TGA Analysis graph with overlaid mass spectrometry curve

This is an example of serious crazing in a glaze. The lines have gotten darker with use of the bowl! That means the color is organic, from food. This cannot be healthy.

Fired to cone 10 oxidation. Although feldspar is a key melter in high and medium temperature glazes, by itself it does not melt as much as one might expect.

These are prepared for the SHAB test procedure. From them we can measure drying shrinkage, fired shrinkage and porosity over a range of temperatures.

The center portion was covered and so it lagged behind during drying, setting up stresses that caused the disk to crack. This test is such that most pottery clays will exhibit a crack. The severity of the crack becomes a way to compare drying performances. Notice the test also shows soluble salts concentrating around the outer perimeter, they migrated there from the center section because it was not exposed to the air.

Particle size distribution and root-of-two stack of sieves from 48-325 mesh

These are part of the procedure for the SHAB test. The length of the bars is entered into a recipe record in your account at When Insight-live has these numbers it can calculate the drying and fired shrinkages.

Bottom: cone 2, next up: cone 02, next up: cone 04. You can see varying levels of maturity (or vitrification). It is common for terra cotta clays to fire like this, from a light red at cone 06 and then darkening progressively as the temperature rises. Typical materials develop deep red color around cone 02 and then turn brown and begin to expand as the temperature continues to rise past that (the bottom bar appears stable but it has expanded alot, this is a precursor to looming rapid melting). The top disk is a cone 10R clay. It shares an attribute with the cone 02 terra cotta. Its variegated brown and red coloration actually depends on it not being mature, having a 4-5% porosity. If it were fired higher it would turn solid chocolate brown like the over-fired terra cotta at the bottom.

These recipes have the same chemistry but the 1215U uses frit to source the MgO and CaO. This demonstrates that it is not just chemistry that determines melt flow. Raw materials are crystalline and have different melting patterns than frits (which have already been melted and reground).

Drying disks used for the DFAC test are 12cm in diameter and 5mm thick (wet). A crack pattern develops in almost all common pottery clays as they shrink during drying. This happens because the center portion is covered and stays soft while the perimeter dries hard. This sets up a tug-of-war with the later-drying inner section pulling at the outer rigid perimeter and forcing a crack (starting from the center). If the clay has high plasticity and dry strength it can pull so hard from the center that cracks appear at the outer dried edge to relieve the tension. Or, it can create cracks that run parallel to the outer edge but at the boundary between the inner and outer sections. The nature, number and width of the cracks are interpreted to produce a drying factor that can be recorded.

By preparing these three tests you can measure many properties of a clay body. These include drying shrinkage, fired shrinkage, porosity, drying performance, soluble salts content, water content and LOI.

Data for hundreds fired clay test bars was logged into a portable Epson Programmable HX-20 computer and uploaded to a Radio Shack TRS-80 Model III where is was stored first on cassette, then floppy disk, then a loop tape. That data was later migrated to the Digitalfire DOS 4Sight lab record keeping system (as SHAB specimens) where it lived for more than 27 years (expanding to more than 200,000 tests) until being imported to an account in 2014.

This type of stamp is deal for stamping mix and ID information on SHAB (and many other test types) clay test bars. Set up the run or recipe number on the left and the specimen number on the right.

Fired strength tester

Dialometric chart produced by a dilatometer. The curve represents the increase in thermal expansion that occurs as a glass is heated. Changes in the direction of the curve are interpreted as the transformation (or transition) temperature, set point and softening point (often quoted on frit data sheets). When the thermal expansion of a material is quoted as one number (on a data sheet), it is derived from this chart. Since the chart is almost never a straight line one can appreciate that the number is only an approximation of the thermal expansion profile of the material.

Albany Slip was a pure mined material, Alberta Slip is a recipe of mined materials and refined minerals designed to have the same chemistry, firing behavior and raw physical appearance.

