Volume 2 Number 5 Dennis R. Dinger 1 March 2004
An Update
Rheology for Ceramists and Particle Calculations for Ceramists can be purchased at the Books and Downloads page of the web-site. Quantity discounts are available on the paperback books. If interested, please contact me for details.
Please forward any issues of the E-zine, or my web address, to any ceramists or materials engineers who might be interested. If this is the first issue you've seen, you can add your name to the mailing list by clicking HERE. All back issues can be accessed from the Publications page at the web site. Questions, suggestions, and/or requests for topics to be covered in future issues of this e-zine can be sent to QuestionsandComments@DingerCeramics.com .
A comment made to me recently suggested this topic that applies to most of you.
Raw Materials Consistency -- Who's Responsible?
This question comes up frequently when I teach short courses and consult. Some engineers simply don't want to do the characterization and control of their raw materials. They want the suppliers to do it. In other words, they don't want the responsibility for the day to day consistency of their bodies. But if they're production engineers in charge of a process, they've got the daily responsibility whether they like it or not. They can't push it off on anyone else. In this article, I want to weigh in on this subject.
One engineer wanted me to go into the business and produce his body formulation so he could buy it from me. There are two problems with this: (1) In my opinion, this is the toughest part of any production process; and (2) If I did this, I'd be responsible every time he had production problems. Since I believe body control (#1 on this list) is the toughest part of the production process, if I'm going to make body (and do the tough part), then I'm also going to produce the whole product (and do the easy part) as well. And if I produce the product, then I'll invariably be responsible anyway when the process goes awry (which was #2 on this list.) But in this case, I'll be responsible to my own company for my own body control and my own process over which I will have control.
In my mind, body control is a necessary part of the production process. And it can't easily be pushed off onto anyone else.
But I don't want to be a ceramic producer, or a raw materials supplier. I want each of you who are ceramic producers and raw materials suppliers to do your own jobs well. Where possible, I'll help each of you meet those goals.
Here is my thinking on this subject:
The Production Engineer Is Responsible & Here's Why
I have visited a variety of raw material suppliers' locations, and as sophisticated as the controls on their raw materials may be, they are coarse controls relative to the fine tuning of processes required at the ceramic producers' sites. No offense to the suppliers in the audience -- this is not a criticism -- just a statement of fact.
By trying to push all materials controls for production processes off onto the suppliers, the customer is asking them to do fine tuning that they are simply incapable of doing. It isn't that they don't know how to do it or don't have the capabilities to do it -- it's because they don't have all the detailed, day-to-day, process performance information they need to tweak their raw materials to produce perfectly constant process performance of each of their customers' processes. How is a supplier supposed to adjust his material to produce the appropriate rheological properties of a customer's body in the customer's plaster casting shop and in the customer's pressure casting shop (which possibly requires two different body adjustments)? When the ceramic producer has a forming problem in an extruder or in an injection molding system, how is any supplier supposed to adjust his powder for that?
The other important factor to consider is that most bodies consist of several raw materials from several suppliers. How can I (as supplier A) be expected to always produce appropriate materials to work well with materials supplied by companies B, C, D, E, and F? If the answer is, "Always supply the same, identical material!", that's a dream because it can't be done. No matter how closely controlled any raw material is, it will always have fluctuations in all of its properties -- especially in its particle size distributions.
Let's say that all six suppliers (Companies A - F) supply perfectly consistent materials and the production company still has processing problems. Who's to blame? Who's going to solve the problem? I once had a computer problem of this sort. The day the software guy showed up to install the software (this was 25 years ago when computers were still in their infancy), the hardware didn't work. A week later, when the hardware guy showed up, all of the computer's components worked fine. A week later, when the software guy returned, the hardware didn't work again. Each said the problem was within the other guy's territory. And in the mean while, we were getting nowhere! I had to insist that both guys show up on the same day to get the system fixed. But this solution doesn't work with ceramic processes. The six representatives from Companies A through F are not responsible for a production body. They can't be held responsible for fixing a production problem. The local production engineers and their staff are responsible. (If it can be demonstrated that one of the raw materials is out of spec, that's a different story. The assumption here is that all powders are perfectly in spec, but when combined, the body has processing problems.)
