Volume 1 Number 4 Dennis R. Dinger 1 February 2003
An Update
Readership of this E-zine continues to grow. Continue to feel free to forward this or any issue to friends and associates. If this is the first issue you've seen and you want to add your name to the mailing list, click HERE. All back issues are stored on the web site and they can be accessed from the Publications page. Questions, suggestions, and/or requests for topics to be covered in future issues of this e-zine can be sent to QuestionsandComments@DingerCeramics.com .
My new book, Rheology for Ceramists, is available on the Books and Downloads page in PDF format as a downloadable self-extracting zip file. A free downloadable preview of the Rheology book, which contains the Table of Contents, Preface, and Chapter 3, is also available on the website. The hot link to a free copy of Adobe® Acrobat® Reader®, which is required to access the PDF file, is at the bottom of the Books and Downloads page.
For those of you who are curious about the book and my writing style (my thought process???), in addition to discussions about the details of the rheological properties of ceramic suspensions, there is also a section in the particle/fluid interactions chapter concerning bugs colliding with the windshields of cars, a section which explains dilatant blockages using the example of a fire drill in a university classroom, and discussions of the rheological behaviors of ketchup, mayonnaise, whipped cream, and corn starch. The goal, of course, is to cause you to think about rheological properties every time you're eating hamburgers, dessert, and/or gravy, when you're sitting in a classroom, and when you're driving your car. Next time you're driving down the road and your wife or husband asks, "What are you thinking?", you can gesture towards the windshield and respond, "I was just thinking about the rheological properties of the insect population interacting with the aerodynamic slip stream around this car." If you can say it with a straight face, your wife or husband should be impressed.
The topics in this issue are subjects that I think are timely, based on recent discussions I've had. Some discussions were about tap densities, so that triggered this first subject. Some discussions and questions were about firing behaviors and the use of fluxes, so that triggered the second subject. The interrelationships between particle size distributions, mixing, and fluxes is a really broad topic, so I narrowed the scope of the second article considerably.
How to Test the Packing Capacity of Suspended Powders
If you were handed a powder sample and asked to test its capacity to pack, how would you do it? An obvious test is a tap density test. That method specifically applies to dry processing. Are you dry pressing? Are you making and using suspensions in your process? Is there an easy way to test the packing capacity of suspended powders? Indirectly, yes. If you are using suspensions in your process, or even if you're doing dry processing, measuring the viscosities of suspended powder samples can be used to gauge the packing potential of powders. These two methods (tap densities and suspension viscosities) will be discussed in this article.
Do You Need Dense Packing?
Experience has shown that powders that can pack really well (let's call these good distributions) provide excellent rheological properties. Powders that don't pack well (let's call them bad distributions) cause rheological problems. Generally speaking, good distributions allow a wide range of solids contents and a wide variety of rheological properties to be controlled and used. Generally speaking, bad distributions require high moisture contents, low solids contents, and high levels of deflocculation. Under highly deflocculated conditions, suspensions tend to be dilatant during processing.
Good distributions usually contain powders distributed evenly across broad ranges of particle sizes. Bad distributions tend to contain particles that are all within very narrow ranges of particle sizes (e.g., monodispersions). We won't go beyond these two statements at this time, because the subject of particle size distributions (PSD) and packing is huge.
Many engineers aren't interested in dense packing. It's simply not an important concern in their processes. Even when dense packing is of no primary concern, good PSDs should be of interest to anyone processing powders and suspensions because they can provide excellent rheological properties and lots of processing flexibilities. Good, broad PSDs can give you such flexibility. Bad, narrow PSDs can't. Good distributions are desirable, regardless what solids contents are used in ceramic process systems. Because good distributions are desirable, it's handy to have a means to test the packing capabilities of powders and to learn whether the distributions are good or bad.
