Volume 2 Number 8 Dennis R. Dinger 1 June 2004
An Update
All Short Courses scheduled for June, 2004, have been CANCELED for lack of sufficient enrollment. My apologies to those who planned to attend. If any of you continue to be interested in attending one or more of these courses, send me a note stating your interest and suggesting a good time to offer the course. Be aware, also, that each of these courses can be taught in-house at your site. In-house courses are offered for 10 or more people at any given location. Contact me if interested.
If this is the first issue of this E-zine that 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. For those of you whose e-mail programs don't properly show the figures in these E-zines, go to the Publications page of the web site using your web browser to open any and all issues. All figures should open properly when issues are accessed from 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 .
The two books, 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. Downloadable versions of each book are also available at the web-site.
As with all other issues of the E-zine, please forward this issue to any ceramists or materials engineers who might be interested. Or simply point friends and associates to the Dinger Ceramics web site.
This topic is the result of an excellent suggestion from a reader.
Laboratory Tests for Day-to-Day Characterization
Depending upon the approach used to control processes in a plant, one might use different instruments to characterize day-to-day production batches. This explanation applies to a plant that has implemented PPC (Predictive Process Control) techniques for all process control activities. This explanation applies to the ideal case where all necessary bins, hoppers, and tanks are available to fully implement PPC. Process engineers can compare their current practices to those described in this article.
Raw Materials
Bulk Storage
Each raw material should be stored in clean, separate storage areas which minimize contamination from other raw materials. Incoming shipments should be tested for acceptance on arrival at the plant. All important specifications should be tested. These tests should include particle size distribution, specific surface area (SSA), methylene blue index (MBI) for clays and kaolins, and wet chemical tests to identify and quantify soluble ions that travel with the powders. When there are especially critical properties of raw materials that are known to affect production performances, those properties should be tested.
Water Supply
Based on experience, the water supply used in all batch operations should be routinely tested. What soluble ions are in the water supply? Is all batch water deionized? Are water treatment columns (de-ionizing columns) functioning properly?
General Comments
Each supplier has their own schedule of routine tests used on their materials prior to shipment. Production companies should test incoming raw materials based on their experience with each supplier, and on their knowledge of the particular instruments, techniques, and sampling methods used by each supplier.
Several reminders: All properties of all raw materials vary from day to day. Sampling techniques which produce representative samples (or non-representative samples) affect the accuracy and applicability of measured properties. Results from fundamentally different instruments measuring the same property of a material will also vary.
Good supplier/customer relationships generate trust which can minimize the extent of acceptance testing. Information should be passed from supplier to customer detailing the procedures used by the supplier (sampling, types of instruments, etc.) When one type of instrument is used by the supplier, and a different type by the customer, that property should be carefully measured by the customer so local results are comparable to all other local tests. For example, when the supplier measures particle size distributions (PSD) using an automatic sedimentation particle size analyzer, and the customer uses a laser-scattering PSD analyzer, PSD analyses of all raw materials should be performed on site so all PSDs at the production site are comparable. This applies to other important analyses as well.
Tests in Preparation for Batching
PPC procedures require the measurement of all important properties prior to batching. With this information in hand, one can then predict the values of properties in production batches comprised of mixtures of ingredient raw materials. The ideal way to implement this is to set batch-sized portions of ingredients aside in bins, hoppers, carts, tanks, etc., and then to characterize each carefully. Three portions of each allows the following daily preparation procedures: one portion is measured and set aside, one portion (from the previous day) is characterized, and one portion (from two days prior) is used in the daily batch.
Once again, sampling is important. Since batch portions are considerably smaller than whole shipments, good sampling procedures can provide valuable and specific details of each portion. One example comes to mind: A suspension of a material was shipped to the customer in barrels. The supplier sampled and characterized the contents of each barrel and provided that information to the customer. The customer, however, employed no special procedures to insure that barrel contents were well mixed before dividing the contents into small plastic containers, nor did they employ special procedures to insure that the contents of each small container were similar. So the smaller containers of suspension at the production facility contained suspensions with widely varying properties -- all of which, however, were assumed to be perfectly equivalent to the measured contents of the larger shipping containers. Many (if not all) contained suspensions that were very different from the average, well-mixed suspensions contained in the larger barrels. This might not have posed any significant problems had the customer sampled and analyzed the contents of each small container. But they did not. They just assumed that each container held suspension with properties identical to the suspension in the larger barrels. Process results were poor (at best).