Raw materials suppliers do an excellent job of producing consistent materials day after day, year after year. I applaud their efforts and skill. (Actually, I think we Americans are spoiled because we have good raw materials and good suppliers. I've seen some of the poor raw materials that good suppliers in other countries must deal with.) But consistency from the supplier does not mean that all raw materials are totally, exactly, perfectly consistent all of the time. It simply won't happen and it should not be expected to happen. Meeting specs should always happen. Total, exact, perfect consistency will not.
Suppliers' production personnel have no idea what their customers are going to do with their raw materials after they arrive at the production sites. The person in the kaolin plant who controls the coarseness/fineness of the air floated kaolin product has no idea whether the current batch he/she is producing will end up in toilet bowls, dinnerware, automobile tires, paper, etc. He has no idea how his product will be processed after it arrives at the customer site, what properties it will impart to a body, or how tiny fluctuations of any property in his material will affect any particular property in the ceramic production process. His whole world is the milling circuit he's controlling, and that's as far as his contribution goes. He takes great pride in producing consistent product, and he's credited and rewarded for it (we hope) when he does an outstanding job. But it's not his fault when a customer's dry or fired shrinkages are off on any particular day. The same can be said of all other employees at raw material suppliers' sites.
The only persons who really have all the fine details of a production process are the production engineering staff (which sometimes is one person). He's the person whom I suggest is responsible for day-to-day body consistency. (I'm using the word he but in all cases, it could be a he or a she.) He needs to work closely with the batch house supervisor to make sure that the body is adjusted and controlled properly to meet process requirements and produce the intended properties. He's responsible for determining which of the raw materials is at fault if there is a production problem and for making appropriate adjustments to eliminate the problem. If he's fortunate, it will be a single raw material that's causing the problem. Otherwise, it will be several raw materials, or the combinations of (or interactions between) several raw materials in a batch that are causing problems. In any case, a very good understanding of the process, its variables, and all interactions between variables will help to eliminate processing problems. Such details and skilled understandings of the processes are only available to the local production people and particularly to the process engineer in charge of the process. Such problems must be solved at the local production level.
There are considerations that must be made, however, with each form of raw material used. We will consider three general categories of materials: sedimented materials such as clays and kaolins, materials that are milled from large rocks and pebbles, and chemically prepared materials. We will look at each of these separately.
Sedimented Minerals Such As Clays And Kaolins
Most clays and kaolins used today are sedimented materials. Generally speaking, the particle sizes of these minerals are not reduced during processing, even though mills are used. Mixing and/or milling of these materials tends to deagglomerate and disperse the particles into carrier fluids if they are shipped as slurries. If these materials are dry processed, they are dried, consolidated, agglomerates are reduced in size, and the materials are shipped as dried powders (and dried, small agglomerates). Few if any particles are broken or reduced in size from the original sizes that existed when they rested untouched in the seam. Some particles (not all) are liberated to report as individuals in slurries. Many particles are compacted during the drying and air-floating process, but individual particles mostly retain their original sizes.
As sedimentary deposits formed, particles flowed downstream over long periods of time to then sediment in quiescent pools. As a result, such deposits are mixtures of particle sizes -- when streams flowed slowly, fines floated along with the current. During flood times, larger particles (boulders) flowed along as well. All of the particles then sedimented and were densified into the clay and kaolin seams that are mined today.
The characteristic particle size distribution moduli, n, for clays and kaolins tend to be small and even negative. Negative distribution moduli indicate that there are more particles in the fine and colloid size classes than in the coarse classes. This is opposite, by the way, to the results achieved by milling -- where most particles are in the coarser size classes and fewer particles are in the fine size classes.
Distribution Modulus Review
By way of review, the D-F equation is:
CPFT/100% = (Dn - Dsn) / (DLn - Dsn) (1)
where D = particle size, Ds = smallest particle size, DL = largest particle size, and n = distribution modulus.
