The progress in computer and video technologyhas created a movement toward electronic imageacquisition. These images can be used in softwareapplications such as word processing or desktoppublishing programs allowing for fast report gen-eration and electronic distribution. In addition,associated data can be stored with the images ina database allowing for easy retrieval of informa-tion through searches.

When adding imaging capabilities to the labora-tory, it is important to consider your goals.Oftentimes, images are just another step in thedocumentation process. For example, in a failureinvestigation it is useful to capture the image of acomplete component before the sectioning pro-cess. Or, an image of the microstructure might beattached to a report as an indication of a pass/failcondition. Additional functionality of an imagingsystem might include a scale marker overlay andpoint-to-point or other operator interactive mea-surements. Fully automated image analysisincludes detection of the features of interest basedon grey level or color differences as well as mor-phological characteristics such as size and shape.Automated imaging applications are based on afundamental series of steps shown in Table 45.Depending on your goals, some or all of thesesteps may be utilized.

Aligned with these imaging goals, Buehler offersa series of upgradeable OmniMet imaging prod-ucts (Figure 52). The OmniMet Image Capture andReport System focuses on the basics of imagecapture and generating simple image reports. TheOmniMet Archive imaging product has additionalcapturing features that enable image comparison

and multi-layer focusing as well as a powerful da-tabase for tracking projects and measurements.The OmniMet Express and OmniMet Enterpriseproducts include automated measurements in ad-dition to the operator interactive measurementsfound in the other products. OmniMet Express hasspecific application modules while the OmniMetEnterprise product enables the user to tackle anyapplication.

AcquisitionImage capture is a term used to describe imageacquisition by means of a camera and frame grab-ber or a digital camera. Because of the manychoices of camera types, a video microscopy sys-tem must be flexible. Analog CCD cameras, blackand white or color, are most frequently used. Com-ponent video (Y/C or S-Video) and compositevideo signals and a number of color video stan-dards such as NTSC, PAL and SECAM aregenerally supported. Images acquired in the ma-terials laboratory are optimized in real time byadjusting brightness, contrast, and color satura-tion. The analog output camera signal is thendigitized utilizing an analog input frame grabberboard.

Table 45. Primary Imaging Steps

Imaging Steps Description

Acquisition Capture, load, or import an image

Clarification Develop the necessary contrast and clarity for detecting the features of interest

Thresholding Detect the features of interest

Binary Operations Clean-up any detection discrepancies, categorize features, and overlay grids

Measurements Conduct field or feature measurements

Data Analysis Evaluate statistics and relevance of the measurements

Archive Store images, annotations, and associated measurements in a database

Distribution Printing or electronic distribution of images and results

Figure 52. OmniMet Digital Imaging System

9 4

Image Capture and Analysis

Alternatively, digital cameras may be integratedwith a SCSI firewire or USB connection. Since thelate 1990’s digital video cameras have become in-creasingly popular for scientif ic imagingapplications. This is largely accounted for by twofactors: 1) higher pixel densities in the acquiredimages potentially provide higher spatial resolu-tion and 2) the cost range of digital cameras hasmerged with the combined cost of an analog cam-era and specialized capture board. Capturing animage with a digital camera is often a two-stageprocess, first a preview image is viewed for fo-cusing and selecting the field of interest and thena snapshot of the image is taken. The previewimage window displays the image at a lower pixeldensity and the refresh rate of the camera willdetermine the amount of lag time between the op-erator adjusting the microscope and the new imagebeing displayed. In order to be comparable to ananalog camera, the refresh rate should approach25 frames per second.

The resulting image files are based on two pri-mary formats, bit-map and vector. The majority ofscientific imaging programs are based on bit-mapped images. Bit-mapped graphics areessentially grids, consisting of rows and columnsof pixels. When the image is zoomed to observedetails, the individual pixels will become visible.Pixels, or picture elements, are the smallest partof an image that a computer, printer or monitor dis-play can control. An image on a monitor consistsof hundreds of thousands of pixels, arranged insuch a manner that they appear to be connected.Figure 53 displays a cast iron image with a smallsegment zoomed such that the pixels are ob-served.

Each pixel in a bitmap is stored in one or moredata bits. If an image is monochromatic (black andwhite), then one bit is enough to store informationfor one pixel. If an image is colored or uses vari-ous shades of grey, then additional bits are requiredto store color and shading information. Table 46lists some of the more common bit depths referredto in imaging software and camera literature.

When working in grayscale, 256 values are usedto represent each shade of grey, starting withblack at 0 and transitioning to white at 255. Forexample, the information stored for the image inFigure 53 would be a combination of the x and ypositions and a number from 0 to 255. For colorimages there are two common models used to nu-merically represent color, RGB and HLS. Eachrequires storing three values, in addition to the xand y position of the pixel, in order to recreate theimage. In general, you will notice the file sizes forcolor images are larger than grey scale images ofthe same pixel density.

The RGB model is based on three primary colors– red, green and blue – that can be combined invarious proportions to produce different colors.This scheme is considered additive because,when combined in equal portions, the result iswhite. Likewise, an absence of all three is black.The three primary colors are each measured as avalue from 0-255. The colors produced by com-bining the three primaries are a result of the relativevalues of each primary. For example, pure red hasa red value of 255, a green value of 0, and a bluevalue of 0. Yellow has a red value of 255, a greenvalue of 255, and a blue value of 0. This system ismodeled as a cube with three independent direc-tions: red, green, and blue, spanning the availablespace. The body diagonal of the cube contains lu-minance, in other words, colorless informationabout how light or dark the signal is.

The HLS model uses the concept of Hue, Lumi-nance, and Saturation. Hue is the color tone. For

Table 46. Bit Depth

Bits

1 21 or 2 tones, black and white

8 28 or 256 grey scale shades

24 224 or 16.7 million colors

32 232 or 4.29 billion colors

Figure 53. A small segment of an image zoomed todisplay the individual pixels..

9 5

example, an etchant might highlight the differentphases as unique shades of brown or blue. Satu-ration is complementary to the hue. It describeshow brilliant or pure – as opposed to washed outor grayish a particular color is. Luminance is cor-related with the light intensity and viewed ascomparable to the greyscale values where zerois black and 255 is white. The HLS color spacecan be viewed as a double cone, in which the z-axis of the cone is the grayscale progression fromblack to white, distance from the central axis isthe saturation, and the direction or angle is thehue.

The illustration in Figure 54 represents pixels ofnine different color values. These values are shownin Table 47 for the greyscale, RGB, and HLS mod-els. When the pixels are not true grayscale tones,the greyscale number is not given.

Resolution is primarily the imaging system’s abil-ity to reproduce object detail by resolving closelyspaced features. In digital microscopy, the reso-lution is examined in terms of both the microscopelimitations and the pixel array of the camera. Theclarity and definition of a digital image dependslargely on the total number of pixels used to cre-ate it. Typical pixel array densities range from 640x 480 to 3840 x 3072. Multiplying the two numbersgives you the total number of pixels used to cre-ate the image. Because a large number of pixelsleads to sharper images, you might conclude thatit is always desirable to capture images with thelargest pixel array possible. That is not necessar-ily always true. The problem with large array imagesis that they consume a lot of storage space. As aresult, these images are not recommended forposting on web sites or email use. It is important

Image Capture and Analysis

Table 47. Values based on color models

HLS Method RGB MethodPixels Greyscale Hue Luminance Saturation Red Green Blue

1 192 0 192 0 192 192 192

2 150 0 150 0 150 150 150

3 128 0 128 0 128 128 128

4 0 127 255 255 0 0

5 85 127 255 0 255 0

6 170 127 255 0 0 255

7 60 98 100 255 255 0

8 310 36 79 148 31 129

9 25 100 176 169 113 31

to take into account what you intend to do withyour images when selecting the appropriate cam-era.

In order to perform measurements, the image orthe image source must first be calibrated. Calibra-tion is achieved by assigning a known distance toa pixel count. When using a light microscope thisis accomplished with a stage micrometer. Each ofthe objectives will have its own unique calibrationfactor. If the aspect ratio of the camera pixels isunknown, it can be determined by calibrating inboth the x and y direction. If the aspect ratio isknown, then calibration only needs to be completedin the x direction. This calibration technique is re-peated for each objective or working

Figure 54. A representation of nine pixels withdifferent color values.

9 6

can be as simple as a label that is referenced inthe image caption or a complete text descriptionof a feature within the image. A recommendedannotation is a scale marker in the lower corner ofthe image. This is particularly important for dis-tributing the image electronically because the finalsize of the printout is at the discretion of the re-cipient.

Automated MeasurementsWhen more detailed measurements are requiredor a large quantity of measurements is important,it is useful to automate the process. For example,consider the coating that was measured in Figure56. It is possible to duplicate that measurementas well as to evaluate any variation across thecoating and determine the percentage porositywithin the coating. A typical process for determin-ing coating thickness is outlined in Figure 57 andis described below.

ThresholdingThresholding is the method for representing rangesof pixel grey or color values with different colorbitplane overlays. The bitplanes are a binary layerplaced over the image plane superimposing thephase(s) of interest. Most of the measurementsare accomplished on this layer not on the actualimage.

The first step of the thresholding process is thedevelopment of a threshold histogram. Each pixelis assigned a value and then the frequency isgraphed. When working in grayscale with an 8- bitimage, 256 values are used to represent eachshade of grey, starting with black at 0 andtransitioning to white at 255. When working with a

magnification. It is also possible to calibrate im-ages that have been imported from another source.The only requirement is to have a feature withinthe image that is of a known dimension. A scalemarker is usually the best choice.

ClarificationImage clarification is also referred to as imageenhancement and is largely achieved through theuse of greyscale filters. Filters perform one of sev-eral standard functions: edge detection,photographic enhancement, or grey level modifi-cation. Image clarifications, which modify thevalues of the pixels across an entire image, areconducted in one of two ways: adjusting limits suchas contrast, brightness and gamma or comparingthe values of neighboring pixels. The latter is re-ferred to as a convolution or kernel operation. Theneighborhood sizes are typically squares of pix-els from 3x3, 5x5, 7x7, etc. A typical need is toincrease the local contrast at the phase bound-aries. Often times, the use of a neighborhoodtransformation filter results in a more narrow grayscale distribution of a given phase, making thesubsequent thresholding or detection processeasier.

Operator Interactive MeasurementsInteractive measurements require the operator touse a mouse to select the start and end points foreach measurement. The most common of theseare linear measurements (point-to-point), parallel,radius or diameter, curvilinear, and angle (Figure55). An operator will typically make several mea-surements over the area of interest. These areeither electronically transferred to a file or tran-scribed by hand. This method is appropriate for alimited number of measurements or where an av-erage is of interest rather than variation. Figure 56represents the most direct way to take an aver-age measurement of thickness using a parallel linetool. One line is aligned with the interface and thenthe opposite line is aligned with the median of therough surface.

In addition to measurements, other annotations areoften placed on electronic images. Annotations

Image Capture and Analysis

Figure 55. Operator interactive tools founds inimaging software.

Figure 56. An example of an operator estimatingcoating thickness with the use of a parallel tool.

9 7

Image Capture and Analysis

Figure 57. A flowchart of the logic used when creating an automated measurement routine for coatingthickness.

9 8

color image, the same 256 values correspond toluminance. Figure 59 demonstrates a yellowbitplane being assigned to the coating (Figure 58)based on the HLS values of the pixels as dis-cussed previously.

Binary OperationsAfter the thresholding process has been accom-plished, the different phases and features in animage are represented by different bitplane coloroverlays. It is possible that more than one phaseor feature of interest is detected by the samebitplane color because of a similar gray level range.A binary operation is the process that separatesand classifies features within the same bitplane,based on morphology or size. Additionally, not allimages are detected as desired. There may beover-detected or under-detected regions that needto modified. In general, binary operations allow theoperator to isolate the features that are of interest.

Two common measurements on this type of coat-ing would include the relative area fraction ofporosity as well as the average thickness. Beforeeither of these measurements can be made it isnecessary to define the total area of the coating.A combination of the binary commands Fill andClose are used to eliminate the holes (undetectedareas) that are totally enclosed by the detectedbitplane. The thickness is determined by averag-ing a series of chord measurements across thecoating, rather than a single measurement of theyellow bitplane. An additional binary layer is su-perimposed in the form of a grid using a Gridcommand. Figure 60 demonstrates the effects ofthese commands on the coating.

To define the individual chords, where the gridoverlaps the coating, a Boolean command is em-ployed. Boolean operations perform logicalfunctions between bitplanes. The two most com-mon are OR and AND. AND takes those portions

Image Capture and Analysis

Figure 58. A TSC color image captured at 1200x 1600 pixels.

Figure 60. The effect of the Binary commands,Close, Fill, and Grid on the detected coating.

Figure 59. Thresholding the coating with ayellow bitplane overlay using the HLS colormodel.

9 9

Image Capture and Analysis

of two bitplanes that overlap and integrates theminto a single destination bitplane. In the coating,for example, AND considers only those segmentsof the red grid bitplane that overlap the yellow coat-ing bitplane. Figure 61 shows the red chords whichresult from the AND command. OR combines twosource bitplanes together to form a single desti-nation bitplane. For example, if two differentinclusion types, oxides and sulfides, are detectedinitially, they can be combined to measure the to-tal inclusion content.

One of the most powerful binary functions is theability to categorize individual features based onshape or size. It can be a single specification(>5-μm) or multiple specifications (>2 and <5)separated by an AND or an OR. If a maximumallowable pore size is specified, all pores that ex-ceed that limit can be flagged. The limit itself canbe based on average diameter, maximum diam-eter, area, aspect ratio, etc. Additionally, the shapeof the pore may have some significance. For ex-ample, in cast materials, a gas bubble tends to becircular while a shrinkage cavity will have a morecomplex shape. Common shape factors examinethe relationship between the perimeter and areaof the feature. For this coating, the pores wereseparated from the elongated oxides to insure thatthe porosity level was not overestimated.

Returning to Figure 61, where grid lines are placedacross the coating, the length of all the individualgrid lines can be measured. The data can beshown as a histogram offering the minimum, maxi-mum, mean and standard deviation of themeasurements across the coating (Figure 62).

Most measurements fall into one of two catego-ries: field or feature-specific measurements. Thegrid lines across the coating are typical of featuremeasurements. Each individual feature is mea-sured and then plotted based on frequency in thehistogram and contributes to the overall statistics.In the case of grain or particle size measurements,features on the edge of the image are eliminatedbecause the complete size of the feature is un-known.

Field measurements are performed on the entirefield of view or a framed portion of the image, pro-viding the sum of the individual measurements inthe selected area. Statistical information for fieldmeasurements is only generated if multiple fieldsare analyzed. This can be employed to observemicrostructural variations within different fields ofa specimen. A fairly common field measurementis area percentage. An area percentage measure-ment results from dividing the number of pixels inthe bitplane of interest by the total number of pix-els in the image. In the case of the coating, thereis a need to measure the area percent porosityrelative to the coating area and not the entire im-age area. This is calculated by dividing the numberof pixels in the bitplane of interest, i.e. the pores inthe coating, by the total number of pixels repre-senting the coating.