Preparing a representative powder sample for sieve analysis

The coarsest screen is at the top, the finest on the bottom. The opening for each is shown on the label. They are chosen such that each successive screen going down has an opening that is about half the area of the one above it. Using this series you can produce a practical measurement of the distribution of particle sizes in ceramic materials and bodies used in traditional ceramics (structural products industries, like brick, measure coarser particles than this, starting at perhaps 10 mesh and ending at 70). The 325 screen on the bottom is only used sometimes, it is difficult to finer-that-325 particles to pass through it because it blinds. It is not possible to shake powder through sieves that are this fine, samples must be washed through.

These are invaluable for creating a consistent environment in which to perform drying and shrinkage tests.

These disks concentrate the solubles on the outer edge (because of the way they are dried). Soluble salts can enhance the visual appeal of a fired clay but they can also do the opposite.

This is unlike some raw materials which melt suddenly.

Example of sedimentation test to compare soluble salts water extracts from suspended clay. This simple test also reveals ultimate particle size distribution differences in clays that a sieve analysis cannot do.

Example of serious glaze shivering using G1215U low expansion glaze on a high silica body at cone 6. Be careful to do a thermal stress test before using a transparent glaze on functional ware.

Particles from each category in a particle size distribution test of Skagit Fireclay

The Redart clay bars (left) are fired at cones 06, 04, 2, 4 & 5 (top to bottom). The Plainsman Blue Grey Plastic bars (right) are fired at 06, 04, 03, 02, 2 & 4. The SHAB test procedure (used to make these) gives us the firing shrinkage and porosity at each temperature, these are direct indicators of the fired maturity. Notice how much the fired color changes with increasing temperature. The fired maturity is pretty similar but the BGP is a little browner in color. It is also much more plastic (the drying shrinkage quite a bit higher).

The concentrations are not serious and are typical of what you might find on a commercial body.

Low temperature clays are far more likely to have this issue. And if present, it is more likely to be unsightly. The salt-free specimens have 0.35% added barium carbonate.

This is a ball clay. They are known to produce this type of soluble salts when fired at high temperature reduction (the inner salt-free section is such because that part of the tile was covered during drying, so the soluble salts from there had to migrate to the outer exposed edge). If soluble salts fire to a glassy surface they can affect the overlying glaze. But in this case they are not and have a minimal effect.

The stated particle size of a material and fired appearance can both be misleading. For example, these are Volclay 325 bentonite particles fired to cone 8 oxidation. They are from a water washed sieve analysis test, the oversize particles from a 325 mesh screen (left) make up 2% of the total and 1% are from the 200 mesh screen (right). Although the 325 particles appear ominously dark, individually they are likely to small to produce apparent fired specks in a porcelain. However 200 mesh sizes can produce visible fired specks, but that fraction of oversize does not have nearly as high iron or flux content. Still, the finer darker particles could agglomerate, it might be better to use a cleaner bentonite to plasticize a porcelain.

These bars have been fired at cones 4, 2, 02, 04 (top to bottom) using the SHAB testing procedure. We can measure fired shrinkage and porosity in each to get an indication of their fired maturity. The Redart (left) is much more vitreous and reaches almost zero porosity by cone 4 whereas the Lizella still has 11% porosity at cone 4. Lizella also has a much higher drying shrinkage because it is way more plastic. Two red clays could not be much more different than these yet sometimes they are substituted for each other in recipes!

This shows the soluble salts in the material and the characteristic cracking pattern of a low plasticity clay. Notice the edges have peeled badly during cutting, this is characteristic of very low plasticity.

This clay is used by traditional potters in the Mazatlan, Sinaloa, Mexico area. This DFAC test shows a very wide main crack and number of edge cracks. These indicate very high shrinkage and plasticity. Although the clay has some coarser grains that help channel water out, this is a very poor showing for this test, no large scale manufacturer could tolerate this. Yet they use it with success, having learned how to adapt. Note also that soluble salts are fairly low.

This 1000 ml 24 hour sedimentation test compares Plainsman A2 ball clay ground to 10 mesh (left) with that same material ball milled for an hour (right). The 10 mesh designation is a little misleading, those are agglomerates. When it is put into water many of those particles break down releasing the ultimates and it does suspend fairly well. But after 24 hours, not only has it settled completely from the upper section but there is a heavy sediment on the bottom. But with the milled material it has only settled slightly and there is no sediment on the bottom. Clearly, using an industrial attrition ball mill this material could be made completely colloidal.