Most dry-milled particle distributions have moduli, n, greater than about 0.6. Non-first-order milling can produce distributions with moduli from ~0.2-0.6. Mixtures can produce distributions with any moduli (which includes distributions with moduli <0.2 and distributions with negative moduli.) The fact that clays and kaolins can have negative distribution moduli is consistent with the idea that all such deposits are mixtures of powders.
Clay and kaolin suppliers do an excellent job of exploring their formations to know the general properties and PSDs (particle size distributions) of the minerals present. Some companies remove the minerals very selectively from the formations to keep similar materials together. Some companies just quickly remove all the minerals in the order in which they're exposed and then they carefully mix and blend to achieve desired products. Regardless how they do it, the companies all offer various types of clays and kaolins that they control within tight specifications. Most have been offering the same trade name materials for years.
For example, if I was in the business, I might have offered a Dinger Special Grade back in the 1960s. For those customers who use this material, I would be offering a Dinger Special Grade today with essentially the same properties this material has had over the last 40 years. The current seam and excavation site might be miles away from the original seam and excavation site, but the materials shipped will be very similar. Actually, today's companies have remained in business because they have successfully offered a variety of named clays and kaolins with consistent properties throughout all their years of existence.
Keep in mind that clays and kaolins in the seams are mixtures. And as each seam is depleted, the companies continue to move along, open, and mine other seams in their reserves. They do this for each of their different named materials.
Milled Materials
All minerals that are mined as huge chunks of rock or as finer pebbles, which are then crushed, milled, and then sized to produce the desired particle sizes fall into this category.
As mentioned above, distribution moduli, n, (see Equation 1) for dry-milled materials and for low solids content wet-milled materials are usually >~0.6. Wet-milling conditions can be adjusted to achieve lower distribution moduli. When all else fails, it may be necessary to mix distributions to achieve desired PSDs.
Suppliers are responsible for finding mineral deposits with consistent properties and then mining, beneficiating, milling, and sizing them to the PSDs required by customer companies. They do this very well. Their products are generally very consistent over time. If they have sizeable reserves of reasonably pure minerals, they mostly need to mine, crush, and size them to produce the various particle size grades. If the desired mineral, however, is mixed with other minerals in a formation, suppliers need to beneficiate and separate the various minerals before crushing and sizing. In this case, the suppliers' task is more difficult (not impossible, just more difficult.)
Chemically Prepared Materials
Lest those of you who use chemically prepared materials start to think that you're off the hook, forget it! Chemically prepared materials have similar problems -- and in some cases, they're worse.
When a batch of material is made in a reaction vessel, a little variation in time of reaction will change the product. If it's a polymerization reaction, polymer chain lengths will vary. If it's a precipitation reaction, powder crystallite sizes will vary. Slight variations in time of reaction, mixing intensity, homogeneity, chemical ingredient concentrations, and firing/heat treating conditions, for example, will all cause slight variations in product consistency.
When several companies sell the same powder, it is likely that they will all use different processes (because they each have their own patents). This means each of the powders will be fundamentally different, and each will contain different surface impurities as a result of the specifics of each process. Can each of these powders be freely substituted for one another? One would hope so, but not necessarily. An engineer once told me concerning a well known, chemically prepared ceramic powder, "There are only three suppliers of this powder in the US and each powder can be identified by analyzing for its impurities." By impurities, he meant surface residue on the powder but impurity atoms could also reside within the crystal structure itself. If different chemical residues are present on the powder surfaces, each will behave slightly differently during processing because each of the surfaces will be slightly different. Atomic substitutions and impurities within the structure may not cause processing problems, but they can significantly alter other material properties.
If you think the surfaces of your chemically prepared powders are perfectly clean, think again. Ask yourself this question: If you were chemically preparing a powder for sale to a customer, how much water would you use to wash the powder prior to shipment? Let's provide another detail: Add the fact that you know that 10 gallons of water will produce an excellently washed pound of powder. Now, how much water will you use?
