CMS : CLJ :MPCVII 1-5'.

DHPAIT! .71:: T OF C01DURHAM of sta

A-SHTNGTQN(March #T, d9 33)

LetterCircularLC 363

TEL PROFHDTIFS AND IFSTIPG OF FOUNDRY SANDS

I. IntroductionII. General properties of molding sandsIII. The determination of the properties of molding sands;

test methods1. Preparation of sands for study

a. Determination of moisture2. Fineness3. Strength

a. Compressive strengthb. Other tests

4 . Ferine abi li ty5. Sintering6. Miscellaneous

a. Hardness of mold surfacesb. Gore sandsc. Shane of graind . C hemi c a 1 analys i s

IV. Grading and choice of sands1. Grading2. Choice of sands

V. Selected bibliography

Introduction

The use of sands ana other materials as a means for puttingmetals into useful forms by casting is knoun to date from thetime of the earliest written records end, v:ith a high- degree ofcertainty, can be projected back far into prehistoric times.Although casting is one of the oldest of all metallurgicaloperations, it has been until very recently one of the mostconservative and has been guided largely by tradition and custom.As an example, it is only within the past two decades that asystematic investigative study has been made of the essentialand characteristic properties of foundry sands.

The greater precision now required in foundry work,resulting in large measure from competition with other pro-cesses, lias necessitated careful study and control of thevarious factors which influence the production of good metalcastings. The mold into which the metal is poured and the sandout of which the mold is formed rank foremost among those fac-tors. A suitable and satisfactory mold is a prime requisitein all successful foundry operation. ITo substitute for it exists.

The Bureau of Standards receives many inquiries relatingto new sand deposits, the reclamation of' used sands, as wellas the technique, 'necessary' for developing the best propertiesfor foundry use of any particular grade of sand. The immediateaim of this circular is to furnish information on the character-istic properties of typical commercial foundry sands as deter-mined by approved test methods. Much of the best availableinformation on the properties and testing of foundry sands hasresulted directly from the extensive research sponsored by theAmerican Foundrymen’ s Association and much of the informationpresented herewith has been obtained in the Bureau’s co-operativestudies carried out with that organization.

Most of the results used later to illustrate the character-istics of foundry sands have been taken from tests of rep-resentative commercial sands which have been carried out as partof this Bureau’s work. The use of these data is not to beconsidered in any way as an -endorsement of the particular sands.There are numerous other sands with nearly identical propertiesthat might be mentioned.

II. General properties of molding s ands

A brief account of the general principles involved in theuse of sand molds will be worth while. A casting is ordinarilymade by pouring the molten metal into a mold which consistsessentially of a cavity of suitable shape and size in ‘a quantityof sand. The sand in such a mold may be either moist or dry oronly the sand which forms the surface walls of the cavity may bedry. By far the greater amount of sand used in the foundry isused in the moist, or green, condition. The molding qualitiesof a sand and its other characteristic properties are determinedon the sand in this condition. For this reason, most of thediscussion presented deals : with the green sand.

The term :? sand :' is used in various ways, usually in a

qualified cense, for example, glass sand, to indicate a specificusage. In the foundry, the term is restricted to sand consist-ing entirely of grains of silica. A natural molding sandconsists essentially of the grain material, made up of grainsof silica, and the bonding material, which is largely clay.

In addition, there may he small amounts of other substances thecharacter of which depends upon the nature of the rod: fromwhich the sand originated. These substances, usually undesir-able, are sometimes referred to as the ;, fluxing ingredients'.

The essential characteristics which any molding sand mustpossess are apparent to any one who observes the making of a

mold and its use in the casting of a metal. Naturally, thesand must possess moldability when wet, or when tempered",with a suitable amount of water. The strength or cohesivenessof the tempered sand after being compressed or rammed issurprising. The need for considerable strength becomes apparenthowever

,as ..one continues his observations. The forming of an

impression, accurate in -all details, of a pattern and removingthe pattern from the sand which' "ha-s -been rammed about it isonly a small part of the story. The strength of the compactedsand must he sufficient to ensure freedom from crumbling of thesand from any part of the walls of the mold, especially fromthat portion 'which lies in the cope, that part which, when theentire, mold is assembled, is inverted and forms the upper partof the mold cavity. Strength is also required. for holdingfirmly in place the cores which are placed in the mold cavityand around which the molten metal solidifies to form cavitiesand openings in the finished casting. Strength in the sandis especially necessary to withstand the erosive and cuttingaction of the molten metal as it is poured into the mold as wellas to withstand- the pressure of the metal itself within the mold

Although a molder always makes special provision for the'’venting 1' of a mold for the escape of airaand other gases duringpouring, it is evident, even to a casual observer, that the sanditself must play an important part in the venting of the moldduring and subsequently to 'pouring. Copious amounts of air andsteam escape from the surface of the sand. The property of thesand’ which permits this to occur is termed 'permeability 1

' and,while this is related to the porosity of the sand, the two arenot synonomous

.

Occasionally one sees in a foundry, castings which, whentaken from the mold, are heavily encrusted with burned-onsand. The removal of this is tedious and often difficult andthe surface of the underlying metal is often rough and unsightlyand details of the pattern are not accurately reproduced. Insuch cases, it is evident that the sand, as a whole, was notsufficiently infusible to withstand the heat of casting. Thisinfusibility is a property of molding sand' to which much im-portance is attached. It is related to certaiv constituents,especially the fluxing constituents and not to the grain of thes an d

.

-A-

While the characteris ti ca described cover the salientfeatures of good molding sands, the list is by no means complete.It is evident, however, from the discussion that molding sandis s. more or less specialized product, the selection and prepar-ation of which' should be carefully supervised and controlled.The testing methods by which this can readily and accurately bedone' form much of the subject matter of this circular.

III. The determination of the properties of molding sands;test methods

1. Preparation of sands for study

As in all testing work, r epresentative s an pies mustalways be selected. The Committee on Molding Sand Researchof the American Foundrymen’s Association has made definiterecommendations for suitable methods for the sampling of sands,in order that representative samules may be obtained.