This test shows the incredible dry shrinkage that a ball clay can have. Obviously if too much of this is employed in a body recipe one can expect it to put stress on the body during drying. Nevertheless, the dry strength of this material far exceeds that of a kaolin and when used judiciously it can really improve the working properties of a body giving the added benefit of extra dry strength.

Large particle kaolin (left) and small-particle ball clay (right) DFAC drying disks demonstrate the dramatic difference in drying shrinkage and performance between these two extremes (these disks are dried with the center portion covered to set up a water content differential to add stresses that cause cracking). These materials both feel super-smooth, in fact, the white one feels smoother. But the ultimate particles tell the opposite story. The ball clay particles (grey clay) are far smaller (ten times or more). The particles of the kaolin (white) are flatter and lay down as such, that is why it feels smoother.

These are DFAC drying performance tests of Plainsman A2 ball clay at 10 mesh (left) and ball milled (right). This test dries a flat disk that has the center section covered to delay its progress in comparison to the outer section (thus setting up stresses). Finer particle sizes greatly increase shrinkage and this increases the number of cracks and the cracking pattern of this specimen. Notice it has also increased the amount of soluble salts that have concentrated between the two zones, more is dissolving because of the increased particle surface area.

This disk has dried under heat (with the center part protected) for many hours. During that process it curled upward badly (flattening back out later). It is very reluctant to give up its water in the central protected section. Obviously it shrinks alot during drying and forms a network of cracks. When there are this many cracks it is difficult to characterize it, so a picture is best.

This DFAC drying performance test compares a typical white stoneware body (left) and the same body with 10% added 50-80 mesh molochite grog. The character of the crack changes somewhat, but otherwise there appears to be no improvement. While the grog addition reduces drying shrinkage by 0.5-0.75% it also cuts dry strength (as a result, the crack is jagged, not a clean line). The grog vents water to the surface better, notice the soluble salts do not concentrate as much. Another issue is the jagged edges of the disk, it is more difficult to cut a clean line in the plastic clay.

These are the oversize particles (from the 79, 100, 140 and 200 mesh sieves) from 100 grams of a commercial Gleason ball clay. They have been fired to cone 8 oxidation. There is 1.5 grams total, this is within the limits stated on their data sheet even though the material is sold as 200 mesh grade. Firing the samples shows whether the particles contain iron that will produce specking in porcelains and whiteware. In this case there are a few. We do this test on many materials and this is typical of what we see.

These are the oversize particles (from the 70, 100, 140 and 200 mesh sieves) from 100 grams of a commercial ball clay. They have been fired to cone 10 reduction. As you can see, this material is a potential cause of specking, especially in porcelain bodies. It is not only wise to check for oversize particles in clays, but firing these particles will tell you if they contain iron. A 200 mesh screen would be a good start for this test, it would catch all of these.

SHAB (Shrinkage, Absorption) test bars ready to unload. These are measured for length after drying and firing and for weight after firing and boiling. This data is plugged into my account at and it calculates shrinkage and porosity numbers. If you fire bars of a clay to a range of temperatures you can characterize key properties of a clay very effectively.

Dolomite is a key material for glazes, especially mattes. When you are forced to adopt a new brand it needs to be tested. Here, three tests were done to compare the old long-time-use material (IMASCO Sirdar) with a new one (LHoist Dolowhite). The first flow test is a very high dolomite cone 6 recipe formulated for this purpose; the new material runs a little more. The second is G2934 cone 6 MgO matte with 5% black stain; the new material runs a little less here. The third test is the high dolomite glaze on a dark burning clay to see the translucency and compare the surface character. They are very close. It looks like it is going to be OK. Does your supplier test new materials when they are forced to switch suppliers?

A porcelain mug has pulled slightly oval because of the weight of the handle. This happens in highly vitrified porcelains (e.g. translucent ones). The amount of feldspar or frit in the body determines the degree of maturity, the correct percentage is a balance between enough to get the maximum translucency and hardness but not so much that ware is deforming excessively during firing. This is Plainsman Polar Ice at cone 6, this degree of warp is acceptable and can be compensated for.

