to know cleaning you have to know sou--especially particles · to know cleaning you have to know...

4
To Know Cleaning You Have to Know SoU--Especially Particles By John B. Durkee, Consultant in Metal and Critical Cleaning The title is one of favorite (at least professionally) expressions. And it makes sense: it is, after all, why my every country's military operation h~ it the is capable of and what enemy ~they are likely to do. If one has at them, one may have to know what the soil is and where it is located in order to remove it. In 2007, this column will focus on soils--greases, oils, particles, and others - and what to do about them. That soils can be so broadly characterized, cleaning so challenging. Soils are a gift to cleaning consultants; with, ABOUT SOIL More often than not, the soil we have to remove is mate- rial we added to enable something we wanted to happen. It can be a left-over amount--an excess beyond that which was needed. It can be something that was formed in the operation we enabled. Examples are: • Particles from a deliberate machining or polishing operation; • and particles from wear between surfaces. And these soils, which often are carriers of particles: • Metalworking or heat-transfer fluids used in finishing operations; • grease used for lubrication, either fluid or solid; • grease applied for long-term protection; • lubricating oils added to avoid wear; • rust-preventive fluids; • acid brightening chemicals; • and spent plating solutions, etc. IGNORANCE CAN BE BLISS Very often, a user doesn't have to know how to clean some unknown material from parts. Users with a TCE (trichloroethylene) degreaser probably won't spend time doing chemical analysis of an unexpected grease found on purchased materials; they'll just immerse a basket of the parts in boiling TCE. The same can be true of particles on metal surfaces. A user with an aqueous cleaning bath having ultrasonic transducers probably doesn't know or care about the chemical identity of tramp particles, how they're bonded to the metal, or even where they are located on the sur- face. They'll just immerse a basket of the parts in the water bath and turn on the ultrasonic power. This column is about those circumstances when neither of those methods of treatments works - probably because the particle isn't what it was or where it was expected. PARTICLE THEORY For a consultant, and a user, removal of particles presents the most difficult problem. Ultrasonic (20,000 to 250,000 cycles/sec or 20 kHz to 250 kHz)--generated pressure waves are omnidirectional. They reach, generally, in all directions. A user doesn't need a GPS to locate a particle so he can aim pressure waves at it. But this information-free approach probably won't suc- ceed the if particles are small. Because of their size, laws of adhesion dictate that they will be more tightly bonded to metal surfaces. Further, if these particles (submicron) are shorter than the fluid boundary layer associated with the pressure waves is tall, the waves will wash over them, leaving the particles untouched by pressure force. The limit of effectiveness is somewhere around 1 to 5 microns. The boundary layer problem is lessened with pressure waves of higher frequency (250,000 to 1,000,000 cycles/sec). Unfortunately, these waves (called megasonic) are unidirectional. Users have to aim their transducers at the particle-laden surface in order to impact these articles with pressure waves 1. Megasonic waves aren't really use- ful below levels around 1/~ to 1 micron--even if the partio cle's location is known. The problem with particles gets worse and never gets bet- ter. When particles with nano-sized dimensions must be removed (or moved), other technology 2 that requires addi- tional specific information must be involved 3. So it's worth- while to know how to manage and classify small particles. PARTICLE CLASSIFICATION SCHEMES For those trying to remove particles from surfaces, there are several types information that are significant: • Hardness: If the cleaning and process liquids do con- tain particles, their composition matters. If they're made of talc, the surface won't be worn. If they're made of diamond from a polishing wheel, it's the opposite case. (Particle hardness is characterized by the Mohs relative scale, as shown in Table I.) 54 www.metalfinishing.com

Upload: others

Post on 24-Aug-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: To Know Cleaning You Have to Know SoU--Especially Particles · To Know Cleaning You Have to Know SoU--Especially Particles By John B. Durkee, Consultant in Metal and Critical Cleaning

To Know Cleaning You Have to Know SoU--Especially Particles By John B. Durkee, Consultant in Metal and Critical Cleaning

T h e title is one of favori te (at leas t professionally) expressions. And it makes sense: i t is, a f ter all, why my every country 's mi l i t a ry opera t ion h~ it the is capable of and wha t enemy

~ t h e y are l ikely to do. I f one has a t them, one m a y have to know wha t the soil is and where i t is located in order to remove it.