Most of the tests are made on moist (tempered) sand. Itis necessary, therefore, that there should be a specific methodfor tempering. The method used by the Bureau is as follows : Asample of the sand, sufficient in amount for all of the tests,should be dried in an oven at a temperature between 105°C and110°C (221 to 230°F) for at least one hour. The sample of sandis allotted to cool and then tempered with water. It is usuallydesirable, to temper different portions of the dried sand withdifferent amounts of water; three are geie rally used, thesmallest amount being chosen so as to leave the sand a littletoo "dry" . The amounts used for the other portions are in-creased progressively and one must be guided large!?/ by exper-ience in their choice. Thorough mixing and uniform distributionof the water throughout the sand are necessary. The sand aftertempering should be kept in an air-tight container (for example,a glass quart jar with screw top and rubber seal) for 24 hoursbefore being used for testing purposes.

a . De t

e

rmi nation of mo i s t ur

e

The water content of the tempered sand should bedetermined before further tests are carried out. This is donely drying a weighed sample (100 g) on a watch glass or otheropen dish in the same manner as the original drying of the sand.The dried sample is allowed to cool in a dessicator or coveredvessel and then re weighed-. The loss, of course, is numericallyequal to the percentage of water contained in the tempered sand.

The amount of v’ater used in tempering most commercialmolding sands for testing purposes lies between 3 and 9 perc e nt

(appr oxina tely ) .

A rapid method for determining the water content of asand is often very desirable. A method ty which satisfactoryresults have been obtained at this Bureau consists in mixing50 ml of alcohol with 100 g of the tempered sand in a wideevaporating dish, igniting: the alcohol and allowing it toburr until the flame is extinguished. Continuous stirring ofthe sand during the burning is advisable. The difference betweenthe weight of the sample after cooling and the initial weightgives the water content in per cent and agrees well with thevalue obtained by drying in the oven as is shown by the follow-ing: results (Table 1).

Table 1 - Water content of tempered sand, obtained by two methods

vrater ContentSand ~dven~ Drying fgni ti~on with Ale ohol

per cent per centAlbany 00 r.

VJ * 5.6Bowner 4.7 4.7Pettinos 5.9 5.6Steel Foundry 7.5 7.5

2. Fineness

The determination of the fineness of a molding sand isundoubtedly the most important si ngle test which is made on asand. Although fineness is a rn.ee- sure only of the relative sizeof th<: constituent particles of the sand, the information gainedby a tost of this kind *is much broader than the name would sug-gest, and much can be inferred from the r suits as to the otherimportant characteristics of the sand, viz., permeability,strength, surface finish of castings, etc.

The method is essentially one of 'mechanical analysis 1 '

and Consists usually of a combination of floatation and sieving.By means of floatation, particles whose- diameter is approx-imately less than 2 0 microns, (1 micron (Al) = 0.001 mm) areseparated from those having a greater diameter. The fine-grained portion consists largely of the bonding clay of thesand as well as some other very fine material and is designatedby the name -'clay substance 5

. The coarser materiel, consistingessentially of the silica, is termed the ” grain 5

. By means ofsieving, the relative amounts of the predominating grain sizesor 'texture' of the grain material, is determined after drying.

-6 -

Since the bonding power of a molding sand is directlyrelated to, and the permeability may also be affected by theclay content, the determination of the clay substance is ofvery considerable importance. The quantity of clay substancemay be determined in various ways, most of which depend uponthe difference in rate of the settling, of particles of differentsizes when suspended in water.

The test is carried out on a 50 g. sample of the driedsand. The separation of the fine clay particles from theremainder of the material is expedited if the water used forthe suspension contains a small amount of sodium hydroxidewhich assists in deflocculat ing the clay. The common approvedpractice is to use 25 ml of a one per cent aqueous solution ofsodium hydroxide diluted ”

: ith distilled water to 500 ml. Thesand and water are placed in a glass container which should besecurely closed. A quart glass jar such as is in common usein the household i s' generally used. It is necessary to agitatethe sand-water suspension very thoroughly so as to separate theclay completely from the sand -grains. A simple device whichutilizes a small electric stirrer such as is used for preparingsoft drinks is shown in Figure 1. Five minutes’ stirring withthis device has been found ample in most cases. A motor-drivendevice for repeatedly turning the jar upside down is also usedand has teen recommended by the American Foundrymen’s Association.The process 'is continued for one-hour at a rate of 60 reversalsof the bottle -per minute.

In order that different workers may obtain comparableresults, it is necessary that a definite, method of separationof the suspended clay from the coarser particles be carefullyfollowed. After stirring the sand-water mixture, the top of thecontainer is removed, adhering particles washed into the jar andthe jar filled with distilled hater to a reference mark nearthe top, allowed to stand for 10 minutes and the water, with thesuspended clay, is removed by siphoning until the level has beenlowered 5 inches ‘below the initial level. The siphon is thenremoved, the jar filled with distilled water, and the wateragitated sufficiently to bring the grain into suspension. Thejar and contents are then allowed to stand for 10 minutes andthe siphoning repeated. The operation is repeated again andagain with a settling periods of 5 minutes, however ,

instead of10 minutes, until the supernatant t niter becomes entirely clearafter standing for 5 minutes. By this means all particles whichsettle at a rate of less than .1 inch, per minute are separatedfrom the remainder. The finer particles removed by siphoningconstitute the 'clay substance’.' The remainder is the ’’grain’ 5

.

This is removed from the jar, filtered, dried and weighed. Thedifference between the two weights of the sample represents the

- 7-

clay substance; no attempt is made to recover the clay which i.

carried off by the water in siphoning.

The various amounts of clay substance determined in thismanner on samples of a lumber of commercial molding sands aregiven in Table 2.

Table 2 - Clay substance in various commercial foundry sands

ClayName of Sand State Substance

per centAlbany, #00 New York 12.4Downer New Jersey 11.6Downer

,( open hearth) do 3.6

Lumberton do 20.0Pett inos do 15.0Red gravel (backing sand) Ohio 19.8Zanesville do 15.8Rush Run do 24.8"Campbell’s (core) Virginia 3.2Ottawa (synthetic molding Illinois 10.0

sand)French molding sand (foreign) 19.2

The bonding power of molding sand decreases with use. Thisdecrease is caused primarily by the driving off of the chemicallycombined water. The bond is practically 'destroyed in '’burnt 1 '

sand. The need for the treatment and systematic "control' ofsands within the foundry is thus apparent.

The grain remaining after the clay substance has been re-moved is dried and the distribution of grain sizes is determinedby sieving. The sieves (shown in Table 3) selected from the Bure a-

of Standards series are used for this purpose.

Table 3 - Sieves selected from the Bureau of Standards seriesfor foundry use

Sieve.

OpeningNo

.