If 5 gallons does a pretty good job, the product meets specs, and the customer doesn't complain -- is 5 gallons sufficient? If, after some experimentation, 2 gallons does a fairly good job -- is 2 gallons enough? I suggest that in today's world of cost reductions, the least possible number of gallons of water will be used that still allows specs to be met and that produces no customer complaints. The less wash water is used, the more surface impurities will be present. This reasoning applies not just to washing, but to all processes.
What impurities could possibly be in your raw materials that will cause problems? Will impurities on different raw materials interact? Will different impurities behave similarly? Knowing that a particular powder could have a surface chemical impurity means you can analyze to determine the presence of that chemical. If a problem arises but the specific chemical cause is unknown -- and the impurity chemical could be literally anything -- good luck!
Possible Problems and Examples
Additive chemicals used in suppliers' processes can cause problems for production processes. If such problems arise, the reason is that the specs weren't written properly. If ceramic producers check and control all appropriate powder properties, this should not be a problem. If ceramic producers have missed or ignored some variables on the specs, however, this can be a major problem.
Three examples come to mind. One has to do with chemistry. One is a PSD issue. And the third is an impurity issue. In all cases, specs were not tight enough to identify the problems. Why? None of these problems were foreseen.
Coal Slurry Example
This happened to us on the coal slurry project at Alfred University in the late 1970s - early 1980s. We were making coal/water slurries (coal powder plus water plus chemical additives.) We had sent one of the guys with a pickup truck to West Virginia to bring back a truckload of coal to use on the project. After he returned, we had a nice supply for quite some time.
A little more than a year later, we ran out. It was November. We called and made arrangements to go pick up some more coal, but they said, "No need -- we ship that same coal by rail to Buffalo, NY. We'll give you the location in Buffalo to pick it up. That will save you a long drive." So we sent the same guy in the same truck to Buffalo to get the coal.
The original coal made wonderful slurry. The new coal made terrible slurry (in fact, it didn't really make slurry at all.) The two coals were different like night and day. We had a well-equipped lab so we carefully characterized both coals. All tests showed the two coals were identical. We called several times to the coal supplier to find out what was different about the new coal. "Nothing!" was the answer. "They're identical."
We weren't getting anywhere and I was getting exasperated, but I continued to try. Refusing to believe there were no differences, I called again. I don't remember how many people I talked with, but since I insisted there was a fundamental difference, I kept pushing to identify it. (For those of you who don't know me, I'm a stubborn Pennsylvania Dutchman -- and I live up to that characteristic whenever necessary.)
Finally, someone said, "Wait a minute! It's November!" Here's what we finally learned: Companies that use hundreds of cars of coal every few days don't empty hopper cars using the chutes in the bottoms of the cars. They grab the cars in huge clamps, roll them upside down, and dump them. In the winter, to prevent all the coal from freezing into one giant ice cube which would not release from the car and fall out on cue, they spray the cars as they leave the mine site with CaCl2. It's routinely done. It's so routinely done that hardly anyone pays any attention to it.
By the way, calcium chloride is the material dumped by the ton on the roads up north in winter to minimize ice formation and to turn all vehicles into rust buckets. Calcium chloride is also a wonderful flocculant. In our case, there was enough calcium chloride on the coal to prevent us from making slurry. By design, our slurries were almost totally deflocculated, and it's almost impossible to produce total deflocculation in the presence of excess calcium chloride (assuming, of course, that one realizes that the calcium chloride is present -- which we didn't.) In the summertime, they obviously don't spray all railcars filled with coal -- and when you pick the coal up at the mine site by pickup truck, it doesn't get sprayed either.
After learning this, we washed and washed and washed the coal and then we were able to use it. But to learn this, we had to be persistent and not take "We didn't treat it differently than your first shipment!" for an answer. There obviously was a difference. They had used a chemical (that was poison to our process) to insure they could empty the cars at the delivery sites. Calcium chloride doesn't hurt air-swept ball milling or combustion processes at power plants. Another point to note is that calcium chloride spraying was such a common process that those who knew about it didn't think it was anything unusual, or important, or even relevant to our questions. The other point to note is that it was the last thing done to the coal as the rail cars departed the loading facility. It's also possible that many coal company employees didn't even realize it was being performed.