Homogeneity of Mixing
The first concern for any density test measurement should be to achieve and maintain homogeneity of distribution throughout the test sample. Even when loose dry powders are well-mixed (if it's possible to ever achieve a well-mixed dry powder), the powders can unmix by size as soon as the mixture is subject to any kind of vibration or motion. The simple answer to the question, "How do you keep dry powders mixed after removing them from the mixer?", is, "You don't!"
Put some Folger's coffee crystals and dry powdered coffee creamer into a cup. Stir them up with a spoon to achieve a 'homogeneous' mixture. When you're satisfied that the coffee and creamer are mixed well, remove the spoon, and tap the cup on the table several times. The 'mixture' will separate quickly again into coffee and creamer. Unless dry particles have some reason to cling to each other, which many don't, they will start to unmix as soon as they're removed from the mixer. This is a good experiment because the coffee is dark brown and the creamer is white, so you can see which particles are which, and you can see when the two are mixed and when they're not. But what are the colors of most ceramic powders? White, white, white, off-white, cream, tan, tan, and tan? It is not possible to tell whether most ceramic bodies are mixed or unmixed simply by looking at them.
When suspensions are used in a process, it's likely that the level of mixedness achieved during mixing (unlike dry powders) will remain after mixing is complete and suspensions are removed from the mixing tank. Only at really low solids contents do settling, and unmixing caused by settling, become problems. The question still remains, whether or not suspensions are really well mixed. HID systems (high intensity dispersion) can achieve fairly high levels of homogeneity. LID (low intensity dispersion) usually cannot.
In dry pressing operations, many powders are suspended first, mixed, and the body suspensions are then spray dried. For the most part, spray dried granules lock the mixedness of the feed suspensions into the dry granules. This is good.
Lab/Plant Procedure Similarity
Mixing, unmixing, and the actual state of mixedness achieved should always be considered before any kind of density test. Another consideration should be to insure that lab test methods match production methods. If they don't, results will be meaningless.
One engineer rather excitedly told Professor Jim Funk (from Alfred U), "We always use a HID kitchen blender to prepare all of our lab samples!" When Jim asked him, "Can you show me the HID kitchen blender you use to prepare all of your production batches?", the fellow rolled his eyes, slapped his forehead, and let out an "Ohhhhhhhhhhhh!!!!" (I wasn't there to see this first hand, but I heard Jim tell it at least 43 times over the years.) The point is this: If you're going to use a special process in the lab, make sure you're also using it in the plant, or it will change the samples and the results won't correlate with process results.
Tap Densities
Tap densities are routinely used to measure packing potentials of dry powders.
Loose Powders
Because tap densities are performed using vibration to densify dry powders, the results are affected by powder surface properties and friction. In tap density tests, powder particles are resting on other powder particles. Densification occurs under vibration because the particles slide over each other until they all reach their lowest, densest packing positions in the test container. When powder surfaces are rough, the finer powders (least mass, highest SSA) will have the most difficulty sliding against one another. When powders are agglomerated, they may also have trouble sliding over one another due to their weird shapes. When powders can't slide easily, they won't easily sift down through the pack to achieve their densest packing positions.
Even when particles are smooth and friction is low, it's still relatively easy for a particle to block an entrance hole to a pore, which will stop further densification. If you have three softballs in a triangular array and you drop a golfball into the top center site formed by the three, it will sit there on top and not fall through the entrance passage between the softballs into the 'pore' below the softballs. Finer particles above the golf ball, even those that could fit through the entrance passage to the volume below the softballs, can be blocked by the golf ball. When this happens, densification ceases.
In a tap density test performed several years ago on fine alumina powders, the results correlated with the specific surface areas (SSA) of the powders, not with the expected packing potentials based on the particle size distributions of the powders. The finer the powders and the higher the SSAs, the less dense were the tap densities. The results supported the argument that particle friction prevents tap density tests from achieving maximum densification.