Inch mm6 0.1320 3.360

12 .0661 1.68020 .0331 0.84030 .0232 .59040 .0165 .42050 .0117 .29770 .0083 .210

100 . 0059 .149140 .0041 .105200 .0029 .074270 .0021 .053

-8 -

Tlie sieves, which are 8 inches in diameter and of the half-height construction, are nested securely together in order withthe coarsest at the top and a pan beneath the finest (Ho. 270).The sample is placed on the upper sieve and covered and the wholeassembly shaken by means of a Rotap mechanical sieve shaker orits equivalent for a period of 15 minutes. The amount remainingon each sieve is weighed and its percentage of the sample cal-culated. The approximate grain size of any of the variousfractions secured in this manner is assumed to' be that of thesize of the openings of. the sieve immediately preceding the oneon which the sample was retained.

The results of sieve analysis of a number of commercialfoundry sands are given in Figure 2, in which the percentage ofthe '’grain" has been plotted against the approximate grain sizeas indicated by the sieve openings. The arbitrary value of 3has been assigned by the Committee on Holding Sand Research,American Foundrymen’s Association, to material which fails topass through the first sieve (No. 6) and 300 to the particleswhich pass the finest sieve and are caught on the pan. Twofurther "arbitrary changes have been made in that the number 5has been assigned to those grains which pass sieve No. 6 andremain 6n sieve No. 12 and 10 to those grains which pass sieveNo. 12 and remain on sieve No. 20. By means of the diagrams inFigure 2, the difference in the distribution of grain sizes ofthree representative foundry sands is clearly shown.

For purposes of classification and specification, an averagegrain fineness number is often advantageous. This is readilyobtained as is shown in Table 4, the calculation being analogousto the method used for determining the center of mass of a systemof rigidly connected bodies. The average grain fineness numberis then approximately proportional to the total surface area ofthe particles ef a unit weight of the material.

Figure 1. - Stirrer used in the sep-aration of the clay and grain of amolding sand

The suspension of sand and dis-tilled water (which contains a smallamount of sodium hydroxide) isforced by the stirrer past thestationary "hair-pin'’ baffles whichare integral with the screw cover.

Figure 3. - Device used in determiningthe compressive strength of sand spec-imens

The specimen is placed beneath thecompression head and the spring com-pressed by turning the handle. The dis-tance to which the graduated rod pro-jects as the spring is compressed, in-dicates the compressive stress ( lb ./in .

2

)

transmitted to the specimen. When thespecimen fails, the projecting rodremains fixed in position.

LUMBERTON MOLDING S AN D Cnew jersey)

1 ll

,-GRAIN FINENESS NUMBEROR AVERAGE CRAIN SIZE.

5 10 20 30 40 50 70 1001

Z< 20 .

ce

o

I 0.

o

q: o

FRENCH MOLDING SAND C FRANCE DCRAIN FINENESS NUMBER

fOR AVERACE CRAIN SIZE

a L10 20 30 40 50 70 100

ALBANY MOLDING S/\ N D C N EW YORK }

CRAIN FINENESS NUMBER/OR AVERACE CRAIN SIZE

1102030 40 50 70 100 140 200 300

GRAIN SIZE Cmeshes per lineal inch of sieve just PASSING CRAIN , approximate)

—GBAIN FINENESS NUMBEROR AVERACE CRAIN SIZE.

OTTAWA CORE SAND Gllinois)

20 30 40 50

ll

100 140

DOWNER MOLDING SAND Cnew jersey)

-CRAIN FINENESS NUMBEROR AVERACE CRAIN SIZE.

10 20 30g40 50 70 100

I

RED GRAVEL BACKING SAND Como)

-CRAIN FINENESS NUMBEROR AffCRACE CRAIN SIZE

JL_l3310 20 3 0 4 0 30 70 100 140 2 00 100

CRAIN SIZE CUCSHCS PER LINEAL INCH OF SIEVE JUST PASS IN C CRAIN , APPROX I MAT E >

FIGURE 2 - DISTRIBUTION OF THE GRAIN IN A NUMBER OF COMMERCIALFOUNDRY SANDS . AFTER THE CLAY HAS BEEN REMOVED THE RECOVEREDGRAIN IS DRIED AND THE VARIOUS FRACTIONS SHOWN ABOVE ARE SEPARATEDBY SIEVING. IN PLOTTING THE DIAGRAMS, THE TOTAL GRAIN HAS BEENCONSIDERED AS 100 PER CENT (NOT THE PERCENTAGE SHOWN IN TABLE 10).

-9-

Table 4 - Calculation of grain fineness number of a foundry7-

sand (Downer molding sand)

Sieve No. Amount (A) Retainedon Sieve

Relative

(

a )

. Grain Size (S)

ProductAS

per cent

6 0.0 3 012 .0 5 020. .8 10 830 1.2 20 2440 3.4 30 10250 4.4 40 17670 29 .6 50 1480

100 32.2 70 2254140: 12.2 100 1220200 1.5 140 210270 0.4 200 80Pan 1 • 300 390

Total (I A )

=• 87. Total (IAS' = 5914Average grain fineness numb e r

r AS i- EA = 5914 - 87 - 67

(a) Proportional to the reciprocal of the grain diameter.

The results plotted in Figure 2 show the relation of theaverage fineness number to the distribution of grain sizes ofthe sand.

Efforts have been made to represent grain size distributionnumerically in a manner similar to average grain fineness. Ifthe complete results of the mechanical analysis by sieving areavailable, a numerical t? distribution factor” is, of course, notneeded-. A method which is mathematically sound consists in anextension of that used in Table 4.

Instead of the summation of the products, AS, the summationof the products, AS 2 is divided by the mass (£A)

,

that is, thetotal percentage of grain. A much simpler method consistsessentially in designating the number of consecutive sieveswhich retain at least 75 per cent of the material. In esse oflack of complete results of sieving, such a factor, togetherwith the average fineness number, would serve to define a sand,closely enough for practical use.

-10-

5. Strength

(a) Compressive strength

It was pointed out that ” strength! is one of theimportant characteristics of molding sand. In a strict sense,the term, strength, should be restricted to the molded sandspecimen and not to the loose sand. Of the various .methods inuse, the one which determines the. strength of the molded spec-imen in compression is preferred and used most frequently atthis Bureau. The specimen consists of. sand which has beentempered in the manner described, compacted in a known andreproducible manner, into a cylinder, 2 inches in diameter and2 inches in height (tolerance ± l/l6 inch in height).