Common ApplicationsThroughout the past 20 years of Buehler’s involve-ment in image analysis, a number of applicationsstand out and may be of interest to most imageanalysis users. Below is a listing of these appli-cations:

Grain Size. Image analysis provides a rapid andaccurate means for determining grain size accord-ing to ASTM E 112. Even if etching is unable to

Figure 61. The chords that result from theoverlapping coating and grid lines.

Figure 62. The histogram of resultsrepresenting the chords shown in Figure 61.

100

Image Capture and Analysis

produce complete grain boundaries, or if there aretwins that could skew the data, binary modifica-tions can be employed to make corrections. If aspecification cites maximum grain size limitations,the excessively large grains may be transferredto a different bitplane color to provide visual andnumerical feedback.

Porosity. Porosity is detrimental to the physicalproperties of most engineering materials. Imageanalysis is able to characterize the pores accord-ing to the total number of pores, number per unitarea, maximum size, average size and the sizedistribution in the form of a histogram.

Linear Measurements. Simple point-to-point mea-surements are widely used for making occasionalmeasurements; however, in cases where a highquantity of measurements and more statistics arerequired, automated image analysis is time sav-ing. A coating or layer is detected based on thepixel values and then, after binary isolation of thecoating, grid lines are superimposed. Using Bool-ean logic to evaluate the common pixels betweenthe layer and grid lines, the result is many chordsrepresenting the thickness at given points in thecoating.

Feature Shape and Size. The shape of the graph-ite constituent in ductile irons is critical. Ductileiron was developed such that the graphite wouldoccur in the form of spherical nodules with the re-sult of dramatically improved mechanicalproperties. However, variations in chemistry andother factors can cause the nodules to be irregu-lar, leading to some degradation of the properties.The ability to monitor the graphite shape or deter-mine “nodularity” is another ability of imageanalysis. These same techniques are applicableto any constituent that can be detected.

Phase Percentage. The area percent of variousphases in a microstructure influences the proper-ties. The tensile strength of grey iron, for example,is directly related to the percentage of pearlite inits microstructure. In addition, in a single image,multiple phases can be detected, measured andpresented in one graph. For example, when evalu-ating the inclusion content of steels, it would beuseful to examine the overall inclusion content aswell as isolating particular inclusion types, suchas oxides and sulfides.

HELPFUL HINTS FOR

IMAGE CAPTURE AND

ANALYSIS

Determine the requiredpixel density of a camera system based onthe size of the features of interest and ac-ceptable file size

When distributing an image, always add ascale bar

When saving images for analysis at a latertime, verify that the image is stored as cap-tured, i.e., without any filters applied.

101

Laboratory Safety

LABORATORY SAFETY

The metallographic laboratory is a relatively safeworking environment; however, there are dangersinherent to the job. Included in these dangers isexposure to heat, acids, bases, oxidizers andsolvents. Specimen preparation devices, such asdrill presses, shears and cutoff saws, also presenthazards. In general, these dangers can be mini-mized if the metallographer consults documentssuch as ASTM E 2014 (Standard Guide on Metal-lographic Laboratory Safety) and relevant MaterialSafety Data Sheets (MSDS) before working withunfamiliar chemicals. Common sense, caution,training in basic laboratory skills, a laboratorysafety program, access to safety reference books– these are some of the ingredients of a recipe forlaboratory safety. Table 48 lists the main require-ments for a comprehensive safety program.

Safe working habits begin with good housekeep-ing. A neat, orderly laboratory promotes safeworking habits, while a sloppy, messy work areainvites disaster. Good working habits include suchobvious, commonsense items as washing thehands after handling chemicals or before eating.Simple carelessness can cause accidents (seeFigure 63). For example, failure to clean glass-ware after use can cause an accident for the nextuser. Another common problem is burns due tofailure to properly clean acid spills or splatter.

Most reagents, chemical or electrolytic polishingelectrolytes, and solvents should be used undera ventilation hood designed for use with chemi-cals. Many of these chemicals used in metallo-graphy can cause serious damage on contact. Itis best to assume that all chemicals are toxic andall vapors or fumes will be toxic if inhaled or will bedamaging to the eyes. A hood will prevent theworking area from being contaminated with thesefumes. However, you will not be protected whenyour head is inside the hood. In most cases, aprotective plastic or shatterproof glass shield canbe drawn across the front of the hood for furtherprotection from splattering or any unexpected re-actions.

All laboratories should be equipped with a showerand eyewash for emergency use. This equipmentshould be near the work area so that the injuredcan reach it quickly and easily. Fire alarms and

Figure 63. Carelessness in the laboratory can resultin dangerous chemical spills and a clean-upnightmare.

1. Emergency Responsea. First Respondersb. Emergency Phone Numbersc. First Aidd. Spills

2. Chemical Procurement, Distribution, and Storagea. MSDS Managementb. Labeling/Storage

3. Laboratory SOPsa. Handling Chemicalsb. Mixing Chemicalsc. Rulesd. Hygiene

4. Employee Health and Exposurea. Personal Protective Equipmentb. Monitoring of exposure levelsc. Medical Evaluationsd. Additional protections for employees working withparticularly hazardous substances (carcinogens, toxins)

5. Disposal of Hazardous Materialsa. Etchants and Chemicalsb. Recirculating Tanksc. Lubricants and Abrasives

6. Equipment Use (JSAs)a. Sectioningb. Mountingc. Grinding and Polishingd. Electropolishinge. Etching

7. Facilitya. Facility and Equipment Maintenanceb. Housecleaningc. Fume Hood Air Flowd. Fire Extinguishers and Fire Protectione. Spill Responsef. Showers and Eye Washesg. Air Lines and Filtersh. Plumbingi. Electricals

8. Employee Traininga. Safetyb. Indications and Symptoms of Exposurec. Job Functions

9. Scheduled and Documented Safety Inspectionsand Meetings

Table 48. Elements of a Comprehensive Metallography Laboratory Safety Plan

102

Laboratory Safety

fire extinguishers (CO2 type) should be availableand tested periodically. A good first aid kit and achemical spill treatment kit should be readily avail-able.

Laboratory EquipmentSpecimen preparation devices used in metallo-graphic laboratories are generally quite safe to use.Information supplied by the manufacturer usuallydescribes safe operating procedures. It is goodlaboratory practice to prepare a job safety analy-sis (JSA) detailing potential hazards anddescribing the safe operating procedure for eachpiece of equipment. This information should be pro-vided to all users and it must be revised andreviewed periodically.

Band saws or abrasive cutoff saws are commonlyused by metallographers. The cutting area of bandsaws is exposed and potentially dangerous. Yourhands should never be used to guide the workpiece during cutting. A guiding device or block ofwood should always be used between the workpiece and your hands. After cutting is completed,the saw should be turned off before pieces nearthe blade are removed. Samples should behandled carefully, because considerable heat canbe generated. In addition, sharp burrs are oftenpresent, which should be carefully removed by fil-ing or grinding. Abrasive cutoff saws are safer touse because the cutting area is closed off duringuse. The chief danger is from flying pieces from abroken wheel. Fortunately, the closed cover con-tains these pieces within the cutting chamber.Wheel breakage usually occurs when the part isnot firmly clamped in place or if excessive pres-sure is applied, a bad practice from the standpointof specimen damage as well.

Dust produced during grinding of metals is alwaysdangerous. For certain metals, like beryllium, mag-nesium, lead, manganese, and silver, the dustsare extremely toxic. Wet grinding is preferred bothfor dust control and for preventing thermal dam-age to the specimen. Bench grinders must be firmlymounted to prevent sudden movement. Care mustbe exercised to avoid grinding one’s fingers orstriking the edge of a grinding belt, which will causepainful lacerations. With nearly all materials, wetgrinding is preferred and produces best results.For routine handling of dangerous metals, grind-

ing should be done wet under a ventilation hood.Waste must be handled carefully and disposed ofproperly. Radioactive materials require specialremote- handling facilities and elaborate safety pre-cautions.

A drill press is frequently used in the laboratory.Drilling holes in thin sections requires secureclamping; otherwise the sample can be grabbedby the drill and spun around, inflicting serious lac-erations. Hair, ties, and shirt cuffs can becometangled in a drill, inflicting serious injuries. Safetyglasses should always be worn when using drillpresses or when cutting or grinding. Mountingpresses or laboratory heat-treatment furnacespresent potential burn hazards. It is a good prac-tice to place a “hot” sign in front of a laboratoryfurnace when it is in use. Gloves should be wornwhen working with these devices. Modern mount-ing presses that cool the cured resin back to nearroom temperature dramatically reduce the poten-tial for burns.

It is occasionally necessary to heat solutions dur-ing their preparation or use. Although Bunsenburners are commonly employed for this purpose,it is much safer to use a hot plate or water bathand thus avoid the use of an open flame. If a Bun-sen burner is used, the flame should never beapplied directly to a flask, beaker, or dish. Plain-or asbestos-centered wire gauze should alwaysbe placed between the flame and the container.

Personal Protective Equipment (PPE)Metallographer must take certain precautions toinsure their personal safety. A laboratory coat isuseful for protecting the operator’s clothing, andshould be changed regularly and cleaned profes-sionally. When handling caustics, a rubberized orplastic coated apron provides better protection.Gloves should be worn when handling bulksamples, working with hot material, or using haz-ardous solutions. Lightweight surgeons’ gloves arevery popular, because the operator retains theability to “feel.” Many metallographers wear theseto protect their skin when mounting with epoxies,and polishing with oxide suspensions. When us-ing these gloves with chemicals, always inspectfor holes, as they are easily punctured. Thick rub-ber gloves are often used for handling specimensduring macroetching, chemical polish

103

Laboratory Safety

ing, pickling, etc. The gloves should always bechecked first for small holes or cracks,because gloves can impart a false sense of se-curity. The operator’s hands generally perspirewhen using rubber gloves and it is sometimes dif-ficult to tell if the moisture is due solely toperspiration or to leakage. Safety glasses shouldbe worn during processes that generate particu-late matter. Goggles are appropriate for use withchemicals, and a chemical face shield is recom-mended when handling large quantities ofhazardous liquids. The appropriate PPE will bespecified on the MSDS for most laboratory chemi-cals and products.

Chemicals, Storage and HandlingMany of the chemicals used in metallographyare toxic, corrosive, flammable, or potentiallyexplosive. Whenever possible, purchase smallquantities that are likely to be used within a rea-sonably short time. Flammable solvents should bestored in fireproof steel cabinets. Acids and basesshould be stored separately, again in fireproof steelcabinets. Strong oxidants must not be stored alongwith acids, bases or flammable solvents.

Reagent-grade chemicals or solvents of highestpurity are recommended. Although more expen-sive, the amounts used are small and the gain insafety and reliability compensates for the cost dif-ference. Chemicals may deteriorate during storage.Exposure to light can accelerate deterioration ofsome chemicals. Hence, they should be stored ina closed metal cabinet.

EtchantsMost laboratories mix commonly used reagentsin quantities of 250 to 1000 mL and then store themas stock reagents. Many reagents can be safelyhandled in this manner. It is best to store only thosereagents that are used regularly. Glass-stopperedbottles are commonly used as stock reagentbottles. If these bottles are opened regularly, thestopper will not become “frozen”. However, if theyare used infrequently, a frozen stopper often re-sults. Holding the neck of the bottle under a streamof hot water will usually loosen them. If thermalexpansion does not free the stopper, the stoppercan be gently tapped with a piece of wood. Glassbottles with plastic screw-on tops can be used so

long as the solution does not attack the plastic.These bottles are useful for holding solutions, suchas nital, that can build up gas pressure within atightly stoppered bottle. A small hole can be drilledthrough the cap top to serve as a pressure reliefvent. Tightly stoppered bottles of nital and someother solutions have exploded as the result of pres-sure buildup. Be certain that the reagent is safe tostore and store only small quantities. All bottlesshould be clearly labeled. Polyethylene bottles arerequired for etchants containing hydrofluoric acid,which attacks glass.

Most recipes for etchants or electrolytes list theingredients by weight if they are solids and byvolume if they are liquids. In a few cases, allamounts are given in weight percentages. In mostcases, reagent compositions are not extremelycritical. An ordinary laboratory balance providesadequate weighing accuracy, while graduated cyl-inders provide acceptable accuracy for volumetricmeasurements. These devices should be cleanedafter use to prevent accidents to the next user.For weight measurements, a clean piece of filterpaper, or a cup, should be placed on the balancepan to hold the chemical, to protect the pan sur-face, and to facilitate transfer to the mixing beaker.A large graduated beaker is usually employed formixing solutions.

With many etchants, the mixing order is impor-tant, especially when dangerous chemicals areused. When water is specified, distilled watershould always be used, because most tap watercontains minerals or may be chlorinated orfluorinated. Tap water can produce poor results orunexpected problems. Cold water should alwaysbe used, never warm or hot water, which cancause a reaction to become violent. In mixing, oneshould start with the solvents, such as water andalcohol; then dissolve the specified salts. A mag-netic stirring device is of great value, as shown inFigure 64. Then, the dangerous chemicals, suchas acids, should be added carefully and slowlywhile the solution is being stirred. Whenever sul-furic acid (H2SO4) is specified, it should be addedlast. It should be added slowly, while stirring, andit should be cooled, if necessary, to minimize heat-ing. Never just pour one liquid into another, asshown in Figure 65. If sulfuric acid is

104

Laboratory Safety

added to water without stirring, it can collect at thebottom of the beaker and enough local heating canoccur to throw the contents out of the beaker.

The literature contains references to a great manyformulas for etchants, chemical polishes, and elec-trolytes that are potentially dangerous or extremelydangerous. Few of these references contain com-ments regarding safe handling procedures orpotential hazards. Fortunately, metallographic ap-plications involve small quantities of thesesolutions, and accidents do not usually producecatastrophic results. However, even with small

solution volumes considerable damage can be,and has been, done. Table 49 lists examples fromthe literature (27-29) of chemical polishing solu-tions, electrolytic polishing solutions and etchantsthat have been involved in accidents. Table 50lists a number of commonly used chemicals andincompatible chemicals.

SolventsNumerous organic solvents are used for cleaningor are ingredients in chemical or electrolytic pol-ishing solutions or etchants, where they are usedto control ionization or the speed and mode of at-tack. Commonly employed solvents include water,acetone, ethyl ether, ethylene glycol, glycerol(glycerin), kerosene, petroleum ether, trichloroet-hylene, butyl cellosolve, and alcohols, such asamyl alcohol, ethanol, methanol, and isopropyl al-cohol. Most are flammable and their vapors canform explosive mixtures with air. They should bekept closed when not in use and should be storedin a cool place away from heat and open flames.

Acetone (CH3COCH3) is a colorless liquid with afragrant mintlike odor. It is volatile and highly flam-mable. It is an irritant to the eyes and the mucousmembranes. Its vapor is denser than air and cantravel along the ground and can be ignited at adistance. Acetone can form explosive peroxideson contact with strong oxidizers such as aceticacid, nitric acid and hydrogen peroxide. It is anirritant to the eyes and respiratory tract, and willcause the skin to dry and crack.

Butyl cellosolve (HOCH2CH2OC4H9), or ethyleneglycol monobutyl ether, is a colorless liquid with arancid odor that is used in electropolishing solu-tions. It is combustible and may form explosiveperoxides. It is toxic in contact with the skin, canbe absorbed through the skin, can cause seriousdamage to the eyes, irritation to the skin and res-piratory tract.