It is difficult to divide large volumes of raw materials into representative smaller batches with identical properties. Some variations will always occur, regardless how careful the preparation techniques used. But variations are less important when each smaller batch is characterized individually. Yes -- this requires lots of analyses, but it removes the guesswork.
If one wants to implement PPC to the greatest effect, one will prepare and keep on hand not just 3 portions of each material (as described above) but 5 portions (or more, when possible): each day one portion is prepared, one portion is characterized, and three portions (or more) are available for use. If each of the three that are available for use have slightly different properties (as expected), any one or combinations of the three can be used to produce the exact properties of each material required in each batch. Of course this depends on the spread of properties of the 3 portions about the mean of each target property. If we use PSD as an example, if the three portions contain one that is coarser, one that is about right, and one that is finer than the target PSD to be added to each batch, materials from each of the three can be mixed to closely approximate the target PSD. If all three portions are coarser than desired, or all three portions are finer than desired, it will be impossible to achieve the daily target PSD. But since each portion is well-characterized, the production people will know that yesterday's portion, which was characterized today, is too coarse (or too fine) for tomorrow's batch. And with this knowledge in hand, attempts can be made to make the next portion finer (or coarser) to compensate.
As another example, when a specific surface area (SSA) of 6m2/g is the target, if each bag of powder has been sampled, analyzed, and labeled with the SSA of its contents, bags labeled 4m2/g can be matched with bags labeled 8m2/g, 5m2/g with 7m2/g, 6m2/g with 6m2/g, etc., to achieve target batch values. The alternative is to randomly select bags for each batch, mix the powders, and then test the final batch properties. This alternative, however, is not control -- it is a form of gambling.
Without the many analyses of all important properties required by PPC, one simply measures out portions of each raw material, mixes them into the day's batch, and then eventually tests the batch properties to learn what has been achieved. When batch tests do not meet desired specifications, everyone scratches their heads and says, "I wonder what happened?" When the properties do happen to meet specs, everyone says, "Boy, we did a really great job today!"
Implementation of PPC requires many bins, hoppers, tanks, etc., and lots of characterizations of all ingredient materials. But when PPC techniques are well-implemented, guesswork becomes a thing of the past, and batch properties are well-controlled. Control will not be perfect, but it will be as perfect as can possibly be expected.
Ingredient Properties to Be Characterized Daily
Four major properties should be measured on each ingredient on a daily basis. During batch preparation, three properties should be routinely measured and tightly adjusted and controlled. Other minor properties may be monitored as required by specific body requirements, but certainly, these following important properties should be measured and controlled in all applicable bodies.
Particle Size Distribution (PSD)
PSD is the most important property that should be measured and controlled in all ceramic process systems. The term PSD does not just mean an average size, or one, two, or even three size parameters that could be used to define a distribution. PSD refers to the detailed analysis of the whole distribution. The word whole refers to a complete square root of two or a complete fourth root of two size classification that covers the whole range of particle sizes from coarsest to finest.
In the past in many plants, PSD has meant only one or two parameters that characterize a distribution. For instance, coal slurries were required to be 98.5% finer than 300 micrometers. That was it! That was the only specification! If this specification were a hole, you could drive an 18-wheeler through it. But that was the specification. Period! And many ceramic companies had similar PSD specifications. Why? In the case of coal combustion systems, particles larger than ~300 micrometers cannot burn in the time allowed for particles to enter, fall through, and burn in power plant combustion chambers. All smaller particles can burn completely in this length of time. In the case of coal combustion, this specification shows that coal combustion technology does not consider the PSD of small particles to be important. They only need to insure that there are not many large particles, unburned remnants of which may eventually build up on the floor of combustion chambers. But when slurrying coal, or when making ceramic forming bodies, the whole PSD is important because rheological properties depend on the whole PSD.
In my opinion, ceramic companies don't have any good excuses for similar specifications on their powders. Most ceramic bodies are processed as slips and plastic forming bodies in which rheological properties are critical! The whole PSD is therefore important in all such ceramic process systems. Years ago, it was difficult and time-consuming to measure PSDs and SSAs, but with today's automatic computer-controlled instruments, that excuse is not acceptable.