The sand is compacted by ramming within a cylindricaltubular container, by means of a freely sliding (falling) weight(17.5 lb.), so arranged that the entire cross-section of thecylinder is uniformly affected. Three blows, under a 2-inchdrop of the weight, are used. If the. .height of the rammedspecimen does not come within the prescribed limits a new spec-imen is made with a slightly different initial amount of sand.

The device for determining the compressive strength is ofrelatively simple construction (Figure 3), the calibrated springbeing the most important feature. By turning the handle and com-

pressing the spring, load is applied through the movable compres-

’ to the sand specimen at a rate of from 20 to 40 poun s

t nilrte until the specimen collapses. The distance trough

Shlh the swing has been compressed at the time tne specimenvhicn he sp o

calil3rated rod which remains fixed mposition. From"the calibrated rod the pressure required to break

the specimen may be read in pounds per square inch.

Various other means t°f me asui essive strengthall.n

•ing the comp

of sands have been devised and are to some extent used,

of them, specimens of the above form are used.

For any particular .molding sand, the strength is dependent

upon the amount of water contained and. the compactness of e

sand. For specimens'

prepared by ramming m. the manner described,

the water content is the principal controlling factor.^

ue ie

suits shown in Figure 4 illustrate, this for a number oi comnei

cial molding sands.

FIGURE 4 - COMPRESSIVE STRENGTH AND PERMEABILITYOF A NUMBER OF COMMERCIAL MOLDING SANDS INEACH CASE FOUR DIFFERENT DEGREES OF COMPACT-NESS (RAMMING) HAVE BEEN USED.

II

H_

A-

B—

V

lei «-£

AIR

WATER

frti

atdi

A-bellB-tankC- AIR OUTLET TUBED- GUIDEE- VENTSF- SAND SPECIMENG- 3-WAY VALVEH-WEIGHTI - SAND CONTAINERJ-MANOMETERK -RUBBER STOPPERL* ORIFICE PLATE

Figure 5. - Diagram showing the essentials of the methodused in determining permeability of sand specimens

The modified method which depends upon the use of acalibrated orifloe, is shown in the smaller diagram.

Figure 6. - Apparatus used in determining the "sinteringpoint" of foundry sands

A is a U shaped yoke, between the free ends ofwhich, is stretohed the platinum ribhon (resting on thesand specimen, H) which is heated by an electric currentcontrolled by tUe carbon-plate rheostat, B, and read byneons of the ammeter £ {range, 50 ampere sf. D is used inoonnection with the optical pyrometer S for measuringthe temperature of the hot platinum riPbon and F is thelight-proof cabinet used to cover the specimen and heat-ing element during the test.

-11-

(1) Other tests

The American Poundrymen’s Association, through itsCommittee on molding Sand Research, has approved the compressiontest in all cases, as in disputes, in which comparisons of theresult's ol‘ differ at investigators are necessary. The deter-mination of the tensile and the shearing strength in place ofthe compression strength, is in rather wide usage. A method,formerly widely used, consists essentially in slowly sliding astraight bar of uniform rectangular cross-section molded fromthe sand over the edge of a supporting table so that, the oneend overhangs. The weight of the maximum overhanging portionis determined. On account of the difficulty of producingspecimen bars of uniform and reproducible properties, the useof this method is decreasing.

Since the strength of a molding sand is primarily dependentupon the bonding material, any test which gives information onthe nature and amount of this bonding material should be ofinterest. Much of the bondingLixe colloids of this .-type

,it

the coloring matter from dye solutions. Hence, the corn.par.isonof the color of two glass tubes, one of which contains a watersolution of a dye (crystal violet) and the..other the same amount

substance is colloidal clay,has the property of adsorbing'

of a similar solution \

the dried molding sandcolloidal character ofthe dye solution whichd irnin i s he d , t he d e gre eW it*n the unchanged dye

with a sample.on on theThe color ofthe sand isby comparison

hioh has been shaken up with a s amT>le ofwill give some informatthe "bond”, of the sand,has been shaken up withof dimunition (measuredsolution of similar concentration)

gives qualitative information on the bonding characteristics

of the clay. In carrying out the test, it is advisable to add

a small amount of an electrolyte (for example, acetic acid

accelerate the settling of the colloidal substancea clear solution. Considerable experience, however, in the

use and testing of molding sands is necessary in applying the

information. The test is' unreliable for used sands, especiallyif they have been treated by the addition of various substancesas is common practice in certain lines of foundry work.

toobtain

4 . Pe rrae ahl 1 i ty

A measure of the relative permeability of a chosen sandis obtained by measuring the rate of flow, of air under pressurethrough a standard specimen. The specimen is identical withthat prepared for the determination of compressive strength,with the important exception, however, that it is not removedfrom the container in which it is rammed. In routine testing

-12

the determination of compressive strength r?.k] permeability aremade on the same specimen, permeability being determined first.The time required for forcing ,‘3000 cm3 of air through thestandard specimen under a known' pressure is measured with a stopwatch. From the data obtained is calculated the volume of air>assing through a unit v )lume of sand (a cube, 1 cm on eachedge) in one minute under a pressure of 1 g per cm2. This is,by definition, the permeability of the sand and may be expressedby the formula

P = v* h .

p . a . t

4

in which P -

v =

h =

a =

P =

t -

the permeability "number”

,

volume (cm3) of gas (air) forced throughthe specimen,height of sand specimen (cm),cross-sectional area of sand specimenexposed co the gas (cm2),pressure used in forcing the air throughthe sand (g),time (minutes).

By substituting the values for v, h 'and a in the formula( 2000 .cm3,, 5.08 cm and 20.27 cm2, respectively), the formulareduces to 1

p = 501. 2,

p. t /"

The .pressure is read directly by means of a manometer.

A diagram shoving the essentials of the apparatus is giveni n F i gur e 5

.

‘The .determination of permeability can be greatly expedited,for routine testing work , by a relatively simple modification ofthe. apparatus and with only slight sacrifice in accuracy. -Themodification consists essentially in- controlling the rate offlow of air from the reservoir to t ne sand specimen by insertingin the supply tube a plug having an orifice of known* diameterand shape. To insure permanence against corrosion and otherpossible deterioration, gold is used as the material surroundingthe orifice. In this modified method, it is necessary only toread the pressure at which the air is being forced through thesand and obtain the corresponding permeability number from atable based upon previous determinations.