Carbitol (C2H5OCH2CH2OCH2CH2OH), or diethyl-ene glycol monoethyl ether, is a colorless, viscoussolvent that is compatible with water and is usedin electropolishing solutions. It irritates the skin,eyes, mucous membranes, and upper respiratorytract, and is harmful if inhaled or swallowed.

Figure 64. When mixing etchants, use a magneticstirring plate with a magnetic bar for stirring.Slowly add the liquid ingredients to the solvent bydripping them down a glass stirring rod. If thesolution is more dangerous than this one, wearprotective gloves and use a face shield. If mixinggenerates substantial heat, it is a good practice toplace a cooling jacket around the beaker.

Figure 65. Illustration of a bad mixing practice. Theacid ingredient was poured into an empty acidbottle and the solvents were added without stirringor cooling. The solution may erupt in the metallo-grapher’s face at any moment. It is recommended tokeep the hands clear of the area and remove rings.

105

Laboratory Safety

Table 49. Chemical and Electrolytic Polishing Solutions and Etchants Known to be Dangerous

Solution Use Problems

5 parts Lactic Acid Chemical polishing Lactic and nitric acids react autocatalytically.5 parts HNO3 solution for Zr Explosion can occur if stored.2 parts Water1 part HF

50 parts Lactic Acid Chemical polishing Lactic and nitric acids react autocatalytically.30 parts HNO3 solution for Ta, Nb Explosion can occur if stored.2 parts HF and alloys

3 parts Perchloric Acid Electropolishing Mixture is unstable and can, and has, exploded,with 1 part Acetic Anhydride solution for Alheating, or in the presence of organic compounds adding to the potentialhazard.

60-90 parts Perchloric Acid Electropolishing Solution will explode at room temperature.40-10 parts Butyl Cellosolve solution Solutions with ≤30% HCIO4 will be safe if T is <20°C.

100g CrO3 Electropolishing CrO3 was dissolved in water, cooled to about 20°C;200mL Water solution the acetic anhydride was added

very slowing with700mL Acetic Anhydride stirring. The solution became warm to the touch.

About 20 seconds later it erupted from the beaker.

1 part HNO3 Electropolishing Mixture is unstable and cannot be stored2 parts Methanol solution for Muntz

(Cu-40% Zn) metal

20mL HF Etchant for Nb, Ta, This etchant is unstable. At 20°C, it reacted after 1810mL HNO3 Ti, V, Zr hours. At 30-35°C, it reacted after 8 hours with30mL Glycerol violence. At 100°C, it will react after 1 minute.

20-30mL HCI Etchant for Ni and Two incidents occurred when a violent reaction10mL HNO3 stainless steels resulted producing NO2 and a spray of acid30mL Glycerol after the etch was left standing for 2-3 hours.

40mL Acetic Acid Etchant for Ni A closed bottled exploded about 4 hours after40mL Acetone it was mixed. Authors state that solutions40mL HNO3 without the acetic acid also cannot be stored.

10mL HNO3 Etchant The solution reacted spontaneously about 2minutes 10mL Acetic Acid after mixing withevolution of heat and fumes (nitrous20mL Acetone and nitric oxides). The mixed acids were poured into

the acetone. The beaker was externally cooled.

50mL Nitric Acid Etchant Mixtures have exploded violently after mixing950mL Isopropyl Alcohol or about 20 minutes after mixing.

Table 50. Some Incompatible Chemicals

Chemical Use in Metallography Do Not Mix With the Following:

Acetic Acid Chemical polishing Chromic acid, glycol, hydroxol compounds,electrolytic polishing nitric acid, peroxides, permanganates

Acetone Degreasing, cleaning, Concentrated solutions of nitric andetchants sulfuric acid

Chromic Acid Electropolishing Acetic acid, flammable liquid, glycerol

Hydrogen Peroxide Chemical polishing, etchants Flammable liquid, organic materials

Nitric Acid (conc) Chemical polishing, etchants Acetic acid, chromic acid, flammable liquids,isopropyl alcohol

Perchloric Acid Electropolishing Acetic anhydride, alcohol, some organics,oil, grease

Sulfuric Acid Etchants Methyl alcohol, potassium chlorate,potassium perchlorate and potassiumpermanganate

106

Laboratory Safety

Ethylene glycol (HOCH2CH2OH) is a colorless,hygroscopic liquid with a sweet taste (but do notswallow as it is poisonous) that reacts with strongoxidants and strong bases. It is flammable. Thesubstance irritates the eyes, skin, and the respi-ratory tract.

Glycerol (glycerin) (CH2OHCHOHCH2OH) is acolorless or pale yellow, odorless, hygroscopic,syrupy liquid with a sweet, warm taste. It is rela-tively nontoxic and nonvolatile but can cause iritis(inflammation of the iris). It is combustible and is amoderate fire hazard. Glycerol should never beused in anhydrous solutions containing nitric andsulfuric acids, because nitroglycerin can form.Glycerol should not be used with strong oxidizingagents, such as chromium trioxide and potassiumpermanganate, as an explosion may occur. Glyc-erol is often added to aqua regia (glyceregia). Thismixture decomposes readily and should be dis-carded immediately after use. This etchant shouldnot be allowed to stand for more than about 15minutes after mixing.

Kerosene is occasionally employed in grindingsamples and with diamond paste as a lubricant.Only the deodorized form should be used. It is flam-mable, but the vapors do not readily explode.Contact defattens the skin and can cause derma-titis, irritation or infections.

Trichloroethylene (CHCl:CCl2) is a stable, color-less liquid with a chloroform-like odor. Effectivelaboratory ventilation is necessary. At ambient tem-peratures it is nonflammable and nonexplosive, butbecomes hazardous at higher temp-eratures. In the presence of strong alkalies, withwhich it can react, it can form explosive mixtures.In the presence of moisture, the substance canbe decomposed by light to corrosive hydrochloricacid. It is carcinogenic to humans, and toxic wheninhaled or ingested, which may cause acute poi-soning.

Amyl alcohol (CH3(CH2)4OH), or 1-Pentanol, is acolorless liquid with noxious odor. It is flammableand the vapors may form explosive mixtures atelevated temperatures. The substance reacts vio-lently with strong oxidants and attacks alkalinemetals. The fumes are irritating to the eyes, upperrespiratory tract, and skin. The substance is toxicthrough ingestion, inhalation, or absorption throughthe skin.

Ethyl alcohol (CH3CH2OH), or ethanol, is a color-less, inoffensive solvent commonly used inmetallography. Ethanol is miscible with water andrapidly absorbs up to 5% water from the air. Thedenatured version is less expensive and contains5% absolute methanol and is suitable for anyrecipe requiring ethyl alcohol. It is a dangerousfire hazard, and its vapors are irritating to the eyesand upper respiratory tract. High concentrationsof its vapor can produce intoxication. Becauseethanol is completely burned in the body, it is nota cumulative poison like methanol.

Methyl alcohol (CH3OH) is an excellent, non-hy-groscopic solvent, but it is a cumulative poison.Ingestion, inhalation or absorption through the skinin toxic levels can damage the central nervoussystem, kidneys, liver, heart, and other organs.Blindness has resulted from severe poisoning. Itis particularly dangerous because repeated low-level exposures can also cause acute poisoningas a result of accumulation. Ethanol should be usedwhenever possible. When using methanol, alwayswork under a ventilation hood. Mixtures of metha-nol and sulfuric acid can form dimethyl sulfate,which is extremely toxic. Solutions of methanoland nitric acid are more stable than mixtures ofnitric acid and higher alcohols.

Isopropyl alcohol [CH3CH(OH)CH3], also knownas 2-propanol, is a clear, colorless liquid that, likeethanol, does not accumulate in the body, althoughit does have a strong narcotic effect. It is a flam-mable liquid and is a dangerous fire hazard.Metallographers have used it as a substitute forethanol but isopropyl alcohol has quite differentcharacteristics and should not be used. Fatal inju-ries and explosions have been reported due to itsuse.

AcidsInorganic and organic acids are common constitu-ents in chemical and electrolytic polishing solutionsand in etchants. The inorganic, or mineral acids,including the very familiar acids such as hydro-chloric, nitric, perchloric, phosphoric, and sulfuric,are highly corrosive and poisonous.They shouldbe stored in a cool, well-ventilated location awayfrom potential fire hazards and, of course, awayfrom open flames. They should not be stored in alocation that receives direct sunlight. When the pureacids contact metals, most

107

Laboratory Safety

liberate hydrogen gas – a fire and explosion haz-ard. The organic acids are naturally occurringsubstances in sour milk, fruits, and plants and in-clude the following acids: acetic, lactic, citric,oxalic, and tartaric.

Hydrochloric acid (HCl), commonly used in met-allography, is a colorless gas or fuming liquid witha sharp, choking odor. It is very dangerous to theeyes and irritating to the nose and throat. It at-tacks the skin strongly causing severe burns.

Nitric acid (HNO3), also commonly used in metal-lography, is a colorless or yellowish fuming liquid,highly toxic and dangerous to the eyes. If it con-tacts organic material or other easily oxidizablematerials, it can cause fires and possibly explo-sions. When it reacts with other materials, toxicoxides of nitrogen are produced. The oxides, whichvary with the conditions, include nitrous acid, ni-trogen dioxide, nitric oxide, nitrous oxide, andhydroxylamine. A commonly encountered probleminvolves pouring nitric acid into a graduated cylin-der that contains some methanol or ethanol fromprior use. The brown fumes given off are quiteharmful. Mixtures of nitric acid and alcohols higherthan ethanol should not be stored. Mixtures of con-centrated nitric and sulfuric acids are extremelydangerous, while strong mixtures of nitric acid andglycerin or glycols can be explosive. Aqua regia,a mixture of one part nitric acid and two to fourparts hydrochloric acid, forms several productsincluding nitrosyl chloride, an exceptionally toxicgas. Aqua regia is a popular etchant but must beused with care under a hood.

Ethanol with additions of up to 3% nitric acid (nital)can be safely mixed and stored in small quanti-ties. Higher concentrations result in pressurebuildup in tightly stoppered bottles. Explosions of5% nitric acid in ethanol have occurred as a resultof failure to relieve the pressure. If higher con-centrations are desired, they can be mixed daily,placed in an open dish, and used safely. Discardthe etchant at the end of the day. Mixtures ofmethanol with up to 5% nitric acid are safe to useand store in small quantities. Mixtures of metha-nol with more than 5% nitric acid if heated aresubject to violent decomposition. Mixtures of 33%nitric acid in methanol have decomposed suddenlyand violently.

Never add nitric acid to isopropyl alcohol. Ander-son (27) reported that a litre bottle of 5% nitric acidin isopropyl alcohol was mixed and placed in acabinet. Although this had been done many timesin the past without problems, twenty minutes laterthe bottle exploded destroying the cabinet, otherstored bottles, and throwing debris up to 20 feetaway. Anderson (27) also reported that a metal-lographer was pouring a freshly mixed litre of 5%nitric acid in isopropyl alcohol into another bottlewhen it exploded. The person died within threehours without being able to tell anyone what hap-pened. As with the other explosion, this sameprocedure had been performed many times previ-ously without mishap. Anderson recommendsavoiding use of isopropyl alcohol completely.

Most metallographers consider nital to be very safeto use, and indeed it is. However, even with suchan apparently safe solution, one can have acci-dents. One such accident occurred when anemployee, not a skilled metallographer, was re-plenishing a stock of 5% nitric acid in ethanol usinga procedure that he had claimed to have performedmany times previously (he was not taught the safeway to mix nital as it was a union chemist’s job tomix nital i.e., not his job). The worker began byadding the desired volume of concentrated nitricacid into the container that contained a small re-sidual amount of stale 5% nital. To his surprise,the contents began “boiling” and spewing out ofthe container along with dense, brown fumes. Theacid splashed the worker, resulting in burns onhis forehead, face and eyes. The small amount ofaged nitric acid solution (the concentration mayhave been increased due to evaporation of thealcohol), present in the container when the freshacid was added, created a dangerous chemicalreaction. An experiment also showed that a simi-lar reaction can occur when nitric acid is pouredinto a graduated cylinder containing only remnantsof ethanol or methanol.

Sulfuric acid (H2SO4) is a colorless, oily liquid thatis highly corrosive, a strong oxidizing agent, anddangerously reactive. It reacts violently with basesand is corrosive to most metals forming flammable/explosive hydrogen gas. It reacts violently withwater and organic materials with evolution of heat.Upon heating, toxic sulfur oxides are formed. Oneshould add sulfuric acid very

108

Laboratory Safety

slowly to water with constant stirring. If added with-out stirring, it will produce a pocket of steam in thebottom of the vessel, throwing the contents out ofthe vessel. Concentrated sulfuric acid can causesevere, deep burns on contact with the skin, andpermanent vision loss on contact with the eyes.Tissue is destroyed by the acid’s dehydrating ac-tion. Lungs may be affected by long-term or chronicexposure to aerosol. Skin lesions, tooth erosion,and conjunctivitis are other long-term effects.

Hydrofluoric acid (HF) is a clear, colorless, fum-ing liquid or gas with a sharp, penetrating odor. Itis very dangerous to the eyes, skin, and upperrespiratory tract. The substance can be absorbedinto the body by inhalation, through the skin andby ingestion. A harmful concentration of the gas inair can be reached quickly, making it very dan-gerous to handle. Exposure by ingestion, inhalationor contact, can be fatal. Undissociated HF posesa unique threat in that it can destroy soft tissuesand result in decalcification of the bone. Moreover,the effects may be delayed. Laboratories whereHF is used should stock an antidote kit to be usedin case of exposure. Although it is a relatively weakmineral acid, HF will attack glass or silicon com-pounds, and should be measured, mixed, andstored in polyethylene vessels. HF reacts withmany compounds, including metals, and will liber-ate explosive hydrogen gas.

Orthophosphoric acid (H3PO4), a colorless thickliquid or hygroscopic crystal, is a medium strongacid. It is corrosive to the skin, eyes and respira-tory tract. Phosphoric acid decomposes oncontact with alcohols, aldehydes, cyanides, sul-fides, ketones, and can react with halogenatedorganic compounds forming organophosphorusnerve-gas type compounds that are extremelytoxic. It reacts violently with bases, and will gen-erate hydrogen gas when it reacts with metals.

Perchloric acid (HClO4) is a colorless, fuming, hy-groscopic liquid. It is extremely unstable inconcentrated form and may explode by shock orconcussion when dry or drying, so commerciallyavailable perchloric acids come in concentrationsof 65 to 72%. In this form, contact with perchloricacid will cause irritation and burns, while its fumesare highly irritating to the mucous membranes.Contact with organic, or other easily oxidizedmaterial, can form highly unstable perchlorates,

which can ignite and cause explosions. Regularuse of perchloric acid requires that the ventilationsystem must be specifically designed and main-tained for perchloric acid. Special fume hoods witha waterfall-type fume washer will remove the per-chlorate fumes before they can enter the exhaustsystem.