In 2007, th is column will focus on soils--greases, oils, particles, and others - and wha t to do about them. Tha t soils can be so broadly character ized, c leaning so challenging. Soils a re a gift to c leaning consul tants ; with,

ABOUT SOIL More often than not, the soil we have to remove is mate-

r ial we added to enable someth ing we wanted to happen.

I t can be a left-over a m o u n t - - a n excess beyond tha t which

was needed. I t can be someth ing tha t was formed in the

operat ion we enabled.

Examples are:

• Pa r t i c l e s f rom a d e l i b e r a t e m a c h i n i n g or p o l i s h i n g

operat ion;

• and par t ic les from wear be tween surfaces.

And these soils, which often are carr iers of part icles:

• Meta lwork ing or hea t - t r ans fe r fluids used in f inishing

operat ions;

• grease used for lubricat ion, e i ther fluid or solid;

• grease appl ied for long- term protection; • lubr ica t ing oils added to avoid wear;

• rus t -prevent ive fluids; • acid b r igh ten ing chemicals; • and spent p la t ing solutions, etc.

IGNORANCE CAN BE BLISS Very often, a use r doesn ' t have to know how to clean some u n k n o w n m a t e r i a l f rom par t s . Use r s wi th a TCE

( t r ichloroethylene) degrease r probably won' t spend t ime

doing chemical ana lys is of an unexpected grease found on

purchased mater ia l s ; they ' l l j u s t immerse a baske t of the

par t s in boiling TCE. The same can be t rue of par t ic les on meta l surfaces. A

use r wi th an aqueous c l ean ing b a t h hav ing u l t r a son ic

t r a n s d u c e r s p r o b a b l y doesn ' t know or care abou t the

chemical iden t i ty of t r a m p part ic les , how they ' re bonded

to the meta l , or even where they are located on the sur- face. They ' l l j u s t i m m e r s e a b a s k e t of the p a r t s in the

wa te r ba th and t u rn on the u l t rasonic power.

This column is about those c i rcumstances when ne i the r

of those methods of t r e a tmen t s works - probably because

the par t ic le isn ' t wha t i t was or where i t was expected.

PARTICLE THEORY For a consul tant , and a user, removal of par t ic les presents the most difficult problem.

Ul t rasonic (20,000 to 250,000 cycles/sec or 20 kHz to 250 k H z ) - - g e n e r a t e d p r e s s u r e waves a re omnid i rec t iona l . They reach , general ly , in al l d i rec t ions . A use r doesn ' t need a GPS to locate a par t ic le so he can a im pressure waves a t it.

But this information-free approach probably won' t suc- ceed the i f par t ic les a re small . Because of the i r size, laws of adhesion dictate tha t they will be more t ight ly bonded to meta l surfaces. Fur ther , if these par t ic les (submicron) are shor ter t han the fluid boundary layer associa ted wi th the pressure waves is tall , the waves will wash over them, l eav ing the par t i c les un touched by p r e s s u r e force. The l imi t of effectiveness is somewhere a round 1 to 5 microns.

The boundary layer problem is lessened with p ressure waves of h ighe r f requency (250,000 to 1,000,000 cycles/sec). Unfortunately, these waves (called megasonic) are unidirect ional . Users have to a im the i r t r ansducers a t the par t ic le- laden surface in order to impac t these art icles wi th pressure waves 1. Megasonic waves a ren ' t rea l ly use- ful below levels a round 1/~ to 1 mic ron - - even i f the partio cle's location is known.