This was an example of a chemical problem that was really difficult to identify. We didn't realize the addition of this chemical was possible, so we had no reason to look for it or to write a spec preventing its use. Such a spec would certainly be necessary if we were using this coal in a full-scale coal slurry production facility. And this spec would definitely cause railcar unloading problems during the wintertime.
Excess Fines Example
This example is a PSD problem. This was Jim Funk's story. I don't know the names of the engineer, the company, or the supplier -- so it's safe for me to tell it.
An engineer at a ceramic production company determined through experimentation that the fines in one of his ingredient powders were hurting his process. So he approached the supplier and they said they could supply him powder without the fines. So they separated out the fines and shipped him only the coarser materials. His process then worked well. But guess what they did with the fines they removed from his materials? They added them back into everyone else's materials. 'Everyone else' included some of his competitors who also had problems similar to his original problem. All of a sudden, his competitors had terrible processing problems -- for no reason known to them.
I should add that typical PSD specifications at that time were something like this: All material must be 90% finer than 100Mesh. Notice: It doesn't say how many (or how few) fines or colloids were needed. It doesn't specify mean particle size. It really doesn't specify much at all. It just said that most of the material had to be smaller than a certain size. This is a spec you could easily drive an 18 wheeler through. So the fines not shipped to this particular company were easily within the specs used by the other companies -- so they received them.
The spec on pulverized coal is of this type. Fine pulverized coal particles burn very well. It's the coarse particles that cause combustion problems. A specification like this insures sufficiently few coarse particles to minimize firing problems. But that's pulverized coal. If any of you in the ceramics industry are still using specs like this, they really need to be changed.
To solve this type of problem, better PSD specs must be used. In fact, a spec on SSA (specific surface area) should also be used.
Impurities
This third example is an impurity problem that showed up in a ceramic final product. After much study, it was learned that the wires used to link the explosive charges in the mining operation were the cause of the problem. Copper oxide impurities vs aluminum oxide impurities in one raw material caused the problems. The impurity was directly related to the use of copper vs aluminum wires in the field during mining operations. Even though this was a relatively minor impurity, it nevertheless caused problems. And I understand that this was a very difficult problem to solve. At least when iron specs show up on a product, one can check the steel trusses in the plant. Where does one look for a source of copper oxide impurities?
When A Change Is Announced
In some cases, changes by a supplier will be announced. An excellent working relationship between customer and supplier can go a long way to working out any difficulties in such times.
Sometimes, changes are announced with the express purpose of assuring customers that consistency will continue to be fine. When this happens, customers need to be paying attention.
In one such case, the workers at a plant went on strike. The company announced they would continue to fill orders, but they were going to fill the orders from a different plant which was being run by its managers. This announced several different changes which were about to occur -- different mining location, different processing facility, and different workers. Although the product continued to meet specifications, it behaved differently. Although this was the 'same material' (according to the specs), it was different. And it showed that the specs being used were rather loose.
When such "Heads up!" announcements are given, prepare to pay close attention to incoming materials.
General Considerations & Questions To Ask
Here are some thoughts that apply to any powders and any of the situations discussed above. These are only some of the obvious questions to ask and thoughts to consider.
Regarding whether today's powders are identical to the same powders used over the years: Are the suppliers' mining, beneficiation procedures, and process equipment the same today as 40+ years ago? Sometimes, yes -- sometimes, no. Today's backhoes are similar to but different from the shovels used 40+ years ago. Some new production equipment, however, is still similar to that used 40+ years ago. Today's tanks and blungers are similar to old tanks and blungers. Are new devices and new improvements being added as they become available? Yes. Will any product today be identical to that sold years ago? ... identical, no -- very similar, yes.
Today's workers are different from those working 40+ years ago. Are the same engineers in charge today as 40+ years ago? Probably not.