Perchloric acid is very useful in electropolishingsolutions. Never electropolish samples mountedin phenolic (Bakelite) or other plastics in perchlo-ric acid solutions as explosions can result. Themixture of perchloric acid and acetic anhydride,which was developed by Jacquet, is difficult toprepare and highly explosive. Jacquet has re-viewed the accidents involving perchloric acid andhas described safety procedures (30). The worstaccident occurred on February 20, 1947, in anelectroplating factory in Los Angeles. In this acci-dent 17 people were killed and 150 were injured(29). Medard, Jacquet, and Sartorius have pre-pared a ternary diagram showing safe com-positions of perchloric acid, acetic anhydride, andwater, Figure 66. Anderson, however, states thataccidents have still occurred with solutions in the“safe” region of this diagram (27). Electropol-ishingsolutions composed of perchloric acid and aceticanhydride are not recommended. Many compa-nies forbid the use of such mixtures, and somecities have banned their use. Electropol-ishing so-lutions of perchloric acid and alcohol, with orwithout organic additions, and mixtures of perchlo-ric acid and glacial acetic acid are safe to use.Nevertheless, in using these “safe” mixtures, oneshould follow the formula instructions carefully, mixonly small quantities, keep the temperature undercontrol, and avoid evaporation. These solutionsshould not be stored.

Mixtures of acetic acid and 5-10% perchloric acidhave been commonly used to electropolishiron-based alloys and are reasonably safe. Do notuse these solutions to electropolish bismuth, ar-senic or tin, as explosions have occurred.Anderson suggests that arsenic, antimony, andtin may also be incompatible with perchloricelectrolytes (27). Do not store these electrolytes .Discard them when they become colored by dis-solved metallic ions (from electropolishing).Always keep these solutions cool; increasing thetemperature increases the oxidizing power of per-chloric acid.

109

Laboratory Safety

Comas et al. have studied the hazards associ-ated with mixtures consisting of butyl cellosolveand from 10 to 95% of 70% perchloric acid (31).Mixtures with 60 to 90% acid were explosive atroom temperature. Acid concentrations of 30% orless were inflammable but were judged to be safeto use as long as the operating temperature doesnot exceed 20°C.

Acetic acid (CH3COOH) is a clear, colorless liq-uid with a pungent odor. It is a weak acid that reactswith strong oxidizers, bases and metals. It is flam-mable and is not easily ignited, although whenheated, it releases vapors that can be ignited andcan travel some distance to an ignition source.Contact with the skin results in serious burns. In-halation of the fumes irritates the mucousmembranes. Anderson states that acetic acid is agood solvent for nitric acid and that a 50% solu-tion can be prepared, but not stored, withoutdanger (27). Sax, however, states that mixturesof nitric and acetic acids are dangerous (32).

Acetic anhydride [(CH3CO)2O], or acetic oxide, isa colorless liquid with a very strong acetic odor. Itcan cause irritation and severe burns to the skinand eyes. Acetic anhydride decomposes on heat-ing producing toxic fumes. It reacts violently withboiling water, steam, strong oxidants (specificallysulfuric acid), alcohols, amines, strong bases, andothers. It attacks metals, and is very corrosive,especially in presence of water or moisture. It isextremely flammable and should be avoided. Theelectrolytic polishing mixtures of acetic anhydrideand perchloric acid (4-to-1 to 2-to-1 mixtures) de-veloped by Jacquet, as mentioned above, areexceptionally dangerous and should never beused. Dawkins (33) reported an accident involv-ing a mixture of chromium trioxide and aceticanhydride that had been used for electropolishing(Table 49).

Citric acid [C3H4(OH)(COOH)3·H2O] comes as col-orless, odorless crystals that are water-soluble. It

Figure 66. Diagram developed by Médard, Jacquet and Sartorius showing safe (A and 1 to 9) nonexplosiveelectropolishing solutions and explosive (C, E and Los Angeles) compositions of perchloric acid, aceticanhydride and water (brought in by the perchloric acid, but not corrected for the effect of acetic anhydride).

110

is an irritant to the skin, eyes and respiratory tract;no unusual problems are encountered except foroccasional allergic reactions.

Lactic acid (CH3CHOHCOOH) is a yellow or col-orless thick liquid. It is damaging to the eyes.

Oxalic acid (COOHCOOH·2H2O) comes as trans-parent, colorless crystals. It is poisonous if in-gested and irritating to the upper respiratory tractand digestive system if inhaled. Skin contact pro-duces caustic action and will discolor and embrittlethe fingernails. It is not compatible with nitric acid,as it reacts violently with strong oxidants. It canalso form explosive compounds due to reacts withsilver.

Picric acid [(NO2)3C6H2OH], or 2,4,6-trinitrophenol,comes as yellow crystals that are wet with 10 to35% water. When picric acid is dry, it is a danger-ous explosive. It is toxic and stains the skin and isincompatible with all oxidizable substances. Pi-crates, which are metal salts of picric acid, areexplosive. When picrates are dry, they can deto-nate readily, possibly spontaneously. Purchase insmall quantities, keep it moist, and store it in asafe, cool place. If it starts to dry out, add a smallamount of water to keep it moist. The maximumsolubilities of picric acid in water and in ethanolare about 1.3 and 8 g per 100 mL, respectively.Picral can be stored safely. During use, the solu-tion should not be allowed to dry out. The etchingresidue should be discarded at the end of the dayto avoid potential explosions.

BasesBases, such as ammonium hydroxide (NH4OH),potassium hydroxide (KOH), and sodium hydrox-ide (NaOH), are commonly used in metallo-graphy, chiefly in etchants.

Ammonium hydroxide is a colorless liquid with astrong, obnoxious odor. Solutions are extremelycorrosive and irritating to the skin, eyes, and mu-cous membranes. It reacts exothermically withsulfuric acid and other strong mineral acids, pro-ducing boiling solutions.

Sodium and potassium hydroxides are strongbases, available as white deliquescent pellets thatare soluble in water. They can rapidly absorb car-bon dioxide and water from the air. Solutions stored

in flasks with ground glass stoppers may leak airand freeze the stoppers, making re-opening diffi-cult. Dissolving NaOH or KOH in water willgenerate considerable heat. Do not dissolve ei-ther in hot water. Never pour water onto thesehydroxides; always add the pellets slowly to thewater. Alkali metal hydroxides react violently withacid, and are corrosive in moist air to metals likezinc, aluminum, tin and lead forming flammable/explosive hydrogen gas. They are very corrosiveto skin, eyes and respiratory tract. Long-termexposure may lead to dermatitis. Potassium hy-droxide is somewhat more corrosive than sodiumhydroxide.

Other ChemicalsHydrogen peroxide (H2O2) is available as a liquidin concentrations of either 3 or 30%. The 3% solu-tion is reasonably safe to use, while the 30%solution is a very powerful oxidant whose effecton the skin is about as harmful as that producedby contact with sulfuric acid. Hydrogen peroxideby itself is not combustible, but if brought in con-tact with combustible materials, it can produceviolent combustion. Hydrogen peroxide is verydamaging to the eyes. Because release of oxy-gen can cause high pressures to develop withinthe container, the container caps are vented.

Bromine (Br2), a fuming reddish brown liquid witha pungent, suffocating odor, is commonly used indeep-etching solutions. It is very corrosive, react-ing violently with easily oxidized substances,including some metals. Bromine is a dangerousliquid that should only be handled by well-qualifiedpersonnel. Its vapors are extremely irritating to theeyes, skin, and mucous membranes. Skin con-tact produces deep, penetrating burns that areslow to heal. Contact with organic matter cancause fires.

Chromic acid (H2CrO4) is formed when chromiumtrioxide (CrO3) is dissolved in water. CrO3 is usedin electropolishing solutions (see previous com-ment and Table 49 about explosive nature ofmixtures with acetic anhydride). Dilute aqueoussolutions are widely used for attack polishing. It isa powerful oxidant; always wear gloves when us-ing it for attack polishing or use automatic devicesand avoid contact potential. Chronic or long-terminhalation exposure may produce asthma-like re-actions.

Laboratory Safety

111

Potassium permanganate (KMnO4), a black crys-talline powder, is a powerful oxidant used inetchants. It is a dangerous fire and explosion haz-ard, especially when in contact with organicmaterials. Ingestion produces serious damage.KMnO4 and sulfuric acid should never be mixedtogether because a violent explosion can result.

Potassium dichromate (K2Cr2O7), a bright orangecrystalline powder, is another powerful oxidant thatis used in etchants. Contact can cause ulcerationof the hands, severe damage to nasal tissue, orasthma and allergies with long-term exposure.

Cyanide compounds are occasionally used in met-allographic applications. Potassium cyanide (KCN)and sodium cyanide (NaCN) are extremely dan-gerous and highly toxic. Exposure by eye or skincontact, or by ingestion, is fatal. NaCN and KCNvapors are intensely poisonous. They are particu-larly hazardous when brought in contact with acidsor acid fumes because of liberation of hydrogencyanide, which is extremely toxic and highly flam-mable. Potassium ferricyanide (K3Fe(CN)6), aruby-red crystalline powder and an ingredient inMurakami-type reagents, is poisonous but stableand reasonably safe to use.

A number of nitrates, such as ferric nitrate[Fe(NO3)3·6H2O], lead nitrate [Pb(NO3)6], and sil-ver nitrate (AgNO3), are employed by metallo-graphers. Since they are powerful oxidizers, theypose a dangerous fire hazard, especially when incontact with organic materials. They may evolvetoxic fumes, such as oxides of nitrogen and lead,and are poisonous and corrosive to the eyes, skinand respiratory tract.

SummaryIn general, the metallographic laboratory is a rea-sonably safe environment. However, dependingon the materials being prepared, dangerous situ-ations can arise. Some hazards, such as thepreparation of radioactive or reactive metals, arequite obvious, while others are not. In the preced-ing discussion, some of the potential hazards thatcan be encountered are summarized; others un-doubtedly exist that are not covered.

Most accidents can be prevented by simplecommonsense rules. It is best to assume that allmetal dust and all chemicals are hazardous. Inha-

Laboratory Safety

lation of dust and fumes, ingestion, or bodily con-tact should be avoided. Personal protectiveequipment is very useful, but it should not be usedas a substitute for good laboratory procedures.The use of such equipment does not guaranteefreedom from injury.

The metallographic literature contains many ref-erences to the use of dangerous materials, oftenwithout any mention of the dangers involved orsafe handling procedures. This is unfortunate be-cause the unwary may be injured. Many of us aretempted to experiment when a recommended pro-cedure does not work as claimed. Thedevelopment of electrolytes, chemical polishingagents, or etchants should be left to those whoare fully versed in the potential dangers. Metallo-graphic laboratories should have some of thereferenced safety publications (see bibliography)readily available in the laboratory and these safetypublications should be consulted when workingwith new or infrequently used materials.

HELPFUL HINTS FOR

LABORATORY SAFETY

Purchase relatively smallquantities of chemicals that

are used routinely, especially those that havea short shelf life, or are dangerous.

After pouring acids, carefully replace the cap,wash the exterior of the bottle under runningwater, and dry off the surface with a papertowel before replacing the acid bottle in itsfireproof cabinet.

Schedule regular time periods for laboratoryclean-ups. These should be weekly for highvolume laboratories. A clean lab is an es-sential ingredient in all laboratory safetyprograms.

112

SUMMARY

Preparation of metallographic specimens is basedupon scientific principles that are easily under-stood. Sectioning creates damage that must beremoved by the grinding and polishing steps if thetrue structure is to be examined. Each sectioningprocess exhibits a certain amount of damage, ther-mal and/or mechanical. Consequently, choose aprocedure that produces the least possible dam-age and use the correct wheel or blade. Grindingalso causes damage, with the depth of damagedecreasing as the abrasive size decreases. Ma-terials respond differently to the same size

abrasive, so one cannot generalize on removaldepths. Removal rates also decrease as the abra-sive size decreases. Through an understandingof these factors, good, reproducible preparationprocedures, such as presented in this guide, canbe established for the materials being prepared.Automation in specimen preparation offers muchmore than reduced labor. Specimens properly pre-pared using automated devices consistentlyexhibit much better flatness, edge retention, reliefcontrol and freedom from artifacts such asscratches, pull out, smearing and comet tailing,than specimens prepared manually.

Summary

113

REFERENCES

1. Metallography and Microstructures,Vol. 9, Metals Handbook, 9th ed., AmericanSociety for Metals, Metals Park, OH, 1985.

2. G. F. Vander Voort, Metallography:Principles and Practice, ASM International,Materials Park, OH, 1999.

3. L. E. Samuels, Metallographic Polishingby Mechanical Methods, 3rd Ed., AmericanSociety for Metals, Metals Park, OH, 1982.

4. J. A. Nelson, “Heating of MetallographicSpecimens Mounted in ‘Cold Setting’ Resins”,Practical Metallography, Vol. 7, 1970,pp. 188-191.

5. G. F. Vander Voort, “Trends in SpecimenPreparation,” Advanced Materials &Processes, Vol. 157, February 2000,pp. 45-49.

6. G. F. Vander Voort, “Metallography forEdge Retention in Surface Treated SinteredCarbides and Steels,” Industrial Heating,Vol. 67, March 2000, pp. 85-90.

7. S. L. Hoyt, Men of Metals, ASM, Metals Park,OH, 1979, pp. 99-100.

8. J. Benedict, R. Anderson and S. J. Klepeis,“Recent Developments in the Use of the TripodPolisher for TEM Specimen Preparation,”Specimen Preparation for TransmissionElectron Microscopy of Materials – III,Materials Research Society, Pittsburgh, PA,1992, Vol. 254, pp. 121-140.

9. G. Petzow and V. Carle, MetallographicEtching, 2nd ed., ASM International, MaterialsPark, OH, 1999.

10. G. F. Vander Voort, “Grain Size Measure-ment,” Practical Applications of QuantitativeMetallography, STP 839, American Society forTesting and Materials, Philadelphia, 1984, pp. 85-131.

11. G. F. Vander Voort, “Wetting Agents inMetallography,” Materials Characterization, Vol.35, No. 2, September 1995, pp. 135-137.

12. A. Skidmore and L. Dillinger, “EtchingTechniques for Quantimet Evaluation,”Microstructures, Vol. 2, Aug./Sept. 1971,pp. 23-24.

13. G. F. Vander Voort, “Etching Techniquesfor Image Analysis,” Microstructural Science,Vol. 9, Elsevier North-Holland, NY, 1981,pp. 135-154.

14. G. F. Vander Voort, “Phase Identificationby Selective Etching,” Applied Metallography,Van Nostrand Reinhold Co., NY, 1986,pp. 1-19.

15. E. E. Stansbury, “Potentiostatic Etching,”Applied Metallography, Van Nostrand ReinholdCo., NY, 1986, pp. 21-39.

16. E. Beraha and B. Shpigler, ColorMetallography, American Society for Metals,Metals Park, OH, 1977.

17. G. F. Vander Voort, “Tint Etching,”Metal Progress, Vol. 127, March 1985,pp. 31-33, 36-38, 41.

18. E. Weck and E. Leistner, MetallographicInstructions for Colour Etching byImmersion, Part I: Klemm Colour Etching, Vol.77, Deutscher Verlag für SchweisstechnikGmbH, Düsseldorf, 1982.

19. E. Weck and E. Leistner, MetallographicInstructions for Colour Etching byImmersion, Part II: Beraha Colour Etchantsand Their Different Variants, Vol. 77/II,Deutscher Verlag für Schweisstechnik GmbH,Düsseldorf, 1983.