The problem with particles gets worse and never gets bet- ter. When par t ic les wi th nano-sized d imensions mus t be removed (or moved), other technology 2 tha t requires addi- tional specific information mus t be involved 3. So it's worth-

while to know how to manage and classify small particles.

PARTICLE CLASSIFICATION SCHEMES For those t ry ing to remove par t ic les from surfaces, there are several types informat ion tha t a re significant: • H a r d n e s s : I f the cleaning and process l iquids do con-

t a i n par t i c les , t h e i r composi t ion m a t t e r s . I f t hey ' r e made of talc, the surface won' t be worn. I f they ' re made of d iamond from a pol ishing wheel, i t 's the opposite case. (Part icle ha rdness is charac ter ized by the Mohs re la t ive scale, as shown in Table I.)

54 www.metalfinishing.com

Page 2: To Know Cleaning You Have to Know SoU--Especially Particles · To Know Cleaning You Have to Know SoU--Especially Particles By John B. Durkee, Consultant in Metal and Critical Cleaning

• Angulari ty. Particles with low angularity, such as spheres, were probably formed intact, and their impact with fin- ished surfaces probably doesn't change the finish level. Particles with high angularity possess sharp or acute angles between facets, were probably formed by inter-par- ticle collisions, and their impact does wear surfaces.

In other words, it is probably less important, in terms of surface protection, to remove round particles than angu- lar particles. See Figure 14 .

Particle composition and angularity can be jointly con- sidered. At one extreme, spherical plastic particles, of any size and amount, may be treated with less urgency. At the other extreme, quartz particles probably compel immediate attention.

My experience is that the smallest particles--the most difficult to remove--are the most angular because they are broken fragments of larg- er particles, and more hard because their surface has been mechanically worked as they were formed. • Size a n d size distr i -

but ion: This informa- tion is needed to design a particle removal system.

Unless process materials change, it probably needs measurement only once '~. So, contract this work with an outside lab.

Represen ta t ive da ta are shown in Table II, and plot ted on log-probabil i ty g r aph paper6 in Figure 2.

0.70

.•.60 e~ ~0.50 e-

~ 0 , 4 0

(,.} ~Eo.3o (1.

0.20

0.10

Crushed Quartz @ Silicon Alumina 3.45 Carbide @ 3.70 @ 2.70

/

D i a m o n d Garnet @ Silica G l a s s

@ 2.60 1.85 Sand @ Bead @ 1.68 0.65

Figure 1

2

6

10

p.

.e

100

10

1 0.10%

Figure 2

r = )8

J

J

t J

f m

1.00% 10.00% ~) ist i rbut ion:A3

, 100.00%

It's important to note two key parameters from the dis- tribution plot. One is the median (50% value) of particle size. In Figure 2, the value is about 8 microns. The other is the 10 or 15% cutoff point. In Figure 2, the value is about 2 to 3 microns. That difference in size is significant (see below).

• A m o u n t . Mass amount data are a component of the particle size analysis in Table IF.

Distribution and amount information can also be jointly considered. It will be used to select and design the particle removal process.

S E D I M E N T A T I O N

This is a low-cost way of separating particles from liquid. But it's not very practical in finishing operations--except as a first stage of particle cleanup.

It's based on a difference of density. Heavier particles set- tle in a lighter liquid. Simple equations govern this behav- ior, plotted as Figure 3.

Particles whose size is between 50 and 100 microns will settle in one to two minutes. It can be useful to remove this size particle if the volumetric liquid throughput does not

February 2007 55

Page 3: To Know Cleaning You Have to Know SoU--Especially Particles · To Know Cleaning You Have to Know SoU--Especially Particles By John B. Durkee, Consultant in Metal and Critical Cleaning

100

80 C 2 ._~ ~ 60

~ 40

~ 20

0 10

Figure 3

/ Time to Settle 12 inches

100 1,000 10,000 100,000 1,000,000

Settling Time Sec

Time to Settle 12 inches

1 1 ~n 8.5 7.5

.2

0.5 ~ ~ 0 1 10 100 1,000 10,000 100,000 1,000,000

Figure 5

Settling Time Sec lO0 × Enhanced

,,==,Gravity Acceleration =mAcceleration

make the settling volume too large. But it's probably not the only treatment step.