Another characteristic of particle size distributions is that they are difficult to modify once powders have been added to batches. The time to control PSD is prior to batching. For this reason, one should have the best possible PSD information available before batching each day's ingredients.
Considering the capabilities of today's computerized PSD analyzers, complete PSD control should be SOP (standard operating procedure) in all plants. Any company which is not taking advantage of modern PSD capabilities is at a distinct disadvantage to their competition.
Specific Surface Area
The second critical property in all ceramic process systems is specific surface area. Chemical additives modify viscous and rheological properties by adsorbing onto particle surfaces. So to provide consistency from day to day of viscous and rheological properties and consistency of the effectiveness of chemical additives, specific surface areas should be controlled.
When any of these properties are to be controlled, more than one ingredient powder must be available to be mixed during batching operations. One engineer asked how to control SSA in his process which required one maxibag of powder per batch. Under such conditions (one maxibag = one batch), SSA can't be controlled. It is what it is in each maxibag. But when at least two sources of each powder are used in a batch, SSA can be controlled. The requirement is that one powder's SSA is higher than desired, and the other's SSA is lower than desired. When this is the case, the two powders can be mixed to achieve SSAs between their individual values, and the SSA of the daily batch can be controlled.
A reminder: When PSD is changed by milling or crushing, SSA changes. When PSD is changed by deagglomeration, however, SSA values do not usually change. PSD and SSA are related. Attempts to modify either one usually also affects the other. It is necessary, therefore, that both PSD and SSA be routinely characterized, modified, and controlled for complete control of batch properties.
Methylene Blue Index (MBI)
In traditional ceramic systems which contain ball clays and kaolins, PSD, SSA, and MBI should be routinely measured and controlled. Although some suggest that MBI is a measure of total surface area, it is not. MBI is a measure of plasticity-producing surface areas. Clays and kaolins have measurable MBIs. Non-plastic powders such as alumina and silica do not. MBI controls the nature of the plastic properties of traditional ceramic forming bodies. The most plastic clays have the largest ratios between MBI and SSA. Regardless of the units that are routinely used for MBI and SSA, when units remain constant, the ratio of MBI/SSA is indicative of plastic properties. Some very high SSA clays also have relatively high MBI values, but their MBI/SSA ratios are not particularly high and their contributions to body plasticities are not particularly strong. Some relatively low SSA clays have relatively low MBI values, but their MBI/SSA ratios are fairly high and they are strong contributors to the plastic properties of ceramic forming bodies in which they are used. The MBI/SSA ratio is important, but each property (MBI and SSA) must be measured independently to obtain, monitor, and control the ratio.
When ceramic bodies do not contain any clay minerals to control plastic properties, plastic properties are then adjusted and controlled using polymeric additives. In such cases with non-plastic powders, SSA measurements are critically important.
Soluble Ions
Soluble ions can reside naturally in mineral deposits, and they can be present as the result of additives used during processing of the minerals. What processes do Companies A, B, and C use in the beneficiation of their powders? Most likely, all three processes are different -- and therefore, all three powders will contain different soluble impurity ions.
If any of you think this article does not apply to you because you are using only chemically prepared materials, you are wrong. MBI may not apply, but PSD, SSA, and soluble ions certainly do apply. Chemically prepared materials are usually prepared by different companies using different processes, and that means each powder will contain different surface impurities, soluble ions, and/or organic chemicals.
Solids Contents
Most ceramic forming bodies and casting slips have tightly controlled solids contents. Solids contents are easy to measure and control, and most production facilities control this very carefully and precisely. This practice should continue.
A concern (which is outside the scope of this article) is "WHY?" has each solids content been chosen? In many cases, tradition is the reason rather than a fundamental processing consideration.
Regarding solids content controls -- keep up the good work!
Chemical Additives
This is the category that is most often used to control and adjust ALL problems after batches have been formed. Viscous and rheological problems caused by improper control of particle physics (i.e., caused by variations in PSD and SSA) are routinely fixed by additive chemical adjustments. Viscous and rheological problems caused by chemical imbalances or by improper chemical controls (i.e., caused by variations in MBI, soluble ion contents, additive chemicals, and solids contents) are also routinely fixed by additive chemical adjustments.