20. E. Weck and E. Leistner, MetallographicInstructions for Colour Etching byImmersion, Part III: Nonferrous Metals,Cemented Carbides and Ferrous Metals,Nickel Base and Cobalt Base Alloys,Vol. 77/III, Deutscher Verlag fürSchweisstechnik GmbH, Düsseldorf, 1986.

21. P. Skocovsky, Colour Contrast inMetallographic Microscopy, Slovmetal,• ilina, 1993.

22. H. E. Bühler and H. P. Hougardy, Atlas ofInterference Layer Metallography, DeutscheGesellschaft für Metallkunde, Oberursel,Germany, 1980.

23. H. E. Bühler and I. Aydin, “Applications of theInterference Layer Method,” Applied Metallog-raphy, Van Nostrand Reinhold Co., NY, 1986, pp.41-51.

24. G. F. Vander Voort, “Results of an ASTME-4 Round-Robin on the Precision and Bias ofMeasurements of Microindentation HardnessImpressions,” Factors that Affect thePrecision of Mechanical Tests, ASTM STP1025, ASTM, Philadelphia, 1989, pp. 3-39.

References

114

References

25. G. F. Vander Voort, “Operator Errors in theMeasurement of Microindentation Hardness,”Accreditation Practices for Inspections,Tests, and Laboratories, ASTM STP 1057,ASTM, Philadelphia, 1989, pp. 47-77.

26. G. F. Vander Voort and Gabriel M. Lucas,“Microindentation Hardness Testing,”Advanced Materials & Processes, Vol. 154,No. 3, September 1998, pp. 21-25.

27. R. L. Anderson, “Safety in MetallographicLaboratory,” Westinghouse Research Labora-tory Science Paper, No. 65 – 1P30 – METLL –P2, March 29, 1965.

28. G. F. Vander Voort, Metallography: Prin-ciples and Practice, ASM International,Materials Park, OH., 1999, pp. 148-159.

29. R. C. Nester and G. F. Vander Voort, “Safetyin the Metallographic Laboratory,” Standardiza-tion News, Vol. 20, May 1992,pp. 34-39.

30. P. A. Jacquet, “The Safe Use of Perchloric-Acetic Electropolishing Baths,” Met. Finish, Vol.47, 1949, pp. 62-69.

31. S. M. Comas, R. Gonzalez Palacin and D.Vassallo, “Hazards Associated with PerchloricAcid-Butylcellosolve Polishing Solutions,”Metallography, Vol. 7, 1974, pp. 47-57.

32. N. I. Sax, Dangerous Properties ofIndustrial Materials, 5th ed., Van NostrandReinhold Co., Inc., N.Y., 1979.

33. A.E. Dawkins, “Chromic Acid-AceticAnhydride ‘Explosion,’” J. Iron and SteelInstitute, Vol. 182, 1956, p. 388.

Selected Bibliography of Books onLaboratory SafetyASTM E-2014, “Standard Guide on Metallo-graphic Laboratory Safety”.

N. Van Houten, Laboratory Safety Standardsfor Industry, Research, Academe, Sci. Tech.Pubs., Lake Isabella, CA. 1990.

E. Gershey, A. Wilkerson and E. Party,Laboratory Safety in Practice, Van NostrandReinhold, N.Y., 1991.

A. A. Fuscaldo et al. (eds), Laboratory Safety:Theory & Practice, Academic Press, SanDiego, CA. 1980.

S. B. Pal (ed.), Handbook of LaboratoryHealth & Safety Measures, Kluwer Academic,Norwell, MA., 1985.

Norman V. Steere (ed.), Handbook ofLaboratory Safety, 2nd ed., CRC Press,Boca Raton, FL., 1971.

A. Keith Furr (ed.), Handbook of LaboratorySafety, 3rd ed., CRC Press, Boca Raton, FL.,1989.

L. J. Diberardinis et al., Guidelines for Labora-tory Design: Health & SafetyConsiderations, J. Wiley & Sons, N.Y., 1987.

S. R. Rayburn, The Foundations ofLaboratory Safety, Brock-Springer Series inContemporary Bioscience, Springer-Verlag,N.Y., 1989.

C. A. Kelsey and A. F. Gardner (eds.),Radiation Safety for LaboratoryTechnicians, Warren H. Green, Inc.,St. Louis, MO., 1983.

N. I. Sax, Dangerous Properties of IndustrialMaterials, 5th ed., Van Nostrand Reinhold Co.,N.Y., 1979.

Prudent Practices for Handling HazardousChemicals in Laboratories, NationalAcademy Press, Washington, D.C., 1981.

Prudent Practices for Disposal ofChemicals from Laboratories, NationalAcademy Press, Washington, D.C., 1983.

N. Proctor and J. Hughes, Chemical Hazardsin the Workplace, J. B. Lippincott Co., Philadel-phia, 1978.

F. A. Patty (ed.), Industrial Hygiene andToxicology, Volume II – Toxicology, 3rd ed.,Wiley-Interscience, N.Y., 1980.

L. Bretherick, Handbook of Reactive ChemicalHazards, 2nd ed., Butterworths, London, 1979.

R. J. Lewis, Sr. (ed.), Hawley’s CondensedChemical Dictionary, 12th ed., Van NostrandReinhold, New York, 1993.

Useful Web sites on laboratory safety:http://ull.chemistry.uakron.edu

http://www.hhmi.org/research/labsafe/lcss/lcss.html

http://www.cdc.gov/niosh/ipcsneng/nengsyn.html

http://www.state.nj.us/health/eoh/rtkweb/rtkhsfs.htm

115

APPENDICES

Appendices

Atomic Atomic Density Melting CrystalGroup Element Symbol Number Weight g/cc@20°C Point,°C Structure

Alkaline Beryllium Be 4 9.015 1.848 1280 cphEarth, IIA Magnesium Mg 12 24.312 1.74 650 cph

Coinage Copper Cu 29 63.545 8.94 1083 fccMetals, IB Silver Ag 47 107.87 10.49 961.9 fcc

Gold Au 79 197 19.32 1064.4 fcc

Group IIB Zinc Zn 30 65.38 7.1 419.57 cphCadmium Cd 48 112.4 8.65 320.9 cph

Group III Aluminum Al 13 26.98 2.7 660 fcc

Group IVA Tin Sn 50 118.69 7.29 232 bctLead Pb 82 207.19 11.35 327.5 fcc

Group IVB Titanium Ti 22 47.9 4.507 1670 cphZirconium Zr 40 91.22 6.44 1853 cphHafnium Hf 72 178.49 13.21 2225 cph

Group V Antimony Sb 51 121.75 6.79 630.5 rhombohedralBismuth Bi 83 208.98 9.78 271.3 rhombohedral

Group VB Vanadium V 23 50.94 6.1 1895 bccNiobium/ Nb/

Columbium Cb 41 92.91 8.66 2468 bccTantalum Ta 73 180.95 16.63 2996 bcc

Group VIB Chromium Cr 24 52 7.14 1867 bccMolybdenum Mo 42 95.94 9.01 2615 bcc

Tungsten W 74 183.85 19.3 3410 bcc

Group VIIB Manganese Mn 25 54.94 7.3 1244 complex cubicRhenium Re 75 186.22 21.04 3175 cph

Group VIII Iron Fe 26 55.85 7.874 1536 bccCobalt Co 27 58.93 8.832 1495 cphNickel Ni 28 58.71 8.8 1452 fcc

Platinum Ruthenium Ru 44 101.1 12.1 2310 cphMetals, Rhodium Rh 45 102.91 12.44 1963 fccGroup VIII Palladium Pd 46 106.4 12.16 1551 fcc

Osmium Os 76 190.2 22.48 3015-3075 cphIridium Ir 77 192.2 22.42 2410 fcc

Platinum Pt 78 195.09 21.37 1772 fcc

Classification of Metals Using Common Characteristics and the Periodic Table

Transformation Temperatures for Allotropic Metals Listed Above

Element Transformation Temperature (C°/F°)

Beryllium (Be) α↔β 1256/2293

Cobalt (Co) ε↔α 427/801

Hafnium (Hf) α↔β 1742/3168

Iron (Fe) Curie 770/1418α↔γ 912/1674γ↔δ 1394/2541

Manganese (Mn) α↔β 707/1305β↔γ 1088/1990γ↔δ 1139/2082

Nickel (Ni) Curie 358/676

Tin (Sn) α↔β 13.0/55.4

Titanium (Ti) α↔β 883/1621

Zirconium (Zr) α↔β 872/1602

Note: The Curie temperature is the temperature where a magnetic material losses its magnetism upon heating above and regains it uponcooling below.

116

Appendices

Abrasive Cutting Troubleshooting Guides

Problem Possible Cause Suggested Remedy

Burning (bluish discoloration) Overheated specimen Increase coolant flow rate;reduce cutting pressure; select awheel with softer bonding (fasterbreakdown)

Rapid wheel wear Wheel bonding breaks down Select a wheel with harderbonding;

too rapidly reduce cutting pressure

Frequent wheel breakage Uneven coolant distribution; Adjust coolant flow to be even onloose specimen fixturing; both sides of the wheel; clamp theabrupt contact with specimen specimen more tightly; start cut

contact carefully

Resistance to cutting Slow wheel bond breakdown Select a wheel with softer bonding;use a “pulse” cutting mode; usecutter with oscillating motion orwith minimal area of contactcutting ability

Stalled wheel Inadequate cutter capacity; Use cutter with greaterpinched blade due to movement horsepower; tighten the clamp onof specimen one side less than on the other

side; reduce pressure or feed rate;use a cutter with oscillating motionor with minimal area of contactcutting ability

Pointed:Wheel too hard

Chisel:Nonuniform coolant flow

Glazed Surface:Wrong wheel

Normal:For structural or

medium wall tubing

Normal:For solid materials

Normal:For light tubing or

other thin wall section

Abrasive Wheel Troubleshooting Guides

117

Appendices

Compression (Hot) Mounting Troubleshooting Guide

Thermosetting Resins (Epoxies, Diallyl Phthalates, and Phenolics)

Defect Probable Cause Suggested Remedy

Radial splits Specimen cross sectional area Use a larger mold size; decreaseor cracks too large; specimen with the specimen size; bevel sharp

sharp corners corners, if possible

Shrinkage gaps Specimen surfaces dirty; specimen Clean and dry specimens carefully;cooled quickly after polymerization; after polymerization, cool underwrong resin used pressure to near ambient; use

Epomet G or Epomet F resins

Circumferential cracks Resin contained moisture Keep resins dry during storage;keep containers closed when notusing; dry resin by baking at38-49°C (100-120°F)

Bulging or soft mount Inadequate curing Increase polymerization time(polymerization) time and pressure

Mount appears Time at temperature too short; Increase polymerization time,grainy and unfused temperature for polymerization temperature and pressure

too low; molding pressure too low

Thermoplastic Resins (Acrylics)

Defect Probable Cause Suggested Remedy

Cottonball Incomplete polymerization of resin; Use less resin; use longernot enough time at temperaure heating and cooling periods;

use controlled linear cooling

Crazing Relief of internal stresses Cool mount to a lower temperatureupon ejection of mount before ejection; use controlled

linear cooling

Comparison of Six Edge Retention Compression Mounting Compounds

Micrographs showing the as-forged surface of a hardened modified 5130 alloy steel part mounted using a varietyof resins showing different degrees of edge retention. The specimens were polished simultaneously in the sameholder and were etched with 2% nital. The magnification bars are 20 μm long. Best results were obtained withEpoMet, ProbeMet and EpoxiCure resin with the Conductive Filler particles.

PhenoCure phenolic resin EpoMet thermosetting epoxy resin ProbeMet Cu-filled conductiveresin

KonductoMet C-filled conductiveresin

EpoxiCure cast epoxy resin withConductive Filler particles

SamplKwick cast acrylic resinwith Conductive Filler particles

118

Appendices

Name Typical Cure Time Uses and Restrictions

SamplKwick Acrylic Resin 5-8 minutes Least expensive resin; used for mounting printedcircuit boards and polymers; good for x-raydiffraction work (no peak interferences); highshrinkage (not recommended for edge retention)

VariDur Acrylic Resin 10 minutes Inexpensive; (good edge rentention for acrylic);used with many metals when the edge detail isnot important

EpoKwick Epoxy Resin 90 minutes Fast curing epoxy; more shrinkage than otherepoxies; viscosity 250-400 cps at 25 °C; notrecommended for vacuum infiltration of voidsdue to fast curing time; physically adheres tospecimen; gives acceptable edge retention; goodfor most specimens

EpoColor Epoxy Resin 90 minutes Provides color contrast between mount andspecimen that can be helpful when studyingedges; viscosity 400-700 cps at 25 °C; goodfor most specimens; but not recommended forvacuum impregnation

EpoxiCure Epoxy Resin 6 hours General purpose epoxy; viscosity 400-600 cps at25 °C, good adherence to specimen; can be usedto vacuum impregnate voids (viscosity can bereduced by warming to 50 °C); good for heat-sensitive specimens (very low exotherm duringpolymerization)

EpoThin Epoxy Resin 9 hours Lowest viscosity (can be further reduced bywarming resin to 50 °C) and best penetration ofvoids during vacuum impregnation; goodadherence to specimen; good for heat-sensitivespecimens (lowest exotherm duringpolymerization)

Curing time is not a precise quantity as it can be influenced by the nature of the mold (how well, or how poorly, the exothermic heat ofpolymerization is extracted), by the volume of the cast resin used, by the volume and nature of the specimen being encapsulated, by thetemperature of the mix, by room temperature and air circulation and by the age of the resin and hardener (i.e., curing time is longer forproducts beyond their normal shelf life). The values listed are typical for metallic specimens cast in 1.25-inch (30-mm) diameter Sampl-Kupmolds at 21 °C (70 °F) using fresh resin and hardener.