Granules
Spray dried granules tend to be relatively narrow size distributions of spheres, dimpled spheres, and donut shapes. Tap density tests of granules pack only the granules, not the powders within the granules. But the results of such tap density tests combine the two types of packing. If tap density results for granules vary considerably, one still has to separate (1) the packing of the granules from (2) the packing of the powders within the granules. If the granules are spheres, they will pack similar to other spheres (and that will be poorly). But even with similarity of packing of the granules from test to test, the mass of a 100cc sample of packed granules can vary considerably because the density of particle packing within the granules varied considerably from test to test.
Tap density results of the loose powders that make up the granules probably won't correlate with the granule densities either, because the tap density test is a dry test, and granules are usually formed with wet processes. The results won't correlate because this is an apples to oranges comparison.
Summary Regarding Tap Densities
Tap densities are easy tests to run, but one has to be careful when interpreting the results. A bad distribution of smooth particles can pack better than a good distribution of rough particles. A bad distribution of coarse particles can pack better than a good distribution of fine particles. Tap density test results for granules don't indicate whether the granules, or the particles within the granules, are the good or bad actors in the samples. When using tap density tests, which may be the best or most applicable density test to use for a process, interpret the results carefully.
Viscosities
When using suspensions in a process, measuring the viscosities of identically prepared samples with a rotational viscometer is an excellent way to gauge the packing density of the powders. Regardless of the state of flocculation/deflocculation used in the process suspensions, this viscosity test requires test suspensions to be highly deflocculated. Other requirements are a fixed solids content, a fixed measuring container size, a single viscometer measuring spindle, and a fixed viscometer rpm. All of the equipment and test conditions must be identical from test to test. Then, as packing potentials of the powders increase, the measured viscosities will decrease. Adjust the solids contents and chemistries to the predetermined values, and measure the viscosities. The viscosities will then correlate with packing potentials. The sample with the lowest viscosity will exhibit the highest packing potential.
Alternatively, a target viscosity can be fixed, and the suspension's solids content can be the variable. In this case, one searches for the solids content for each suspension sample that produces the chosen target viscosity. Distributions with the highest packing potentials will produce the target viscosities at the highest solids contents.
These tests work because suspension fluids perform two tasks: they fill pores, and they separate particles. As the powders' packing potential increases, less fluid must fill pores, and more fluid separates particles. Generally speaking, as packing potential increases, the average distance between particles in the suspension will increase. The greater the distance between suspended particles, the lower will be the measured viscosity. (Or for the alternative form of the test, the average distance between particles that corresponds to the target viscosity will occur at higher solids contents as packing potential increases.)
The viscosity method for gauging packing potentials does not require particles to slide against one another as in tap density tests. If a golf ball wants to pass three softballs in suspension (as in the previous example), the softballs can simply separate a little due to the imposed shear and allow the golf ball to pass between them. None of the particles ever need to be in contact when they're in suspension.
This viscosity method for determining packing capabilities is not a function of particle surface properties, like the tap density test, because the particles never need to come in contact with one another.
Flexibility of Properties During Processing With Dense Packing Potentials
There is a big difference between the packing potential of a powder's particle size distribution and the actual packing that will occur during processing. If it's necessary to achieve dense packs during processing, then packing potentials of the powders must be high and suspensions must be highly deflocculated. Under such conditions, particles will easily collide with one another during flow, rheological properties will be dilatant, and viscosities may be high.
When it's not necessary to achieve dense packs during processing, packing potentials do not need to be high. If they are high, however, solids contents can be set high or low, as desired. If solids contents are low, suspensions can be flocculated, rheological properties will be shear-thinning, and forming should proceed well. If solids contents are higher, it still may be possible to flocculate the suspensions and gain the benefits of flocculated chemistries. When really high solids contents are desired, suspensions can then be deflocculated to achieve process viscosities.
The primary benefit of high packing potentials is the capability for dense packing. When dense packing is not necessary, the secondary benefit of high packing potentials is the wide range of solids contents, rheological properties, and process viscosities that can be achieved.