If a supplier makes a process change, will customers notice it? Some may, some won't. If for example, a company uses high intensity dispersion on a clay, the new HID clay will be very different from the non-HID clay of the same name. Customers would notice the difference. Dinger Special Grade clay would behave very differently from an HID-Dinger Special Grade clay. This would be a major change and suppliers would not substitute an HID material for a non-HID material. They would label the new clay with either a new name, or with the HID label.
Small, routine changes would be, for example, changes in dispersant concentration. More substantial changes would be, for example, a change to a totally different dispersant. The important question is: How big a "small, routine change" is necessary to make it into a "substantial change"? Just because a production company was not notified of a change does not mean that one didn't occur. This actually happens frequently. Even small raw material production changes can be big changes to some customers.
If a company is supplying powders that meet a set of tight specifications, customers should not see major differences from shipment to shipment. Normal day-to-day fluctuations will be present, but major fluctuations should not occur. If there are production problems, it may be (and usually will be) because the specifications ignored one or more important parameters. That is, some parameters that were ignored should have been included in the spec.
When companies update laboratory equipment (and this happens routinely), customers shouldn't notice any raw material changes. Some analyzers, however, operate based on fundamentally different principles, but they give results in the same form. For example, consider particle size analyzers. Sieves, sedimentation, laser scattering, and Coulter method analyses all produce fundamentally different descriptions of particle size. My pencil will fit through a 1/4" sieve, but would anyone give the characteristic size of 6" long pencils as 1/4"? Sedimentation devices report equivalent spherical diameters based on settling phenomena (Stoke's Law.) Laser scattering devices report equivalent spherical diameters based on light diffraction physics. Coulter principle devices report equivalent spherical diameters based on electrical resistance phenomena.
All of these techniques report particle sizes, but all use different definitions of size. Will any of these analyzers substitute one for one with any others? Since they are particle size analysis techniques, they should. But obviously, since they are all based on different measurement techniques, they won't correspond exactly to one another.
An upgrade in a lab from an old instrument to a newer instrument may also mean a change of analysis technology. A new sedimentometer should produce results similar to those from an old sedimentometer. But a new laser scattering instrument will not produce results similar to those from an old sedimentometer. All sizes reported by the new instrument will be shifted slightly from the same sizes reported by the old instrument. The shift will correspond to the differences in the definitions of particle diameter. If a company replaces an old particle size analyzer with a new particle size analyzer, they may assume it's a one-for-one swap and never think anything of it again. This applies to suppliers and production companies alike.
What happens when a production company uses four different ingredient powders and each powder supplier uses a different type of particle size analyzer to report their PSDs? Other than their own powder, each supplier doesn't know whose powders a customer is using (nor should they, nor need they). Each production company needs to be able to compare apples with apples when putting the various powders together to form production bodies. And I suggest this means that each production company should perform all characterizations of ingredient raw materials in-house. Problems at this stage of the production process simply cannot be pushed back upon any supplier.
The Bottom Line
The bottom line is this -- each production facility should be characterizing all incoming raw materials with the techniques and instruments that the production company's engineers have determined work well with their process. If even a minor change occurs in an incoming raw material (totally within the limits of the specs) -- the production facility will detect it and accommodations can be made. The detailed characteristics that control all fine processing properties must be known and should be adjustable. I believe this should be the standard operating procedure (SOP) in all production plants.
There will always be raw material fluctuations in all processes that production engineers must contend with. To accommodate the fluctuations, all important properties should be measured and controls should be implemented to produce consistent behavior during processing. This is the responsibility of the process engineer and his/her staff, in whose care and keeping resides the successful day-to-day output of their company's products.
Miscellany
I've received some good suggestions for future column topics. Please continue to send your ideas. Thanks.
Until next time ...
Thanks for signing up to receive Ceramic
Processing E-zine
To unsubscribe, click the "Remove Me" link below.
Copyright © 2004 Dennis R Dinger
103 Augusta Rd, Clemson, SC 29631 (864) 654-3155
All Rights Reserved.
All Rights Reserved.