Castable Mounting Resin Selection Guide

Greek Letter Name English Equivalent

Α, α Alpha A, a

Β, β Beta B, b

Γ, γ Gamma G, g

Δ, δ Delta D, d

Ε, ε Epsilon E, e

Ζ, ζ Zeta Z, z

Η,η Eta E, e

Θ, θ Theta Th

Ι, ι Iota I, i

Κ, κ Kappa K, k

Λ, λ Lambda L, l

Μ, μ Mu M, m

Greek Alphabet and its English Equivalents

Greek Letter Name English Equivalent

Ν, ν Nu N, n

Ξ, ξ Xi X, x

Ο, ο Omicron O, o

Π, π Pi P, p

Ρ, ρ Rho R, r

Σ, σ Sigma S, s

Τ, τ Tau T, t

Υ, υ Upsilon U, u

Φ, ϕ Phi Ph

Χ,χ Chi Ch

Ψ, ψ Psi Ps

Ω,ω Omega O, o

119

Appendices

Cast Resin (“Cold” Mounting) Troubleshooting Guide

Defect Probable Cause Suggested Remedy

Non-Curing Expired resin and/or hardener; Measure weights of resin and hardener;Wrong resin – to – hardener ratio after prescribed time, and it has not

hardened, heat the mold, liquid andspecimen to 50 °C (120 °F) to force curing;check age before using resin and hardener

Slow Curing Incomplete mixing; too little hardener Mix thoroughly; measure weights of resinused; curing at too low a temperature and hardener; place in warm area*

Rapid Curing Too much hardener; curing at too high Measure weights of resin and hardener;a temperature; mold size too large mix smaller quantities; increase air(excessive resin volume) circulation; place in a cooler place

Gas Tunnels Excessive hardener used; mold size Measure weights of resin andtoo large (excessive resin volume); hardener; increase air circulation;use of a low thermal conductivity mold use higher thermal conductivity

molds

Mount Stuck in Mold Did not use mold release agent Apply Silicon Mold Release or ReleaseAgent to interior surfaces of mold

Solvent Softening Poor resistance to solvents; Heat the mold, liquid and specimenmount may not be fully cured to 50 °C (120 °F) to force curing

Excessive Shrinkage Polymerization is too fast; too much Use an epoxy rather than an arcylic; use ahardener used; wrong resin selected slow curing epoxy; add Flat Edge Filler

particles; use the correct resin – to –hardener ratio

* Placing in an oven at 120°F (50°C) for up to two hours may accelerate the curing of epoxies

Name Characteristics Applications

MicroPolish 1-, 0.3- and 0.05-μm alumina powder Inexpensive general-purpose abrasivealumina powder for polishing most minerals and metals

MicroPolish II 1-, 0.3- and 0.05-μm deagglomerated Higher quality alumina suitable fordeagglomerated alumina powders and 5-, 1-, 0.3- and polishing most minerals and metals;alumina powder 0.05-μm deagglomerated alumina produces a better surface finish thanand suspensions suspensions standard alumina abrasives of the same

size

MasterPrep 0.05-μm sol-gel alumina suspension Unique seeded-gel alumina producesagglomerate-free better finishes than calcined aluminaseeded-gel alumina abrasives; excellent with minerals, metals,suspension carbides, PCBs. Good for percious metals.

MasterMet 0.06-μm amorphous silica colloidal Chemo-mechanical action producescolloidal silica suspension with a pH of ~10 producing a excellent surfaces for metals, minerals,suspension chemo-mechanical polishing action polymers, PCBs, electronic devices,

etc., especially soft metals, but not forprecious metals

MasterMet 2 0.02-μm non-crystallizing amorphous Non-crystallizing version of colloidal silicanon-crystallizing silica colloidal suspension with a pH yields similar results as Mastermet butcolloidal silica of ~10.5 producing chemo-mechanical without the danger of crystallized silicasuspension polishing action particle contamination

MasterPolish Proprietary viscous blend of high Proprietary blend has little water and canblended alumina purity 0.05-μm alumina and colloidal be used with materials that are sensitiveand colloidal silica with a slightly basic pH (can to water, such as Mg and most of its alloys;silica suspension be diluted with water) good for most Fe-, Ni- or Co-based alloys

and MMCs

MasterPolish 2 Proprietary 0.06-μm abrasive suspension Proprietary blend designed for finalproprietary with a pH of ~10 for chemo-mechanical polishing of ceramic materials, nitrides,chemo-mechanical polishing action carbides; and compositessuspension

Guide to Fine Polishing Abrasives

120

Appendices

A Guide to Equivalent Loads for Different Mount Sizes

The two graphs plot the load applied per specimen, and the resulting pressure, as a function of mold diameterfor both Imperial and metric units. An example of the use of the graphs is shown. Assume that we have apractice developed using 1.25-inch diameter mounts that calls for 5 lbs. load per specimen. We want to use thismethod with mounts that are 1-inch in diameter. How much load should we use? Draw a line vertically at 5 lbs.(shown in green) until it intersects the line for a 1.25-inch diameter mount. Then draw a horizontal line (shownin blue) from the intersection point to the line for 1-inch diameter mounts. Then, draw a vertical line (shown inblue) down to the load axis. This point is about 3 lbs. So, we would use 3 lbs per specimen for 1-inch mounts. Insimilar manner, if we want to use 1.5-inch diameter specimens, the graph would suggest that we use about 7lbs. per specimen (red lines).

121

Appendices

Preparation Troubleshooting Guide

Troubleshooting Definitions

Comet Tails: Dull, broad depressed lines emanating from a hard particle or from a hole in a specimen in a pattern thatresembles the tail of a comet. The hard particles may be nonmetallic inclusions or nitrides. This appears to be a material-specific phenomenon as repeated efforts to produce this defect in randomly chosen specimens fail. Historically, withmanually-prepared specimens, it was claimed to be caused by unidirectional grinding; however, repeated efforts tocreate comet tails with only unidirectional grinding fail. Comet tails have been observed on susceptible specimensprepared with automated machines using only complementary grinding and polishing with specific combinations of headand platen speeds. In a specimen prone to comet tailing, using contra rotation in the final step will enhance the problemcompared to complementary rotation.

Edge Rounding: The edge or edges of a specimen are abraded at a faster rate than the interior leading to a change inflatness at the edge such that the edge cannot be brought into focus along with the interior with the light microscope (thiscondition is more critical as the magnification is raised due to the higher numerical aperture of the objective whichreduces the depth of focus).

Embedding: Hard abrasive particles that become fixed in the surface of the softer specimen. This is a common problemwith the low melting point alloys (e.g., Pb, Sn, Bi, Cd, Zn), aluminum (mainly fine diamond), and precious metals, but hasbeen observed with refractory metals such as titanium.

Pull Out: Removal of second phase particles (either softer or harder than the matrix) during preparation. Pull out may beenhanced if the interface between the particle and matrix is weak or if the particle is particularly brittle.

Relief: Excessive height differences between second-phase particles and the matrix due to differences in grinding andpolishing rates between the matrix and the particles. Soft particles abrade faster than the matrix and are recessed belowthe matrix while the reverse is observed for hard particles.

Scratches: A linear cut along the surface of a specimen due to contact with a correctly oriented abrasive particle; agroove caused by the abrasion process. The width and depth of the cut is a function of the size of the abrasive particle, itsangle to the specimen surface, the applied pressure, and other factors. Scratches become a problem when they areexcessively coarse for the abrasive size used, indicating a problem with the preparation method. Very fine scratches maynot be a problem in routine examination for grain size or inclusions but could be a problem in failure analysis if criticalfine detail is obscured.

Smear: Matrix flow over voids, cracks, or second phase particles that make detection of these features and measurementof their extent difficult or impossible.

Stain: A contamination residue on the surface of a specimen, that may form around second-phase particles due tointeractions between the specimen and abrasives and/or lubricants, or may form on the matrix due to inadequatecleaning or drying or may form due to interactions between the specimen and solvents after preparation or etching.

122

Appendices

Defect Probable Cause Suggested Remedy

Comet Tails Poorly bonded very hard phase in softer Use hard, napless cloths; reduce applied pressure;matrix; or, pores in matrix results in reduce step times; impregnate pores with epoxy orunidirectional grooves emanating from wax; use complementary rotation. In manualparticles or hole. preparation, avoid unidirectional grinding.

Edge Rounding Shrinkage gaps between specimen and Avoid gaps by cooling thermosetting mount undermount are main problem and, when pressure afer polymerization. Use EpoMet resinpresent, avoid napped cloths and long (least prone to shrinkage gaps); of castable resins,polishing times. epoxy is best (adding Flat Edge Filler particles to

epoxy improves edge retention); use central forcerather than individual force; use the ApexHerculesrigid grinding discs; use napless cloths.

Embedding Hard abrasive tends to embed in soft Coat SiC abrasive paper with paraffin waxmetals; finer particles are more likely (candlewax is best, soaps are also effective);to embed than larger particles. beeswax is less effective; reduce applied pressure,

rpm and times; avoid finer SiC abrasive papers;avoid using diamond slurries with fine diamondsizes (for ≤ 3-μm diamond, paste embeds less thanslurries). SiC embedded in soft metals (e.g., Pb) canbe removed by polishing with alumina slurries.

Pull Outs Excessive grinding time on worn SiC Grind no more than 60 seconds on a sheet of SiCpaper; excessive polishing time on paper; use napless cloths; reduce appliednapped cloths; excessive pressure; pressure; use proper degree of lubrication.and inadequate lubrication lead to pullout of second-phase particles.

Relief Use of napped cloths, long polishing times, Use napless cloths, higher pressures and shorterand low pressure create height differences times. If relief is observed using contra rotation inbetween matrix and second-phase particles. last step, repeat the last step using complementary

rotation.

Scratches Contamination of grinding or polishing Maintain clean operating conditions, washingsurfaces by coarser abrasives; pull out hands and equipment between steps; for cracked orof hard particles from matrix or broken porous specimens, use ultrasonic cleaning afterpieces of brittle materials; or each step (especially polishing);execute each stepinadequate grinding or polishing times thoroughly (avoid short cutting methods);

when usinG the Apex and pressures can leave aheavy magnetic disc system, store the discs in aclean scratch pattern. Drawer or cabinet and scrubsurfaces if contamination is believed to haveoccurred.

Smear Flow of softer metals may be caused by Use proper degree of lubrication, lower pressuresinadequate lubrication, excessive and rpms; use a medium napped cloth on final step;pressures or rpms. lightly etch the specimen after final polishing and

repeat the last step; use the vibratory polisher toremove smeared surface metal.

Stain Stains around second-phase particles can Some staining issues are unique to the specimenbe induced by interactions between but most result from using impure tap water (switchabrasive and particle, perhaps affected to distilled water; sometimes de-ionized water isby water quality; or they can occur due needed) or inadequate drying. When using colloidalto inadequate cleaning, or improper silica, which can be hard to remove, direct the watersolvents used after the abrasive step. jet onto the cloth and wash the specimens (and

cloth) for the last 6-10 seconds of the polishingcycle. This makes subsequent cleaning easy.Otherwise, scrub the surface with water containinga detergent solution using cotton. Then, rinse withalcohol and blow dry with hot air, such as with theTorraMet dryer. Compressed air systems cancontain impurities, such as oils, that can stain thesurface.

Preparation Troubleshooting Guide Cont.

123

ASTM Metallography Standards

ASTM METALLOGRAPHYSTANDARDSA 247: Visual Classification of Graphite in theMicrostructure of Cast Iron

A 892: Defining and Rating the Microstructure ofHigh Carbon Bearing Steels

B 390: Evaluating Apparent Grain Size andDistribution of Cemented Tungsten Carbides

B 588: Measurement of the Thickness ofTransparent or Opaque Coatings by Double-Beam Interference Microscope Technique

B 657: Metallographic Determination of Micro-structure in Cemented Tungsten Carbide

B 681: Measurement of Thickness of AnodicCoatings on Aluminum and of Other TransparentCoatings on Opaque Surfaces Using the Light-Section Microscope

B 748: Measurement of the Thickness ofMetallic Coatings by Measurement of CrossSection with a Scanning Electron Microscope

B 795: Determining the Percentage of Alloyed orUnalloyed Iron Contamination Present in PowderForged Steel Parts

B 796: Nonmetallic Inclusion Level of PowderForged Steel Parts

B 847: Measurement of Metal and Oxide CoatingThickness by MicroscopicalExamination of a Cross section

C 664: Thickness of Diffusion Coating

E 3: Preparation of Metallographic Specimens

E 7: Standard Terminology Relating toMetallography

E 45 Determining the Inclusion Contentof Steel

E 82: Determining the Orientation of aMetal Crystal

E 112: Determining Average Grain Size

E 340: Macroetching Metals and Alloys

E 381: Macroetch Testing Steel Bars, Billets,Blooms, and Forgings

E 384: Microindentation Hardness of Materials

E 407: Microetching Metals and Alloys

E 562: Determining Volume Fraction by System-atic Manual Point Count

E 766: Calibrating the Magnification of aScanning Electron Microscope

E 768: Preparing and Evaluating Specimens forAutomatic Inclusion Assessment of Steel

E 807: Metallographic Laboratory Evaluation

E 883: Reflected-Light Photomicrography

E 930: Estimating the Largest Grain Observed ina Metallographic Section (ALA Grain Size)

E 975: X-Ray Determination of RetainedAustenite in Steel with Near RandomCrystallographic Orientation

E 986: Scanning Electron Microscope BeamSize Characterization

E 1077: Estimating the Depth ofDecarburization of Steel Specimens

E 1122: Obtaining JK Inclusion Ratings UsingAutomatic Image Analysis

E 1180: Preparing Sulfur Prints forMacrostructural Examination

E 1181: Characterizing Duplex Grain Sizes

E 1245: Determining the Inclusion or Second-Phase Constituent Content of Metals byAutomatic Image Analysis

E 1268: Assessing the Degree of Banding orOrientation of Microstructures

E 1351: Production and Evaluation of FieldMetallographic Replicas

E 1382: Determining Average Grain Size UsingSemiautomatic and Automatic Image Analysis

E 1508: Quantitative Analysis by Energy-Dispersive Spectroscopy

E 1558: Electrolytic Polishing of MetallographicSpecimens

E 1920: Metallographic Preparation of ThermalSpray Coatings

E 1951: Calibrating Reticles and LightMicroscope Magnifications

E 2014: Metallographic Laboratory Safety

E 2015: Preparation of Plastics and PolymericSpecimens for Microstructural Examination

F 1854: Stereological Evaluation of PorousCoatings on Medical Implants

124

ASTM/ISO Standards

ASTM HARDNESS TESTINGSTANDARDS

B 578: Microhardness of ElectroplatedCoatings

B 721: Microhardness and Case Depth ofPowder Metallurgy Parts

C 730: Knoop Indentation Hardness of Glass

C 849: Knoop Indentation Hardness ofCeramic Whitewares

C 1326: Knoop Indentation Hardness of Ad-vanced Ceramics

C 1327: Vickers Indentation Hardness ofAdvanced Ceramics

E 10: Brinell Hardness of Metallic Materials

E 18: Rockwell Hardness and RockwellSuperficial Hardness of Metallic Materials

E 92: Vickers Hardness of Metallic Materials

E 140: Standard Hardness Conversion Tablesfor Metals

E 448: Scleroscope Hardness Testing of MetallicMaterials

E 1842: Macro-Rockwell Hardness Testing ofMetallic Materials

ISO METALLOGRAPHYSTANDARDS

ISO 643: Steels – Micrographic Determination ofthe Ferritic or Austenitic Grain Size

ISO 945: Cast Iron: Designation ofMicrostructure of Graphite

ISO 1083: Spheroidal Graphite CastIron – Classification

ISO 1463: Metallic and Oxide Coatings –Measurement of Coating Thickness –Microscopical Method

ISO 2624: Copper and Copper Alloys – Estima-tion of Average Grain Size

ISO 2639: Steel – Determination andVerification of the Effective Depth ofCarburized and Hardened Cases

ISO 3754: Steel – Determination of EffectiveDepth of Hardening After Flame or InductionHardening

ISO 3763: Wrought Steels – MacroscopicMethods for Assessing the Content ofNon-Metallic Inclusions

ISO 3887: Steel, Non-Alloy and Low Alloy –Determination of Depth of Decarburization

ISO 4499: Hardmetals: MetallographicDetermination of Microstructure

ISO 4524-1: Metallic Coatings – TestMethods for Electrodeposited Gold and GoldAlloy Coatings – Part 1: Determination ofCoating Thickness

ISO 4964: Steel – Hardness Conversions

ISO 4967: Steel – Determination of Content ofNon-Metallic Inclusions – Micrographic MethodUsing Standard Diagrams