Example of the oversize particles from a 100 gram wet sieve analysis test of a powdered sample of a porcelain body made from North American refined materials. Although these materials are sold as 200 mesh, that designation does not mean that there are no particles coarser than 200 mesh. Here there are significant numbers of particles on the 100 and even 70 mesh screens. These contain some darker particles that could produce fired specks (if they are iron and not lignite); that goodness in this case they do not. Oversize particle is a fact of life in bodies made from refined materials and used by potters and hobbyists. Industrial manufacturers (e.g. tile, tableware, sanitaryware) commonly process the materials further, slurrying them and screening or ball milling; this is done to guarantee defect-free glazed surfaces.

Overfired Polar Ice porcelain. This bowl fired with an oval-shaped rim and was sticking to the shelf.

A 10cm mark is made in the plastic bar during preparation. Talc powder is put on the marker so it pulls cleanly away from the clay leaving a crisp mark. These bars are made from the same clay and will be fired at multiple temperatures. Each one needs to be measured.

A batch of fired test bars, organized by temperature, have already been weighed (the weight is written on the side of each bar). Now they will be measured and the SHAB test data (shrinkage/absorption) entered into each recipe record (in an account at From this data Insight-live can calculate fired shrinkage and fired porosity, enabling you to compare the degree of vitrification of different materials and bodies. This is especially good for quality control purposes.

These bowls are fired at cone 03. They are made from 80 Redart, 20 Ball clay. The glazes are (left to right) G1916J (Frit 3195 85, EPK 15), G191Q (Frit 3195 65, Frit 3110 20, EPK 15) and G1916T (Frit 3195 65, Frit 3249 20, EPK 15). The latter is the most transparent and brilliant, even though that frit has high MgO. The center one has a higher expansion (because of the Frit 3110) and the right one a lower expansion (because of the Frit 3249). Yet all of them survived a 300F to icewater test without crazing. This is a testament to the utility of Redart at low temperatures. A white body done at the same time crazed the left two.

An example of a cone 10 porcelain that is over mature. It contains too much feldspar and is vitrifying so much that it is beginning to melt. The weight of the handle is pulling the lip into a oval shape, even though the hourglass shape of the piece should offer stability.

This is a cone 04 clay (Plainsman Buffstone) with a transparent glaze (G1916Q which is 65% Frit 3195, 20% Frit 3110, 15% EPK). On coming out of the kiln, the glaze looked fine, crystal clear, no crazing. However when heated to 300F and then immersed into ice water this happens! At lower temperatures, where bodies are porous, water immediately penetrates the cracks and begins to waterlog the body below. Fixing the problem was easy: Substitute the low expansion Frit 3249 for the Frit 3110.

SHAB test bars, an LDW water content sample and a DFAC drying disk about to be put into a drier. The SHAB (shrinkage-absorption) bars shrink during drying and firing, the length is measured at each stage. The LDW sample is weighed wet, dry and fired. The can prevents the inner portion of the DFAC disk from drying and this sets up stresses that cause it to crack. The nature of the cracking pattern and its magnitude are recorded as a Drying Factor. The numbers from all of these measurements are recorded in my account at Insight-live. It can present a complete physical properties report that calculates things like drying shrinkage, firing shrinkage, water content and LOI from these measured values.

The plastic clay has been rolled to 3/16 inch thickness (using the metal rods and a rolling pin). The disk is cut to 12 cm diameter.

The heat lamp dries the out edge in minutes (this photo makes it appear hotter than it really is). The center section of the disk is protected by the glazed bowl and takes an hour or more to dry. This sets up stresses that cause the disk to crack. The nature and size of the cracks enable establishing a drying factor value for the clay.

The specific gravity of a glaze slurry is simply its weight compared to water. Different glazes optimize to different specific gravities, but 1.4 to 1.5 is typical (highly fritted glaze are higher). To measure, counter-weigh a plastic measuring cup on your scale and fill it with 500 grams of water and note how high the water fills it (hopefully to the 500cc mark!). Fill the container with your glaze to the same place. Divide its weight by the number of ccs (in this case, 500) and you have the specific gravity. The more you weigh the more accurate is the test.