When suspensions are flocculated, gel structures will form. The gel structures and formed wares will be relatively porous, and forming properties should be good. The actual packing densities produced by flocculated bodies will bear no relationship to calculated or measured packing potentials. Flocculated suspensions will not exhibit the primary benefit of dense packing. Flocculation doesn't allow dense packing, so don't expect it. But flocculated suspensions will exhibit the secondary benefit, which is the excellent rheological and viscous properties that can be achieved over a relatively wide range of solids contents.
Conclusions
Just because calculated or measured packing potentials are high doesn't mean that a distribution will pack to high density during forming operations. Suspensions will only ever pack to approach their packing potentials if they are well deflocculated. Dry powders seldom pack as densely as desired because surface friction prevents particles from moving into their densest possible arrangements.
High packing potentials usually indicate that particle/particle collisions will be minimized during suspension processing, and therefore, one can expect rheological properties to be good. Low packing potentials usually indicate the opposite: particle/particle collisions will dominate and rheological properties will be bad (i.e., dilatant). Poor packing potentials almost always require suspension solids contents to be low.
It is quite useful to know the packing potential of particle size distributions. In some cases, the goal is to achieve dense packing. In other cases, the goal is to achieve excellent rheological properties. In those systems using dry processing methods, tap densities provide the most useful and meaningful results. In all systems (both wet and dry), however, packing potentials can be gauged by measuring suspension viscosities under conditions of fixed solids contents and fixed (deflocculated) additive chemistries. The biggest advantage to this method is that it does not force particle/particle contacts during the testing, which is the biggest disadvantage to the tap density method.
How Do the PSDs of Feldspars and Sintering Aids Affect Firing Behaviors?
This topic is small part of a really broad subject. Many of you already realize most of what is written in this article. If so, consider this a review. The quiz will be scheduled at a later date.
The specific question is this: What is the optimum particle size distribution (PSD) to use with these materials to optimize firing behavior? Feldspars are added to ceramic bodies to decrease firing temperatures and to start the reactions between body constituents. In pure systems, such as pure oxides or carbides, sintering aids perform this function.
The Finer The Better
Experience has shown that by reducing the particle sizes of feldspars, firing reactions can take place at lower temperatures. Finer is not always better if you're working with glasses or glazes. It's not that finer particles don't react at lower temperatures in glasses or glazes, but because the fine powders tend to cause tiny bubbles that are hard to remove. But with respect to ceramic bodies, 'the finer the better' generally applies and it can take two forms:
Same Percentage, Lower Firing Temperature
Tests have been performed on whiteware bodies using different finenesses of feldspars. Maintaining constant total feldspar contents in bodies, the use of fine feldspars has produced reductions in firing temperature of as much as 100oC.
Lower Percentage, Same Firing Temperature
Tests have also been performed on whiteware bodies using different concentrations of feldspars. When fine feldspars were used, firing behaviors were held approximately constant with less total feldspar in the body compositions.
In a pure alumina body, for example, a little colloidal alumina can produce similar effects. Colloidal powders can be expensive, but when used sparingly, they can improve firing behaviors. Colloidal versions of other materials should have the same effects on other pure systems.
This type of information is not usually included in phase diagrams. Phase diagrams can take years to create. They show equilibrium conditions, which usually don't apply to most production bodies. In this age of fast firing, low temperature firing, etc., fired bodies will seldom be close to equilibrium bodies. But within the firing times and temperatures allotted, the reactions will certainly move in the direction of equilibrium.
Finer particles (in the sub-sieve and colloidal ranges) react more quickly and at lower temperatures to move firing/sintering reactions along more quickly. As particles sizes are reduced, specific surface areas increase substantially. The larger surface areas of the such fines can be in much more intimate contact with the surfaces of the other body ingredients. Their enhanced activities and their broader distribution throughout bodies both help to improve firing behaviors.
Partly, The Title Question Is A Trick Question
Part of the answer, which is not even part of the title question, has to do with the mixedness (the homogeneity) of the body. If you add the 'optimum' size distribution of fluxes, or colloids, to a body to improve firing, you will not see the desired results if those materials are agglomerated and insufficiently dispersed throughout the body.