Various precautions depending upon the condition of thesand tested are sometimes necessary in the determination of thepermeability number. For example, a screen (60 lesh) is used tc

-13-

support specimens of dry sand containing little or no bondingmaterial. In determining the permeability of baked specimensof a sand containing bonding material, a special holder isused and it is necessary to "seal'’ in the specimen with mercuryand a rubber gasket to prevent the flow of air between thespecimen and the adjacent wall of the container.

Permeability and strength ,of molding sands are, in general,opposing properties. Conditions such as an abundance of bondingclay or a high degree of compactness, both of which are favor-able to high strength, are conducive to low permeability numbers.The effect of progressive ramming in increasing the compressivestrength of molding sand and simultaneously decreasing thepermeability is clearly shown in Figure 4. The. general effectof water content (degree. of tempering) on the strength and per-meability is shown, as well as the rather wide- -differences which,may exist between different sands on account of the differencesin their inherent characteristics, clay content, grain sizes,etc. The need for strictly defining the conditions of preparatioof the specimens upon which strength and permeability are to bedetermined is apparent.

5. Sintering

. Although silica, the chief constituent of molding sand,undergoes two crystalline transformations on heating (870°C, totridymite, and 1470.° C ,

to cris tobalite) ,

it is otherwise stableuntil the melting point, 1710°C ± 10 (3110°F) is reached. Thistemperature is well above the temperatures reached in the castingof iron and non-ferrous alloys. The nearest approach is in themaking of steel castings, A casting steel becomes completelymolten on heating to approximately 130.0°C (2730°F).

The refractoriness of a molding send under conditionsobtaining in the Casting of motels is dependent primarily uponthe amount and nature of '’fluxing constituents' 5 and to someextent the bonding materiel. A determination of thu meltingtemperature is seldom necessary. However, the determination ofthe temperature at which a sand sir.ters or -burns on a castingis important. The method used for this determination consistsin holding a ribbon of platinum, maintained at a chosen temper-ature by passing an electric current through it, against thesurface of a specimen of the sand.. The specimen is prepared inthe manner already described for the determination of strengthand permeability . After removal from the cylinder in which itv/as rammed, enough sand is cut away to reduce the dimensionsto 2 inches by 2 inches by 1 inch and the specimen is dried for

-14-

1 hour, at 105 - 110°C. The current is adjusted so as to pro-duce the chosen temperature, as determined by an optical pyrometerof* the "disappeariHg-f i lament n type, proper corrections foremissivity and. other conditions being made. The ribbon is thenheld for four minutes against the narrow face of the specimenby the weight (170 g.) of the ribbon holder, and then examinedafter the current has been cut off for evidence , of s.intc-rcdsand particles sticking to it. The operation is repeated, eachtime at a temperature 25 degrees higher than the preceding oneand on another portion of the- specimen, until particles of thesand may be seen to have sintered to the ribbon. The temperatureat which this occurs is taken as- the "sintering point” of thesand. The apparatus is shown in 'Figure 6. Table 5 gives thetest results on representative commercial molding sands . It willbe noted that there is a fairly consistent relationship betweenthe average grain size and the sintering temperature. This isdefinitely lower ' for : sands of finer- grain size than for those ofcoarser grain size.

Table 5 - Sintering point of various commercial molding sands

Sintering Clay Average GrainSand Temperature Substance Finene s s Numb e

r

°c °F( a) . Per GentFrench molding 1175 2145 19.2 192.Albany #00 -

. 1200 2190 12.4 188.Zanesvi lie 1200 2190 15.6 196.Lumberton 12 00 2190 20 • 111.5Lush Run 1225 '2235 .

- 24.8 110

.

Pettinos 1225 2235 . 15. 83

.

0ttawa( added bond) 1250 2280 10. 50.DownerRed gravel (backing

1250 2280 11.6 67.5

Sand

)

1300 2370 19.8 35.5Downer open-hearth 1325 2415 3.6 67.Campbell core 1400 2550 3.2 46.

(a) 5Trounded of f ”

.

The approximate pouring temperatures in ordinary foundrywork are given in Table 6.

-15-

Table 6 - Pouring temperatures for foundry castings (approximate)

(The pouring temperature is Vailed with size of castingand other factors .

)

•

CastingsPouring

deg. C

TemperatureDeg. F

Aluminum ialloys 700 - 780 1290 - 1455Brass 1000 - 1520 1850 - 2410Cast iron ( stoveplate

)

1260 - 1425 2500 - 2600 •-

Malleable cast iron . 1425 - 1550 2600 - 2820Steel 1525 - 1600 2750 - 2910

6. Miscellaneous

a. Hardness of mold surfaces

The determination of the surface hardness of green-sand molds is a test which is coming into favpr. The results,however, are more characteristic of the mold and particularlythe compactness of the sand produced by ramming, than of aninherent property of the sand. The- test is somewhat analogousto that of the Brinell hardness testing of metals. A smallportable device having a flat base, through which projects thehemispherical end (indentor). of a movable plunger acting againsta spring, is pressed against a flat surface of the mold or sandSpecimen until there is close contact. The depth to which theindentor penetrates the sand is shown on a graduated scale. Theinverse of the depth of penetration has been accepted as a measureof the hardness of the sand mold surface; a penetration of 0corresponds to a hardness number oT 100 and vice versa. To theexperienced molder the results are indicative of the strengthand permeability of- the mold, particularly the surface layer.

The results summarized in Figure 7 illustrate the relation-ship between the compressive strength of specimens of fourcommercial molding sands, as determined in the manner previouslydescribed, and the surface hardness of the same specimens. Thesurface hardness values shown in Figure 7 are also related to thepermeability values, shown in Figure 4.

b. Core sands

-In general, considerably greater strength is requiredin sand cores than in sand molds. In order to obtain thegreater bonding strength various artificial 'binders' are addedto the sand. Silica sand is usually used, although molding sand

-16 -

is used sometimes. The ' 'Molded core is usually baked in orderto develop the required strength. The permeability of cores ofbaked molding sand is determined in the manner already describedand the strength is usually determined by breaking in tension abriquet somewhat similar in shape to those used in testingcement. In the preparation of 'test specimens of cores, the samecare and procedure should be used as in the making of the coresthemselves. These tests are of greatest value in the study of theproperties of core binders and for such tests a carefullyselected, washed and dried silica sand consisting of well-roundedgrains between 40 and 70 mesh in size, with at least 95 percentof the 70-mesh size, is used.