F I L T R A T I O N Particle removal or collection may start with filtration. But it may not end there.

In metal finishing work, particles remain on surfaces as debris from cutting or polishing operations. As a more finely featured surface is produced by mechanical shear forces, the produced particles are smaller and probably more angular.

Even worse, particles degrade on contact with surfaces. Submicron-sized particles are often made from micron- sized particles where there has been surface wear + .

If the polishing is done in a liquid carrier or heat transfer medium, some particles will be removed from the surface. Usually that liquid is recycled through a filter. The remain- der defines the cleaning task.

But if the filtration equipment isn't correctly sized or operated, the particles won't be effectively removed, the net surface contamination will increase, and the cleaning task worsens.

Sizing of a filter is, as is the case with most things in life, a tradeoff between quality and quantity.

As smaller particles are removed, the capacity of the ill-

100%

80%

.~ 60%

o E LU 40o£

~_ 20O1o LL

Figure 4

,,,, ' /

1 2

f

3 5

Particle Size, Microns

ter to hold them declines as does the volumetric through- put, and, therefore, useful life.

Example data are plotted in Figure 4. Note the interaction between the particle size distribu-

tion information in Figure 2 and the information about expectation for filter performance in Figure 3. The size of the finer particles does matter. They may not be removed.

S U B - M I C R O N P A R T I C L E S Note in Figure 3 that if the median particle size is below about 1 micron, filtration 9 performance won't be satisfac- tory 1°. In that situation, filtration isn't the answer, and neither are pressure waves produced by ultrasonic trans- ducers. Probably, pressure waves produced by megasonic transducers aren't, either.

At least five unusual and specific ways have been used to recover submicron particles. They all involve custom engi- neering and details are beyond the scope of this column: • A col lec tor fluid: This can be an emulsion or a viscous

fluid into which the particles are chemically or electro- statically attracted and then collected. More angular particles, with more surface area for bonding, will be easier to collect. Naturally, this (probably unusual) fluid becomes an additional soil for cleaning operations.

• Centr i fuga t ion: This is sedimentation with centrifugal force providing enhancement of particle separation from fluid vs. gravitational force. But little relief should be expected with centrifugation

of submicron particles. Figure 5 shows that a 100X increase of separation force has little effect in this range of particle sizes. Compare to Figure 3. Centrifuges excel at removing large chips produced by machining opera- tions, but not surface polishing operations.

• E l e c t r o s t a t i c p r e c i p i t a t i o n (EP): Not commonly used in cleaning operations, there are two versions of this approach. Basically, EP involves charging the parti- cles in a liquid by moving them adjacent to an element, which "radiates" an electrostatic charge. Particles close to the element take on this charge. When the liquid

56 www.metalfinishing.com

Page 4: To Know Cleaning You Have to Know SoU--Especially Particles · To Know Cleaning You Have to Know SoU--Especially Particles By John B. Durkee, Consultant in Metal and Critical Cleaning

. . . . . . : :f:,: ~: :: : :

~ : ~ i ' : ~ ] : I : : : : l i :~ : : : : : : : : : : : : ~ : : : : : : : : : : : : :

flows near an oppositely charged (grounding) element, the particles are attracted to it. Obviously, this can't be done in aqueous media. And it's

not a good idea to attempt it in cleaning solvents with a low dielectric constant such as cyclohexane or paraffins 11. But it has been done in formulated lubricating/cutting/grinding oils. A clever variant on this approach is to split the fluid

stream into two portions. Each then passes an opposite- ly charged element. When the stream is recombined, the oppositely charged particles attract and agglomerate. The larger particles can then be removed from the liq- uid by the means described above 1~.