As mentioned, additive chemicals function in the interparticle fluid environment and on powder surfaces. Concentrations in the interparticle fluid are functions of solids contents, exposed surface areas, and the packing capabilities of the powders. As SSA and PSD vary, surface coverage by adsorbed additives, concentrations of the additives in the interparticle fluid, and distances between particle surfaces will vary. As SSA, PSD, and additive effectivenesses vary, viscous and rheological properties will vary -- and these last two properties (viscous and rheological properties) directly control forming processes.
Viscous and Rheological Properties
All of the properties mentioned above should be well controlled to maintain consistency from day to day of viscous and rheological properties. If all six listed properties affect viscous and rheological properties, when there is a viscous or rheological problem, one must determine which one or more of the six properties is causing the problem. When all six properties are routinely monitored and tightly controlled, viscous, rheological, and forming problems should only occur due to minor, unmonitored property variations or due to minor variations of daily properties from target values.
For example, when viscosities are too high, PSDs can be changed to pack better, surface areas can be reduced, MBIs can be reduced, solids contents can be decreased, soluble alkali ion contents can be increased, and/or deflocculant additive concentrations can be increased. All such changes should lower viscosities. When all six properties are closely monitored and controlled, one should have an idea in advance (that is, prior to batching) that some batch properties may be out of spec, and one should have an idea which properties should or could be altered to modify viscous, rheological, and forming properties to return them to specified target values. This is the PPC approach.
Summary
In answer to the question, "What properties should be characterized on a daily basis?", the answer is: PSD, SSA, MBI, soluble ions, solids contents, additive chemistries, and viscous and rheological properties. Process engineers should know the first four of these prior to batching, and they should understand the behavior of batches as functions of solids contents and additive chemistries. They should then use these understandings to carefully monitor and control all properties.
In the past, many particle size analyzers could not see all particles in a powder. Colloids were outside the range of most analyzers. New particle size analyzers, however, can see large fractions of the colloidal particles, so the capabilities in this area of analysis are changing quickly. PSD analysis should be a routine characterization tool in all ceramic plants.
SSA analysis measures all external particle surfaces -- even of the finest particles that are outside the range of particle size analyzers. SSA analysis should be a routine characterization tool in all ceramic plants.
The MBI test measures plasticity-producing surface areas which exist primarily in clay and kaolin minerals. This test should be routinely used in all ceramic plants using traditional clay-based ceramic bodies.
Soluble ions that travel with raw materials should also be closely monitored in all ceramic processes. Powders can be slurried, filter pressed, and the filtrate fluids can be tested for the presence of soluble ions. Some such ions are desirable. Some such ions are impurity ions. Many soluble ions are not a problem, but their concentrations should nevertheless be known. The presence and concentrations of all such ions should be measured and known, prior to batching. Some ions, when present, may need to be removed, or otherwise dealt with, prior to batching operations. The basis for all such process adjustments, however, is good data on all ingredient materials.
Solids contents can and should be routinely tested prior to and subsequent to batching operations. These tests are easy to perform, and adjustments are easy to make, as well.
The effectiveness of chemical additives should be tested on each raw material as well as on whole body compositions. Different additives and concentrations of additives will have different effects on processing properties. Behavioral results of the application of each different possible additive should be known and understood in each prospective body application for each ingredient material. Some additives are incompatible. Tests should be performed to determine which additives are most effective, and which mixtures of additives cause problems. Of course, this parameter represents a huge number of variations and a correspondingly difficult subject to be understood. It is hardly possible to understate the importance of this parameter, and it is hardly possible to spend too much time learning how additives behave in forming bodies.
Finally, the combination of all of these characteristics of ingredient powders produces the forming behaviors and viscous and rheological properties of forming bodies. Not only should rheologies be routinely measured, but each process engineer should understand the cause-and-effect relationships between the first six properties (the causes) and the viscous, rheological, and forming properties (the effects.)
Obviously, there are other important properties (each with their own characterization methods) that are applicable to ceramic bodies. Each process engineer should evaluate his/her specific process and process requirements to identify all properties that are uniquely important to their process. Those properties should then also be added to the Causes list, and (depending on their level of imporance) they may need to be monitored daily as well.
Miscellany
Please continue to send your ideas or questions for future topics. 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.