c. Shape of grain

Although much study has been devoted to the subjectof grain shape, no definite statements supported by the resultsof tests can be made as to the practical aspects of this phaseof the subject. The grain shapes of foundry sands are usuallyclassified as: angular, sub angular

,rounded and compound. By

some writers, the rounded grain is considered as the basis ofthe ideal” molding sand.

d. Chemical analysis

A determination of the chemical composition of amolding sand- is not required in ordinary testing. In the studyof new deposits and of sands which are, in certain respects,unsatisfactory, chemical analysis must often be relied upon.The determination of the nature and amount of the :tfluxingconstituents” is often helpful in this respect. A sand may beunsuitable for foundry use because of the presence of particlesof undecomposed mica or of feldspar, both of which result in alow. sintering point for the sand. The determination of calcium,sodium or potassium by chemical analysis is useful in suchcases in establishing the cause of the inferiority of such sands.

IV. Grading and choice of sands

1. Grading

A number of classifications of molding sands have beenrecommended by the Committee on Molding Sand Research of theAmerican Foundrymen’ s Association. The class designations serveto define rather closely a sand for foundry use. In addition togiving the grain fineness number (page 9) and the grain shape(page 16) the sand may be classified according to the grainfineness and the clay content. These classifications are givenin Tables 7 and 8.

-17-

Table 7 - A.F.A. classification of sands according to grain,f i ne ne s s

ClassDesignation Range in Grain Fineness Number

(

a )

1 200 to 3002 140 fT 2003 100 tt 14-0

4 70 - ? T 1005 50 ft 706 40 Tt 507 30 TT 408 20 t? 309 15 Tt

.2-0

10 10 TT 15

(a) In all cases, except No, 1, the range of fineness numbers'extends up, to but does not include the number marking theupper limit.

Table 8 - A.F.A. ' class ification of molding sands on the basisof clay content

Clay Designation Clay Conte nt•*''' per cenT fa )

A .0 to 0.5B 0,5 ” 2C 2 " .5

D • 5 " 10E 10 " 15F 15 f

' 20G 20 .

. 30H • • 30 45I 45 " 60J 60 fT 100

(a) In all cases, the range of percentage of clay extends _to

but does not include the upper limit.

-18-

2. Choice of sands" • * --

,

Only very general 'recommendations can ha given to coverthe choice of sands for specific uses. The information -givenin Table 9, which is based largely tipon experience, is intendedas suggestive rather than in the nature of a categorical replyto the question often asked of this Bureau as to the' proper sandfor a certain specific use.

Table 9 - Sands for specific uses

Casting A .F.A. Sand' Compressive Perme abilityRei. Size Class Strength Number

''

,

' .--j

..

lbs/ sq. . in.

Brass light .1;D or r7

3 - 4.5 " 8" - 12medium' . 2 '3 : - 5' 10 - 17

• O . 0 • . . o «,

he avy r~y

O5

3.5- 6 - 17 - 25' f :

'

* **

Aluminum'

light, }

' 1 • -> F or G 3 - 7 8 - 15alloys medium

)

t i

?7- • - ... . - -heavy

.

1 or 2 ;E or 4 - 8 10 - 20

: • - • •

Cast iron light ' 2 ,3 or A'-"'2

-i.,

xj 3 - & 15 - 65r? medium 4 or 5

; —

t

4 - 9- - 65 - 100rr heavy d ,6 or 7; E 6 - 10 '• 75 - 150

or F

Malleable light 3; E or F 3 - 8 20 - 50

iron medium .4?

J_i 4 - 9 50 - 100he avy 5 , 6 or 7; E 6 - 10 90 - 125

3 tee 1 light 5; c 4 - 8 100 - 150

if medium :- r 5 or 6 ;

r\KJ 4 - 8 125 - 200

-. ..he avy 6 or 7

;

C 4 - 8 i. 150 - 250

For convenience in .reference: -most of the data on moldingsands given in the course of the preceding discussion has beencollected in Table' 10.

MOLD

SURFACE

HARDNESS

NUMBER

COMPRESSIVE

STRENGTH

MOLD

SURFACE

HARDNESS

NUMBER

COMPRESSIVE

STRENGTH

STRENGTH OF A NUMBER OF COMMERCIAL MOLDINGSANDS. COMPARE FIGURE 4

fO ro M M M M M M M M M Mm o vo 06 “>i a vn f v*i ro m o vo oo ->j cr vn f v>4 go m 2

0

N.

Y.

Ohio

N.

J.

dodoOhio

Va.

Foreign

N.

J.

Ohio

N.

Y.

Ill

.

doMich

.

Ala.Mips

.

Tenn

.

N.

J.

-

Md

.

doD.

C.

0 coH O c+H- M- ®n r+m ro

3

Albany

,00

0.0

Zanesville

—

Pettinos

Downer

Lumber

ton

—

Red

gravel

11.6

Campbell

1

e

core

sand

FrenchDowner

—

open

hearth

and

core

sand

Rush

run

—

Albany

,00

Ottawa

core

sand

Ottawa

steel

—

mold

sand

Buohanan

—

Tuscaloosa

—

Red

molding

—

MoldingSouth

Jersey

—

CampbellMarlboro

2.6

Georgetown

—

U)

All

ar»

ro 3

P. M_ocMwo

cr

aro

ro

M?

3ro

-

O M O O' o3 1 • 1*11 III 1 * 1 1 . • 1 1 1

O I cr irollO ill i ro 1 i v>i m v>4 i i | M Hro

CT

35

OH;

Hi

Mro

<3

ro

s,

P Mm C i—1 v>i ro O3 1 ... | | .11 1 . . • . 1 .

jq i vc v>j f i i cr v>4 i i i v>i a m a ->j. o vn i rv> ro

ro0

a i<- _ mp m Ovji O »- c f vn v>j m o3 1

... 1 1. • 1 1 . • • . . • 1 •

p. y o vn O' i i a mi i m a m vn m o cr i h v>j <5

M MMg 1 M M M O ro O 1 00 O M Vn C Vn f F cr ro OM

1 F VJl VO CS O cr M Vjj O F M O' fO M F O vtf F £ro

M<ro

CO M ro M M V*l Mro O lb 06 -J m ->j m O i vn O f f ovd -j -vi ro ro Oo ro O cr ->i m oq V- v*i a v»i vc vo oo ->j o b bo $ O