Agglomerated Fluxing Agents Can Sinter
The finest feldspars or colloids known to man, if not well-dispersed throughout the body, will not find other materials to react with, and they will sinter. Sintering will take place at higher temperatures than reactions with other ingredient materials would normally begin. In such systems, most of the firing benefits of the fine materials can be lost.
If feldspars, colloidal materials, or other firing aids are agglomerated and not well dispersed throughout the body, the bulk of their mass will be isolated from other body ingredients and the firing reactions will be delayed to higher temperatures. The particles at the surface of agglomerates will tend to react with other ingredients in their vicinities, but most of the benefits of using these materials will be lost or wasted. Wasted is probably the better word, because if a premium is being paid for these materials, their effects will not be realized.
Greater Percentages of Agglomerated Fluxing Agents Will Be Required
If agglomerated constituents aren't adequately dispersed, more will typically be required to do the same job. 'More' doesn't necessarily mean 'more than the original coarser materials', but it means more of the finer materials than should be required. One small agglomerate can contain thousands of colloidal particles. So if the percentage of fines added to a batch is sufficient when the colloids report as individuals, the percentage will be insufficient if the colloids report as agglomerates. Such agglomerates will not allow uniform distribution of the fines, and increased percentages may be required.
This is a similar phenomenon to the glaze pigment example, which has been mentioned before. Well-dispersed (with high intensity dispersion) pigments provide denser colorations than when the same materials are less well-dispersed (using low intensity dispersion). Our experience with this is that lesser quantities are required to give the same coloration (with less variability) when dispersion is adequate. One should expect similar results with feldspars and firing aids. When well-dispersed, lesser quantities should be required. When not well-dispersed, greater quantities may be required.
Some Materials Leach
Some alkali-containing materials that provide great firing benefits are partially soluble, and the alkalis leach into the carrier fluids over time. This can cause stability problems during storage. Nepheline syenite is one material that comes readily to mind. Additions of a little of this mineral can produce remarkable changes in firing behaviors of whiteware bodies. But it is a self-deflocculating mineral. The fine form of this mineral is a great dilatancy demonstration material. It's better than corn starch, because it is self-deflocculating. When added to a body suspension, the alkalis leach out with time to deflocculate the body. If the suspension aging process normally produces slow flocculation, this mineral can balance that behavior. Once the suspension approaches equilibrium, however, the deflocculation process can continue and cause viscosities to decrease.
When alkalis and other ions can leach out of minerals, extremely fine versions of those minerals contain large surface areas which aid the leaching process. Although it may not seem to be much of a problem to deal with this phenomenon, it can be very difficult in the plant to stop viscosity drift that results from ion leaching problems.
Summary
An excellent way to improve firing behaviors and/or to lower firing temperatures is to add finer versions of the feldspars, or to add colloidal materials in pure systems. Replacing all or part of a feldspar with a finer form of the same feldspar should reduce the firing time and/or firing temperature. The use of some finer feldspar should allow the same firing conditions to be achieved with less total percentage of feldspar in the body.
Colloidal forms of materials should behave similarly in body compositions containing one primary material. In pure alumina systems, for example, small additions of colloidal alumina should help the sintering behavior.
Two important points must be considered when using these techniques: (1) Fines or colloids used in this way MUST be well-dispersed throughout the body suspensions or their benefits will not be realized. (2) Fines or colloids that contain leachable ions can cause suspension stability problems in aging and holding tanks.
Miscellany
Don't forget to send suggestions for topics. I've been using discussions I've had with subscribers to help identify and select topics. So if you want my take on a subject, please send me a note.
In case you're interested, it's 90oF, sunny, and warm here in Indonesia.
Thanks. See you next time.
Thanks for signing up to receive Ceramic
Processing E-zine
To unsubscribe, click the "Remove Me" link below.
Copyright © 2003 Dennis R Dinger
103 Augusta Rd, Clemson, SC 29631 (864) 654-3155
All Rights Reserved.
All Rights Reserved.