ISO 4968: Steel – Macrographic Examination bySulphur Print (Baumann Method)

ISO 4970: Steel – Determination of Totalor Effective Thickness of Thin Surface-Hard-ened Layers

ISO 5949: Tool Steels and Bearing Steels –Micrographic Method for Assessing the Distribu-tion of Carbides Using ReferencePhotomicrographs

ISO 9042: Steels – Manual Point CountingMethod for Statistically Estimating the VolumeFraction of a Constituent with a Point Grid

ISO 9220: Metallic Coatings – Measurement ofCoating Thickness – Scanning ElectronMicroscope Method

ISO 14250: Steel – MetallographicCharacterization of Duplex Grain Sizeand Distribution

125

ISO/Other Nattional Standards

ISO HARDNESS STANDARDS

ISO 1355: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Rockwell Superficial Hardness TestingMachines (Scales 15N, 30N, 45N, 15T, 30T and45T)

ISO 4516: Metallic and Related Coatings –Vickers and Knoop Microhardness Tests

ISO 4545: Metallic Materials – Hardness Test –Knoop Test

ISO 4546: Metallic Materials – HardnessTest – Verification of Knoop HardnessTesting Machines

ISO 4547: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Knoop Hardness Testing Machines

ISO 6441-1: Paints and Varnishes – Determina-tion of Microindentation Harness – Part 1: KnoopHardness by Measurement ofIndentation Length

ISO 6441-1: Paints and Varnishes – Determina-tion of Microindentation Harness – Part 2: KnoopHardness by Measurement of Indentation DepthUnder Load

ISO 6507-1: Metallic Materials – Hardness Test– Vickers Test – Part I: HV 5 to HV 100

ISO 6507-2: Metallic Materials – HardnessTest – Vickers Test – Part 2: HV 0, 2 to LessThan HV 5

ISO 6508: Metallic Materials – Hardness Test –Rockwell Test (Scales A, B, C, D, E, F, G, H, K)

ISO 9385: Glass and Glass-Ceramics – KnoopHardness Test

ISO 10250: Metallic Materials – HardnessTesting – Tables of Knoop Hardness Values forUse in Tests Made on Flat Surfaces

ISO 14271: Vickers Hardness Testing ofResistance Spot, Projection and Seam Welds(low load and microhardness)

OTHER NATIONAL STANDARDS

FranceNF A04-10: Determination of the Ferritic andAustenitic Grain Size of Steels

GermanyDIN 50150: Testing of Steel and Cast Steel;Conversion Table for Vickers Hardness,Brinell Hardness, Rockwell Hardness andTensile Strength

DIN 50192: Determination of the Depth ofDecarburization

DIN 50600: Testing of Metallic Materials;Metallographic Micrographs; Picture Scalesand Formats

DIN 50601: Metallographic Examination;Determination of the Ferritic or AusteniticGrain Size

DIN 50602: Metallographic Examination;Microscopic Examination of Special SteelsUsing Standard Diagrams to Assess theContent of Non-Metallic Inclusions

SEP 1510: Microscopic Test of Steels for GrainSize by Comparison with Standard Charts

SEP 1570: Microscopical Examination of SpecialSteels for Non-Metallic Inclusions UsingStandard Micrograph Charts

SEP 1572: Microscopic Testing of Free-CuttingSteels for Non-Metallic Sulphide Inclusions byMeans of a Series of Pictures

ItalyUNI 3137: Extraction and Preparation ofSamples

UNI 3138: Macrographic Analysis

UNI 3245: Microscopic Examination of FerrousMaterials - Determination of Austenitic or FerriticGrain Size of Plain Carbon and Low-Alloy Steels

UNI 4227: Determination of MetallographicStructures

UNI 4389: Nonferrous Metals and Their Alloys:Determination of the Dimension of Their CrystalGrains

126

Other National Standards

JapanJIS B 7724: Brinell Hardness – Verification ofTesting Machine

JIS B 7725: Vickers Hardness – Verification ofTesting Machines

JIS B 7730: Rockwell Hardness Test – Calibra-tion of Standardized Blocks

JIS B 7734: Knoop Hardness Test –Verification of Testing Machines

JIS B 7735: Vickers Hardness Test –Calibration of the Reference Blocks

JIS B 7736: Brinell Hardness Test – Calibrationof Standardized Blocks

JIS G 0551: Methods of Austenite Grain SizeTest for Steel

JIS G 0552: Method of Ferrite Grain Size Testfor Steel

JIS G 0553: Macrostructure Detecting Methodfor Steel, Edition 1

JIS H 0501: Methods for EstimatingAverage Grain Size of Wrought Copper andCopper Alloys

JIS R 1610: Testing Method for Vickers Hard-ness of High Performance Ceramics

JIS R 1623: Test Method for Vickers Hardnessof Fine Ceramics at Elevated Temperatures

JIS Z 2243: Brinell Hardness Test – Test Method

JIS Z 2244: Vickers Hardness Test –Test Method

JIS Z 2245: Rockwell Hardness Test –Test Method

JIS Z 2251: Knoop Hardness Test –Test Method

JIS Z 2252: Test Methods for Vickers Hardnessat Elevated Temperatures

PolandPN-57/H-04501: Macroscopic Examination ofSteel. The Deep Etching Test.

PN-61/H-04502: Reagents for MacrostructureTests of Iron Alloys

PN-61/H-04503: Reagents for MicrostructureTests of Iron Alloys

PN-63/H-04504: Microstructure Testing of SteelProducts. Iron Carbide. Ghost Structure.Widmanstätten’s Structure.

PN-66/H-04505: Microstructure of SteelProducts. Templates and Evaluation.

PN-75/H-04512: Nonferrous Metals. Reagentsfor Revealing Microstructure.

PN-75/H-04661: Gray, Spheroidal Graphite andMalleable Cast Iron. Metallographic Examination.Evaluation of Microstructure

PN-76/H-04660: Cast Steel and Iron. Micro-scopic Examination. Sampling and Preparationof Test Pieces.

PN-84/H-04507/01: Metals. MetallographicTesting of Grain Size. Microscopic Methods forDetermination of Grain Size.

PN-84/H-04507/02: Metals. MetallographicTesting of Grain Size. Methods of Revealing thePrior-Austenitic Grains in Non-Austenitic Steels.

PN-84/H-04507/03: Metals. MetallographicTesting of Grain Size. Macroscopic Method ofRevealing the Prior-Austenitic Grain Size by theFracture Method.

PN-84/H-04507/04: Metals. MetallographicTesting of Grain Size. A Method of Testing forOverheating of Steel.

PN-87/H-04514: Steel, Cast Steel, Cast Iron.Macrostructure Examination. Baumann’s Test.

RussiaGOST 801: Standard for Ball and RollerBearing Steel

GOST 1778: Metallographic Methods ofDetermination of Nonmetallic Inclusions

GOST 5639: Grain Size Determination

SwedenSIS 11 11 01: Estimating the Average Grain Sizeof Metals

SIS 11 11 02: Estimating the Austenitic GrainSize of Ferritic and Martensitic Steels

SIS 11 11 11: Methods for Assessing theSlag Inclusion Content in Steel:Microscopic Methods

127

Other National Standards

Sweden (continued)SIS 11 11 14: Determination of SlagInclusions – Microscopic Methods –Manual Representation

SIS 11 11 16: Steel – Method for Estimation ofthe Content of Non-Metallic Inclusions –Microscopic Methods – Jernkontoret’sInclusion Chart II for the Assessment ofNon-Metallic Inclusions

SIS 11 03 40: Hardness Test – Vickers TestHV 0, 2 to HV 100 – Direct Verification ofTesting Machines

SIS 11 03 41: Hardness Test – Vickers Test HV0,2 to HV 100 – Indirect Verification of TestingMachines Using Standardized Blocks

SIS 11 03 42: Hardness Test – Vickers TestHV 0,2 to HV 100 – Direct Verification ofStandardizing Machine for Calibration ofStandardized Blocks

SIS 11 03 43: Hardness Test – VickersTest HV 0,2 to HV 100 – Calibration ofStandardized Blocks

SIS 11 25 16: Metallic Materials – Hardness Test– Vickers Test HV 5 to HV 100

SIS 11 25 17: Metallic Materials – Hardness Test– Vickers Test HV 0,2 to Less Than HV 5

SIS 11 70 20: Determination of the Depth ofDecarburization in Steel

United KingdomBS 860: Tables for Comparison of HardnessScales.

BS 4490: Methods for Micrographic Determina-tion of the Grain Size of Steel

BS 5710: Macroscopic Assessment of the Non-Metallic Inclusion Content of Wrought Steels

BS 6285: Macroscopic Assessment of Steel bySulphur Print

BS 6286: Measurement of Total or Effective

Thickness of Thin Surface-Hardened Layers inSteel

BS 6479: Determination and Verification ofEffective Depth of Carburized and HardenedCases in Steel

BS 6481: Determination of Effective Depth ofHardening of Steel after Flame or InductionHardening

BS 6533: Macroscopic Examination of Steel byEtching with Strong Mineral Acids

BS 6617: Determination of Decarburisation inSteel

128

REGISTERED TRADEMARKS OFBUEHLER LTD.:

Apex

ApexHercules

Buehler

CarbiMet

ChemoMet

DuoMet

EdgeMet Kit

EpoKwick

EpoMet

EpoThin

EpoxiCure

FluorMet

HandiMet

IsoCut

IsoMet

KonductoMet

MasterMet

MasterPolish

MasterPrep

MasterTex

MetaDi

MetaDi Supreme

MicroCloth

MicroPolish

NanoCloth

NelsonZimmer

OmniMet

OscilliMet

PlanarMet

ProbeMet

SamplKup

SamplKwick

SimpliMet

SuperMet

SurfMet

TexMet

TorraMet

VibroMet

UNREGISTERED TRADEMARKSOF BUEHLER LTD.:

AcuThin

Delta

EpoColor

PhenoCure

Phoenix Beta

Phoenix Vector

TransOptic

TriDent

UltraPad

UltraPlan

UltraPol

UltraPrep

WhiteFelt

Buehler Trademarks

129

Index

AAbrasive cut-off machines:

chop cutters, 10

minimum area of contact cutting, 11

orbital cutting, 11-12

oscillating cutting, 11

pulse cutting mode, 10-11

traverse-and-increment cutting, 11-12

Abrasive cut-off wheels:

bond strength, 9

recommendations, 10

thickness, 9

troubleshooting guide, 116

Abrasive sectioning, 7, 9-12, 116

Acids, 107-110

Acrylic resins, 15-17, 117-119

Alumina, 20, 27, 28, 35, 40

Alumina grinding paper, 20, 29

Aluminum and alloys:

etchants, 70, 71, 73

preparation, 34-35

ASTM Standards, 123-124

Attack polishing:

Au, 52

Be, 37

Co, 50

Hf, 41-42

Nb, V and Ta, 43-44

safety, 41

Ti, 41

Zr, 41-42

BBases, 110

Beryllium

preparation, 37

Bright field illumination, 83,84,86

Buehler trademarks, 130

CCastable resins:

acrylic resins, 15-17, 118, 119

Conductive Filler particles, 16

epoxy resins, 15-17, 118, 119

exotherm, 16-17

Flat Edge Filler particles, 16,18,19

recommendations, 16, 148

troubleshooting guide, 119

Central force, 19, 29

Ceramics:

etchants, 77

preparation, 57-60

Chemical polish, 41-42

Clamp mounting, 14

Cobalt and alloys:

etchants, 76

preparation, 50

Cold mounting (see castable resins)

Colloidal silica, 27-28, 34, 120

“Comet tails,” 23, 121-122

Composites:

preparation, 59-60

Compression mounting, 14-15, 17-19, 53, 117

Conductive mounts, 16, 117

Contra rotation, 29, 30-31

Coolant flow, 9

Copper and alloys:

etchants, 70, 75-76

preparation, 48

Cutting damage, 7-9, 116

DDamage, sectioning, 7-9, 116

Dark field illumination, 65, 85, 86

Diamond, 27, 28

Differential interference contrast, 23, 43, 44, 59,65, 85, 86

Digital imaging:

acquisition, 93

archiving, 93

calibration, 95-96

cameras, 93-94

clarification, 96

color, 94

pixels, 94, 95, 96

resolution, 95

EEdge preservation, 14, 15, 17-19, 121-122

Electronic materials:

preparation, 62-64

Embedded abrasive, 20, 21, 35, 39, 64, 121-122

Epoxy resin, 15-17, 117-119

130

Index

E (CONT.)Etchants, 27, 73-77

chemicals, 110-111

dangerous solutions, 105, 107, 108, 109

mixing, 103-104

storage, 103

Etching, 67-77, 78

anodizing, 71

electrolytic, 71

heat tinting, 71-72

interference layer method, 72

procedure, 67

selective etching, 68-71

FFerrous metals:

etchants, 68-70, 74-75

preparation, 31, 45-47

Fluorescent dyes, 16, 53

GGrain size measurement, 7, 99-100

Grinding:

alumina paper, 20

diamond, 19, 20

embedding, 20,21

emery paper, 20

machines, 22

planar, 20

SiC paper, 20

HHafnium and alloys:

etchants, 74

preparation, 41-42

IImage analysis

binary operations, 98-99

feature-specific measurements, 99

field measurements, 99

measurements, 96, 99-100

thresholding, 96, 99-100

Inclusion measurement, 7

Individual force, 19

Iron and steel:

etchants, 68-70, 74-75

preparation, 31, 45-47

ISO standards, 124-125

K-LKnoop hardness, 88-89

Light microscopy, 78-86

Low-melting point metals:

etchants, 73

preparation, 37-38, 62-64

MMagnesium and alloys:

etchants, 73

preparation, 36

Magnesium oxide, 34

Metals:

allotropic transformations, 116

periodic table, 33

properties, 116

Method Development, 7

Microindentation hardness:

accuracy, precision, bias, 89-92

automation, 91-92

indentation size effect, 88

Knoop test, 88-89

Vickers test, 84-88

Microscope, reflected light:

condenser, 80

depth of field, 83-84

diaphragms, 80

“empty” magnification, 82-83

eye clearance, 81

eye pieces, 81

filters, 80

graticules, 81

illumination modes, 79, 83-86

Köhler illumination, 83-84

light sources, 77, 79

magnification, 81, 82

numerical aperture, 80, 82, 83

objective lens, 77, 79, 80

ray diagram, 78

resolution, 82, 83

reticles, 81

stage, 78, 81-89

stand, 82

131

Index

M (CONT.)Microscope, reflected light (cont):

working distance, 80, 81

Mounting:

castable resins, 14, 15-17

clamp mounting, 14

compression mounting, 14-15, 17-19, 117

conductive resins, 16, 117

edge retention (preservation), 14, 15, 17-19,117, 121-122

electroless nickel plating, 17, 19

low-melting point metals, 38

phenolic resins, 14

presses, 17, 19

purposes, 14

recommendations, 14, 16

shrinkage gaps, 14, 15, 17, 18, 19

thermoplastic resins, 14-15, 18, 117

thermosetting resins, 14-15, 17-19, 117

troubleshooting guides, 117

Mount size, 21

NNickel and alloys:

etchants, 76

preparation, 49

Nomarski DIC, 19, 43, 44, 59, 65, 85, 86

O-POblique illumination, 84-85

Other National Standards, 125-127

Periodic table of the elements, 33

Polarized light, 9, 35, 36, 37, 38, 40, 41, 42, 45,48, 50, 59, 67, 71, 85, 86

Polishing:

abrasives, 27-28, 120

automated, 25

cloths, 19, 24, 25-26, 28

electrolytic, 24

head position, 19, 32

lubricants, 27

manual, 24

recommendations, 120

vibratory, 27, 35-36, 37, 39, 41

Polymers:

etchants, 77

preparation

(see also: castable resins)