H

Oro3ct

ro

c+pM3ro

P-

03

MA<1

ro

D" ro m v>i vo v*i ro ro ro m m ro-(TO O VT O-nI-F’OO' ^ O O O O M M F v>J o 00 -jJ O M3 F ro o-^ao o f " * vn F vo f io ro v>j m cr vo crM

Oro

® M ro M ro MM M V>J M V>J ro Mv*i o vjj-^irorovo m i f m o o v*j ->i- ro -F F as O -j

ro -FMcrMMvnoa -vi o-f ct-nIO O vr ro cr O OOO

M MM M M M MMM Meg -vj -vj oo F oc oo v>j vn

^

o' vn —J -vi —vi m m ft m - co oj cr

^ m f -vj vn vjn cs vo i

’os vn vn -f w Cj os vn o os. ro

HS'rorov^jroMM m m

p. ro os vo F vn m m m i ro os f v*i vo O'-^i m a a ->j

F vr f vn oo cr v*i*

os c cr cr bo vn o -j v*i

roOO

M M ro Hvn vo M v^i vji O vo ->Jli O Ovji O v»J O ->j ro

4= O osmvo!fv*i v>j cs ro vn oj oj vn m vn cs -S os

ro

c

ro m m ro -F vji m ro m -*i v*j-vj vo m a os c ro mii vn ro o mm v>j ro m , v*i o» ov*j F ro ro v»j cr o cs ro vo v>j oo o o ro cr o vo oo

8a

ro ro m ro m ro m m ro m m ro m m m mv>j vr f (T v>j -vj ro -sjOl -F v>j vo v>j vc o m vr vn ro

-vi vraroMcs v>j cs cr ro ro cs O cr o oj -F

Q-S

..

: '*0

ro

m *-r

-vjvjiv>j ocvtv>j os cr -F — o -vj os crvrv^ m a -si cr f crvrv>j -vivrv>j ncf o -Fv>j ~vjvrv>j —jvt^j a -f ro -vjvt -F ->jvnv>j. 4 . H

ro^M m osvc v>j mvj vn O o o ->j Ovji *<4 c» ro->J vdv»j->j ro m f -Fro m vj vjivc cr ro cs cr f^^ vt^j os cs m m ro lovr ro

e+

s0f-nH"c4ro

33D*

ro

Mro

PMd

00

Mc+Mro

a

MMM ro -F^ ro Mm rovn ro ro ro m m cs ro ctvommm f=vt cr v>j -FV-r v>j ro ro oa F'Oov.j rov>i m cs O osvo->j m m m mi

—

j f vo m cr -f-F'O o vc os o cs vc cr vr vr f^ F7 ^ a om cs cr a Ov^vr -^j o ovr vo os so -^j os ro ^j^>iio Osp-^i != -F'-F

cr a F o O vc vt--jv*j v>j vro-Vj jz oaos->j -.>1 vr-^4

cr* ro- H-* 3c+ ro

< p1

MM MMM M M MM MOM CT

v>4i--viv>jroro Mroro vo m ro v>i f=v>jv»j cs i f=vji cr ofoj ro m m vo f=v>i i cr o cr cr a oj fO cr cs cr F^ F -F -Fvr a ro

-sJ b-v vr os-vj vi^vr.o o ro ro mvjiv^i os ->j-viv^ Froro oerer O cr ro OfP* foo-^i Ovd io ro os ro m^-oJ m FO ro

H J3*<

3

00

W 3rt-13

-i Hd ro

3 ro

££J <

ro

1 M,l m m m m mmmm m m m m m m1 - .1111 - - - - - - - - » » o1 ro ro I i l i ro ro i m ro v»j m f v>i ro ro ro ro ro o

vri Q ro vn -viro ro—JO oovnro Q 5o o vn o vnvjivnvno ooovjiOOCO

0 MO 3M c+3 ro

c+- HM3oq

ro M M M MM ro M M M _ H Mo vn ovDv^cra vovnvn f m cr vof v^i m a os so «m vn vorororo ro -vioo sb o -0 roa- vn m —j v^ cr os

m F vn

'

3 cro h

D M3

3

O' 3ro ro

h 1

cNPaa

m ro vai f v>i vn ro ro vn vn mv*i vn ro cr -viv^jvn F f\) ro0 OM Hro ro

ro ^ro 3

0 0 cs^cri^o otfli **j cs 0 **30 •*: 0 n ^ ^ ^0 0-* Mro p>

Dro

Sub-

angular

Com-

pound

Sub-

ancrular

dodododododo

Angular

Sub-

angularAngular

Round

Com-

pound

Sub-

angular

dodo

ro Q3 Hro p

iS-

Cast

iron

,

braes

or

aluminumdodo

Cast

iron

and

steel

Brass

,

cast

iron

MoldingCores

Statuary

bronze,

etc.

SteelIroh

Br^ss

and

aluminum

CoresSteelCast

iron,

brass

or

aluminum

Brass

and

cast

iron

SteelBra

s8and

iron

Steel

IronCast

iron,

brass

or

aluminum

Brass

or

aluminum 5

B

am

3

0PCO

1

f X f 1* _ F

1 MiiroiP- mii 1 1 1 11 9'9'S‘m1 oq 1 1 p 1 0 oq 11 1 1 1 11 OOOOOWID-II^I t=r 1 1 1 1 1 11 ^ 5

33s ro-* M^ Pro m<ro

3PO'H®

O

IP-a

lommerolal

molding

-19-

V. Sele cted bibliography» •

1. American Foundrymen’s Association. Report on sand testsconducted under the auspices of the American Foundrymen’sAssociation, Tr. Am. Fdy. Assoc., 32, pp. 244-358, 1925.

Geological survey data on sands from Alabama, California,Maryland, Michigan, North Carolina, New Jersey, Pennsylvania,Tennessee and Virginia.

2. American Foundrymen’s Association. Testing and gradingfoundry sands, Standards and Tentative Standards, Am. Fdy.Assoc., Chicago, 1931.

A complete description is given of apparatus and methodsof testing and grading foundry sands adopted by the AmericanFoundrymen’s Association, following recommendations of the Com-mittee, on Molding Sand Research.

3. Adams, T. C., Strength tests for foundry sands, Tr . Am. Fdy.Assoc., 33, pp. 404-492, 1926.

The relation between tests and foundry practice and thegeneral problems involved in carrying out the tests and applyingthem to foundry conditions are treated. The apparatus used tomake strength tests and to prepare test specimens is described.