• Bubbles. This approach is significant to those involved with surface finishing at the nano scale. To this author, it's still empirically and not theoretically defined. The basic idea is that pressure pulsations (not necessarily produced by pressure waves) cause collapse of gas (not vaporized liquid, not water vapor) bubbles at surface imperfections 13. This is how nucleate boiling is done. Only here, the surface imperfection is the particle! So one doesn't generally care about the size of the par-

ticle or where it's located. The particle is removed by the mechanical force liberated when these bubbles collapse. Note that this approach doesn't collect (as does a filter)

submicron particles for removal from apparatus or parts, it just removes them from a surface. There is no guarantee they wont' be redeposited!

• P reven t ion . If cleaning is required, one has to know what it is that one is cleaning, then the best way to do cleaning is to eliminate the soil (particle) before it becomes a concern. Often this is practically impossible. But also often the information above which character- izes particles provides clues as to how tiny particles are formed. Polishing surfaces with submicron diamond grit sus-

pended in a mix of viscous oils and fatty acid amides or with fine alumina in a water-based suspension is cer- tain to produce a smoother surface, and submicron debris. But if the submicron particles are found to be formed in some other manner, then prevention may be possible.

tamination, 92(563); 1992. 2. A good example is found in U.S. Patent 5,821,175

which can be an excuse to try the new Google patent server at http://www.google.com/patents. Finding spe- cific patent information has just gotten easier!

3. As indicated in this column, manipulating nano-sized objects will be one future aspect of metal finishing work.

4. Data from Fitch, J., "A Much Closer Look at Particle Contamination," Practicing Oil Analysis Magazine; September 2005. The horizontal axis labels display the material and its relative wear rate.

5. While this measurement may be made only once, that doesn't mean it's done with only one sample-- only one analysis.

6. This graph paper can downloaded as a PDF file for free at http://geolab.seweb.uci.edu/19.pdf.

7. If the information in Table II is produced by a filtration or sieving scheme, the mass amount will be directly known. If the size and distribution information in Table II is produced by optical means (light scattering), the mass amount can be estimated from the reported num- ber of particles and the size and distribution information.

8. The science of wear management, including manage- ment of tiny particles, is called tribology. Those engi- neers and scientists practicing tribology may be involved in lubrication of engines or bearings, design of prosthetics, or design of hard disks.

9. Filter media for removal of particles of submicron size can be useful. But their service life will be short, they will be easily blinded (fouled), and their removal efficiency will be low. Removal of submicron particles with pressure waves or filter media is a challenge best avoided.

10. It is commercially possible to remove particles as small as ~0.1 micron from liquids. It's routinely done by firms that conduct medical research, make semiconductors, work for NASA, and the like. But it's a multiple-stage process, done in a controlled environment (cleanroom), not done with low cost, and supported by a QC lab.

11. Water transmits (conducts) electrostatic charge and allows it to dissipate. Cyclohexane does not; it holds electrostatic charge because it's not conductive. Released charge in a low flash point solvent can pro- vide ignition to a fire.

12. See http://www.isopurfluid.com/bcp.htm. 13. See U.S. Patent 6,418,942 and its successor 6,824,620.

S U M M A R Y Is it possible to make a problem appear too complicated? Absolutely! But generally, more and better information about a problem makes for a more certain solution for it. Knowledge of soils is central to solution of cleaning problems.

REFERENCES 1. This has been accepted for more than a decade. Good

references are Werner Kern, Journal of the Electrochemical Society, 137(6):1887; 1990 and Ahmed Busnania and I. I. Kaskkoush, Proceedings, Micro-con-

John Durkee is the author of the book "Management of Industrial Cleaning Technology and Processes," published by Elsevier (ISBN O- 0804-48887). He is an independent consultant spe- cializing in metal and critical cleaning. You can contact him at 122 Ridge Road West, PO Box 847, Hunt, TX 78024; 830-238-7610; (fax) 612-677- 3170; or (e-mail) [email protected].

[nf

February 2007 57