Precision saws, 12-13

Precious metals:

etchants, 76-77

Preparation, 51-52

Preparation method development: 6, 8, 29, 31, 32

Preparation problems:

“comet tails,” 23, 121-122

disturbed metal, 23

edge retention (preservation), 14, 15, 17-19,117, 121-122

embedded abrasive, 20, 21, 35, 39, 64,121-122

excessive relief, 23, 24, 121-122

pitting, 23

pull outs, 9, 121-122

scratches, 121-122

smear, 63, 64, 121-122

troubleshooting guide, 121-122

water sensitivity, 36, 37

Preparation procedures

contemporary methods, 29-31

Al and alloys, 34-35

Be, 37

ceramics, 57-58

Co and alloys, 50

composites, 59-60

Cu and alloys, 48

electronic devices, 62-64

Fe and alloys, 45-47

Hf and alloys, 41-42

low-melting point alloys, 38-38, 62-64

Mg and alloys, 36

Ni and alloys, 49

polymers, 65-66

precious metals, 51-52

printed circuit boards, 61

Refractory metals, 40-44

Sintered carbides, 55-56

Ti and alloys, 40-41

TSC and TBC coated metals, 53-54

Zr and alloys, 41-42

132

Index

O-P (CONT.)specific procedures, 32-66

development, 32

load, 32

platen size, 32

time, 32

“traditional” method, 29

Printed circuit boards:

etchant, 61

preparation, 61

Pull outs, 7, 121-122

RRefractory metals:

etchants, 74

preparation, 40-44

Relief, 8, 23, 24

Rigid grinding discs, 19, 30-31, 45-47, 49-50,53-54, 55, 57, 63

SSafety:

acids, 107

bases, 110

chemical storage, 103

dust, 102

equipment operation, 102

fume hoods, 101

MSDS, 101, 103

personal protective equipment, 102-103

publications, 114

safety plan, 101

solvents, 104, 106

web sites, 114

Sampling, 7

Sectioning, 9-13

Sectioning damage, 7, 8, 9

Sectioning planes, 7

Shrinkage gaps, 14, 15, 17, 18, 19, 122

Silicon

preparation, 62-63

Silicon carbide paper, 20-22

ANSI/CAMI grit size, 21-22

FEPA grit size, 21-22

Silicon dioxide, 27

Sintered carbides:

etchants, 70-71, 77

preparation, 55-56

Smartcut, 12

Solvents, 104, 106

Sorby, Sir Henry Clifton, 81

Specimen preparation:

goals, 6, 10

Methods

Contemporary, 29-31

Specific materials, 32-66

traditional, 29

(see: preparation procedures)

Problems, 7, 121-122

(see: preparation problems)

Stains, staining, 17, 18, 23, 121-122

Standards, metallography and hardness,123-127

Steels (see: Fe and alloys or Irons and steels)

TThermally spray-coated metals

preparation, 53-54

Thermoplastic resins, 14-15, 18

Thermosetting resins, 14-15, 17-19

Titanium and alloys:

etchants, 74

heat tinting, 72

preparation, 40-41

Trademarks, Buehler Ltd., 128

U-ZUltrasonic cleaning, 28

Vacuum impregnation, 15-16, 53, 57, 59

Vickers hardness, 87-88

Water-sensitive materials, 27, 36, 37, 120

Worldwide Sales Offices, 132-135

Zirconium and alloys:

etchants, 74

preparation, 41-42

133

Worldwide Sales Offices

BUEHLER HEADQUARTERS41 Waukegan Road, P. O. Box 1Lake Bluff, Illinois 60044-1699Tel: (847) 295-6500 • Fax: (847) 295-7979Sales: 1-800-BUEHLER (800-283-4537)

INTERNATIONALArgentina - Buenos AiresARO S.A.Tel: (54) (11) 4331.5766Fax: (54) (11) 4331.3572

Australia - VictoriaBiolab Industrial TechnologiesTel: (61) (3) 9263.4300Fax: (61) (3) 9562.7953

Austria - WolfsgrabenAndreas GrimasTel: (43) 2233.78610Fax: (43) 2233.78619

Bahrain-ManamaYousuf Mahmood Husain, WLLTel: (973) 276.176Fax: (973) 275.819

Bangladesh - DhakaA.Q. Chowdhury Science & SynergyTel: (880) (2) 911.2290Fax: (880) (2) 988.0790

Belgium - NamurAnalis S.A.Tel: (32) (81) 255050Fax: (32) (81) 230779

Brazil - Sao PauloInstrumental Instrumentos de Medicao LTDATel: (55) (11) 5011.0901Fax: (55) (11) 5012.4650

Bulgaria - SofiaET KBR-Ivan MerchandjiiskiTel: (359) (2) 793035Fax:(359) (2) 793626

Canada - OntarioBuehler CanadaTel: (905) 201.4686Fax: (905) 201.4683

Chile - Vina del MarG. Busch & Cia. Ltda.Tel: (56) (32) 692.064Fax: (56) (32) 692.064

ChinaHong Kong - ChinaBuehler Asia/PacificTel: (852) 2307.0909Fax: (852) 2307.0233

Buehler - BeijingTel: (86) (10) 6592-5321Fax: (86) (10) 6592-5320

Buehler - ChengDuTel: (86) (28) 662.8833Fax: (86) (28) 662.5844

Buehler - ShanghaiTel: (86) (21) 6391.2588Fax: (86) (21) 6391.2277Buehler - WuhanTel: (86) (27) 8360.2980Fax: (86) (27) 8360.2779Buehler - XianTel: (86) (29) 766.0608Fax: (86) (29) 766.0603

Hong Kong - Unicon Scientific Co. Ltd.Tel: (852) 2350.6323Fax: (852) 2352.3335

Colombia - SantaFe de BogotaArotec Colombiana S.A.Tel: (57) (1) 288.7799Fax: (57) (1) 285.3604

Costa Rica - San JoseCapris S.A.Tel: (506) 232.9111Fax: (506) 232.8525

Croatia - ZagrebIndustrijska Defektoskopija (IDEF)Tel: (385) (1) 364.6637Fax: (385) (1) 364.6635

Cyprus - NicosiaPNC Prospects Ltd.Tel: (357) (2) 467200Fax: (357) (2) 367831

Czech Republic - PrahaHanyko Praha S.R.O.Tel: (420) (2) 5718.7612Fax: (420) (2) 5718.7610

Egypt - CairoScientific Equipment Trading Institution (SETI)Tel: (20) (2) 574.9198Fax: (20) (2) 575.5662

Estonia - TallinnTallinn Technical UniversityTel: (372) 6203.150Fax: (372) 6203.153

Finland - HelsinkiKnorring OY A.B.Tel: (358) (9) 56041Fax: (358) (9) 565.2463

France - DardillyBuehler S.A.R.L.Tel: (33) (4) 37 59 81 20Fax: (33) (4) 37 59 81 29

Germany - DüsseldorfBuehler GmbHTel: (49) (211) 974100Fax: (49) (211) 974.1079

134

Worldwide Sales Offices

Greece - KoropiB. Caravitis S.A.Tel: (30) (10) 602.8317Fax: (30) (10) 664.4886

Hungary - BudapestGrimas HmbHTel: (36) (1) 420.5883Fax: (36) (1) 276.0557

India - AgraCensico International Private, Ltd.Tel: (91) (562) 524777Fax: (91) (562) 520222

Iran - TehranA B Sangari & SonsTel: (98) (21) 8964958Fax: (98) (21) 8959396

Ireland - LucanLabquip Ltd.Tel: (353) (1) 624.1668Fax: (353) (1) 624.0107

Israel - GivataimMicrotechTel: (972) (3) 573.1122Fax: (972) (3) 573.1123

Italy – FirenzeNikon Instruments, S.p.a.Tel: (39) 055.300920Fax: (39) 005.300993

Japan - TokyoSankei Co., Ltd.Tel: (81) (3) 5805.0515Fax: (81) (3) 5805.0525

OsakaTel: (81) (6) 6327.3850Fax: (81) (6) 6327.1993Shizuoka Pref.Tel: (81) (54) 287.6722Fax: (81) (54) 287.6907TsukubaTel: (81) (298) 52.3061Fax: (81) (298) 52.3063

YokohamaTel: (81) (467) 41.1221Fax: (81) (467) 48.3610

Jordan - AmmanSMS Scientific & Medical Supplies Co.Tel: (962) (6) 462.4907Fax: (962) (6) 462.8258

Kuwait - SafatAl-Sedan Trading & Contracting Co.Tel: (965) 242.0469Fax: (965) 240.0111

Lithuania - VilniusNovatex BaltijaTel: (370) (2) 73.7292Fax: (370) (2) 73.7296

Malaysia - Selangor Darul EhsanCLMO Technology SDN. BHD.Tel: (60) (3) 8942.1238Fax: (60) (3) 8948.1661

Malta - MarsaAttard & Co., Ltd.Tel: (356) (21) 237555Fax: (356) (21) 220186

Mexico - Mexico D.F.Microanalisis S.A. de C.V.Tel: (52) (55) 5660.9828Fax: (52) (55) 5660.9831

Morocco - CasablancaEstablissements BoyerTel: (212) 227.1571Fax: (212) 226.1149

Netherlands - ZoeterwoudePaes Nederland BVTel: (31) (71) 545.0850Fax: (31) (71) 545.0802

New Zealand - AucklandBiolab Ltd.Tel: (64) (9) 980.6700Fax: (64) (9) 980.6788

Norway - OsloNerliens Kemiske-Tekniske ASTel: (47) (22) 666500Fax: (47) (22) 666501

Pakistan – KarachiFaco Trading(92) (21) 452.3096Fax: (92) (21) 452.6898

Peru - LimaCimatec S.A.Tel: (51) (1) 336.5151Fax: (51) (1) 336.5279

Philippines - ManilaEdward Keller (Philippines) Inc.Tel: (63) (2) 864.1600Fax: (63) (2) 864.1698

Poland - KatowiceTestingTel: (48) (32) 757.4597Fax: (48) (32) 757.4815

Portugal - Linda-a-VelhaLaboControleTel: (351) (21) 419.7945Fax: (351) (21) 415.1430

135

Worldwide Sales Offices

Qatar - DohaAL-Obeidly Trading TechnologyTel: (974) 469.4111Fax: (974) 469.4449

Romania - BucarestTotal Control S.R.L.Tel: (40) (21) 619.3054Fax: (40) (21) 212.3151Russian - St. PetersburgRVS Company Ltd.Tel: (7) (812) 186.9516Fax: (7) (812) 252.0136Saudi Arabia - JeddahK.I. Abdulkadir & PartnersTel: (966) (2) 640.5846Fax: (966) (2) 640.4710

Republic of Singapore - SingaporeMicroactive Scientific Pvt. Ltd.Tel: (65) (6) 786.3386Fax: (65) (6) 786.3183Slovakia - BratislavaMITARTel: (421) (7) 6428.8474

Slovenia - SI-LjubljanaSI-MET D.N.O.Tel: (386) (12) 322.094Fax: (386) (12) 322.094South Africa - JohannesburgMicro Met Scientific Co.Tel: (27) (11) 886.3345Fax: (27) (11) 886.5204South Korea - Kyung-KiKey One Engineering Co., Ltd.Tel: (82) (31) 706.1128Fax: (82) (31) 706.1150

Spain - OviedoBiometa Tecnologia y Systemas, S.A.L.Tel: (34) (902) 244343Fax: (34) (985) 269169Sultanate of OmanBusiness International Group LLCTel: (968) 771.4752Fax: (968) 771.3924

Sweden - SundbybergMikron ABTel: (46) (8) 564.82140Fax: (46) (8) 564.82141

Switzerland - DietikonPrufmaschinen AG PrufagTel: (41) 01746.4030Fax: (46) 01746.4039Taiwan - TaipeiSanpany Instruments Co. Ltd.Tel: (886) (2) 2392.3433Fax: (886) (2) 2392.3463Thailand - BangkokJebsen & Jessen Marketing (T) Ltd.Tel: (66) (2) 714.3939Fax: (66) (2) 714.3900

Turkey - IstanbulMegaTel: (90) (216) 326.4535Fax: (90) (216) 326.3286United Arab Emirates - DubaiTamra ElectrocomTel: (971) (4) 223.3259Fax: (971) (4) 228.3405United Kingdom - CoventryBuehler UKTel: (44) (0) 24 7658.2158Fax: (44) (0) 24 7658.2159

Uruguay - MontevideoGustavo E. AriasTel: (598) (2) 600.3703Fax: (598) (2) 600.3776Venezuela - CaracasDequimem C.A.Tel: (58) (212) 242.7512Fax: (58) (212) 241.6354Yugoslavia & Mecedonia - BelgradeTrimid POTel: (381) (11) 444.4149Fax: (381) (11) 456870

136

In today’s competitiveIn today’s competitiveIn today’s competitiveIn today’s competitiveIn today’s competitive

world, material scientistsworld, material scientistsworld, material scientistsworld, material scientistsworld, material scientists

and metallographers needand metallographers needand metallographers needand metallographers needand metallographers need

to find effective solutions toto find effective solutions toto find effective solutions toto find effective solutions toto find effective solutions to

preparation issues quickly.preparation issues quickly.preparation issues quickly.preparation issues quickly.preparation issues quickly.

For over 67 years Buehler has

been providing technical

solutions to the materials

industry on specimen prepara-

tion theory, best lab practices,

and material applications.

In response to this need and to further our quest for excellence in customer service, Buehler’s staff of

professional material scientists have written BUEHLER® SUM-MET™ The Science Behind Materials

Preparation - A Guide to Materials Preparation.

BUEHLER® SUM-MET™ is a summation of Buehler’s knowledge base in metallurgical and materials applications and

includes step-by-step information on sectioning, mounting, and grinding/polishing. BUEHLER® SUM-MET™ gives

extensive theory and information on preparing specific materials, etching, microscopy, microhardness testing,

imaging, lab safety, consumables usage and troubleshooting.

This book on CD is supplemented with the free Buehler e-Club on-line materials database at www.buehler.com.

There you can learn more information on your material by alloy, see the recommended preparation procedure and

view an interpretation of the microstructure. More information and materials are continuously being added. The e-

Club also provides free downloadable technical publications such as the Buehler Tech-Notes and Laboratory

Posters to help your lab with its daily work.

Buehler is the world’s premier manufacturer of scientific equipment and supplies for use in materials analysis.

Buehler’s products are used throughout the world in manufacturing facilities, quality laboratories, and universities

to enable material characterization, ensure quality, and perform materials research.

USA $120.00(CAN $160.00)

Buehler is the Science Behind Materials Preparation & Analysis™

ISBN 0-9752898-0-2 FN01255