4. Adams, T. C., Testing, molding sands to determine theirpermeability, Tr. Am. Fdy. Assoc., 3_2, II, pp. 114-167, 1925.

A complete description of the methods of testing for per-meability in molding sands is given.

5. Ashley, II. T., The colloid matter of clay and its measure-

ment. Department of the. Interior*, U. S .. Geological Survey,

Bulletin 380, 1909.Colloid theories are applied to a description of the proper-

ties of slags. It is shown that the absorption of certain dyesby clays furnishes a means of estimating the colloidal matterpresent. •

6. Batty* George, The most potent variable, Tr. Am. Fdy. Assoc.,38, pp. 309-331, 1930.

The subject of the production and control of green-sandfacings is treated in its application to the elimination of de-fects, in light steel castings.

7. Cole, H. J. , A comparison of natural bonded and syntheticmolding sands for the steel foundry, Tr . Am. Fdy. Assoc., 39,pp. 161-173, 1931.

A comparison is given of the working properties of syntheticand natural bonded steel molding sands as used in actual plantpractice

.

-20-

8. Dietert, H. F.,Commercial application of molding sand

testing, Tr. Am. My. Assoc., 32, II, pp. 24-46, 1925.Control of permeability by proper additions and tempering

with the right amount of water are test controlled by testmethods. The application of test methods in the selection ofnew sands and the control of sands in use and the benefits de-rived from sand control testing are discussed.

9. Dietert, H. XL, Control of hardness and other mold proper-ties, Tr. Am. Fdy. Assoc., 40, pp. 63-71, 1932.

The control of the sand condition in the rammed mold isproposed as a link between sand control of heap sand and castingdefect's. The relation of mold hardness to the other physicalproperties of the sand in the mold is shown.

10. Hanson, C. A., The grading of molding sands, Tr . An. Fdy.Assoc., 34, pp. 373-403, 1926.

An interpretation of the Albany grading scale for the pur-pose of providing a logical basis for the orderly classificationof sands is given. The results of screen analysis are discussedand the general requirements of a sand indicated.

•

11. Hanson, C. A., The physical properties of foundry sands,Tr. Am. Fdy. Assoc., 32 II

,

pp. 57-97, 1925.An endeavor was made to correlate the various measurable

properties of a few simple sands and to establish relationshipsbetween them.

12. Harrington, R. F. ,Fright, A. S.» Hosmer, M. A., Practical

and technical data obtained from the use of clay bond in moldingsand heaps, Tr. Am. Fdy. Assoc., 34, pp. 307-326, 1926.

The authors endeavor to cover from a very practical stand-point the many problems which were encountered in an effort toemploy clay material successfully in molding sand heaps.

13. Hartley, Laurence A., Flementary foundry technology,McGraw-Hill Book Co., Inc., 1928.

A text suitable for general use in foundry work. Chaptersare included covering the properties and the handling of molding.sands.

14. Kiley, T. F.,Sand control and sand conservation, Tr . Am.

Fdy. Assoc., _36, pp. 359-376, 1928.The value of sand tests is pointed out in connection with

control, conservation and reclamation in a gray iron jobbingfoundry. A general application of the facts may be readily made.

-21-

15. Kerlin, ’ :T

. Kprley, Status of* foundry sand control as seenby the send producer, Tr. Am. Fdy, Assoc., 38, pp. 373-584, 1930,

An attempt was made to point out the relationship betweenthe properties of new sand and of its effect' on the heap, and tooutline the methods available to. the producer for controllinghis- sh i prnent s

.

16. Kuniansky, M. ,

Routine sand control in Pipe foundry,Tr. Am. Fdy. Assoc., 35, pp. 170-178, 1927.

The practical value of sand control is shown. The 'sand waskept within certain limits for permeability, moisture, greenstrength and dry strength.

17. Leun, A. V., The effect of mulling on the physical propertiesof foundry sands, Tr. Am. Fdy. Assoc., 34, pp. 269-306, 1926.

The effect of mulling on the permeability, compressive andtensile strength of sand is shown. An increase in efficiency bymulling the sand is claimed.

18. McMahon, J. F. ,The refractoriness of moldinfg sands, Tr, Am.

Fdy. Assoc., 37 , pp. 501-525, 1929.A comparison of the various methods for determining the re-

fractoriness of molding sands based upon the work of previousinvestigators is given. The Saeger sintering test is recommended.

19. Kevin, Charles M. ,Notes on the grading of sands with

special reference tc Albany sands, Tr. Am. Fdy. i.ssoc., 32 , II,

pp. 182-219, 192 5

.

Grading may be based upon the graphical representation ofthe sieve tesus linked with permeability values but the bondingstrength should be expressed separately.

20. Phillips, H. D., Sand control in the steel foundry, Metals &Alloys, 1, pp. 759-762, 1930,

Moisture, grain size and bond are considered in their effecton the conditions of permeability, strength and proper molding.

21. Ranis, K.,Sand control in malleable foundry. Tr, Am. Fdy.

Assoc., 3_7 , pp. 577-585, 1929.The general procedure is discussed and the benefits

resulting are pointed out.

22. Reichert, rr. G. , Sand control methods and their developments

in a light casting foundry, Tr, Am. Fdy. Assoc., 36, pp. 213-234,1928.

The practical application of bhe standard methods of sandtesting to the foundry control of molding sand is discussed andthe improvements which are directly associated with these testsare shown.

23. Reichert, rr. G. and Woolley, D , , Factors, which influence

'the surface quality of ‘gray iron castings, Tr. Am. Fdy. Assoc.,39, pp. 205-234, 1931.

A series of investigations to determine the influence ofvarying sand conditions on the face of a casting was carriedout. The conclusions were that grain size and moisture contentare the main factors which influence the surface quality.

24. Hies, II., Report of American Foundry-men’s AssociationCommittee on Molding Sand Research, Tr. Am. Fdy:. Assoc., 59,pp. 541-568, 1921.

The report contains valuable information concerning generalresults of committee work. Tests on clay and sand mixturestogether with the method recommended for. testing', bonding claysand the definition of foundry terms’ are especially noteworthy.

25. Young, S. R. , Characteristics of some steel molding andcore sand materials, Tr. Am. Fdy. Assoc., 35, pp. 202-213, 1927.

A survey of the common materials used in steel and coresand -work is presented.