To measure particle size in a slurry or powder you need sieves. This is the most popular type used in labs. They are made from brass by a company named Tyler. The range screen sizes for testing particle size is very wide. The top screen has an opening of 56 mm (that size and smaller pieces can fall through). The bottom sieve has an opening of 0.1 mm, the wires are almost too small to see. Coarser and finer sieves are available. You can buy these on ebay for a lot less than new ones, just search for tyler sieves. Keep in mind that the finer sieves (especially 325) are fragile and easily ripped. We use a series that bottoms out at 200.

These bars were fired at cone 10, they were straight when dry. The back one is a cone 10 Grolleg body, the front one is a cone 6 Grolleg body. This simple test is valuable to determine susceptibility to warping in porcelains. If the pyro-plastic deformation is too much, for example, the weight of a handle will pull the round rim of a mug into an oval shape, for example.

When porcelains mature in the kiln they progress toward vitrification, getting softer. This simple test enables anyone to quantify the degree to which a porcelain is likely to warp. Bars of plastic clay almost never dry straight, so the measurement (in mm) to which they deviate from straight is recorded and the bar is mounted with the hump upwards. After firing the mm of firing deviation-from-straight are added to the dry value to derive a total pyro-plastic deformation measurement. This can be recorded as an absolute value for comparison with other clays or temperatures.

This is G2926B cone 6 transparent glaze. I am developing a simple test procedure to produce an absolute measurable value for glaze melt flow and it appeared it would be worthwhile to create a mold to make these cone-shaped samples. But I was wrong. Both specimens are exactly 10 grams, but the simple ball flows better. This is likely because of better early heat penetration because there is only a small area of contact with the tile.

This is Plainsman Polar Ice porcelain. It is plastic enough to be pulled very thin on a plaster table. I have cut it in a 2in by 2in grid. This porcelain is ideal for these testers because it has such a glassy surface, this produces minimum surface tension to resist the flow of the glaze.

The weight data from these fired test bars is being collected for the SHAB test in Insight-live (they have just been boiled for five hours and soaked for 19). Compiling this type of data for hundreds of simultaneous tests is possible because Insight-live takes care of all its organization.

This terra cotta cup (center) is glazed with G2931G clear glaze (Ulexite based) and fired at cone 03. It survives 30 seconds under direct flame against the sidewall and turns red-hot before a fracture occurs (the unglazed one also survived 30 seconds, it only cracked, it did not fracture). The porcelain mug (Plainsman M370) is glazed with G2926B clear, it survived 15 seconds (even though it is much thinner). The porcelain is much more dense and durable, but the porous nature of the earthenware clearly withstands thermal shock much better. It is actually surprisingly durable.

I finally gave up trying to dry the inner section of this DFAC test. During that test the inner part of the disk is shielded from the air flow or heat lamp. This sets up a shrinkage gradient that encourages cracking of the sample. But with some clays drying can be so slow that it can take a days. Serious cracking and high drying shrinkage almost always accompanies this phenomenon.

The vast majority of glazes are plastic (but less than clay bodies). They can be dewatered on a plaster surface and formed. Why do this? To make 9-10 gram balls and fire them on flat tiles (or inclined flow testers) to see their melting characteristics. It is surprising how much this can tell you about the glaze. To do it, mix the slurry well and pour a little on the plaster. It should dewater in less than 30 seconds. As soon as the water sheen is gone, scrape it up with a rubber rib, hand-knead it and flatten it back down to dry a little more if needed (leave it only for five or ten seconds and rework it. Repeat until it is stiff enough to roll into balls of about 12 grams. Stamp them with ID numbers and dry them.

This body is used in the sanitary ware industry in China, the supplier sends this report with each shipment. The chemistry and assorted values for porosity, shrinkage, particle size are provided. The factory receiving this report accepts it as gospel and goes into production. However engineers at the plant need to think twice about such reports. These tests are being done at one temperature, they say nothing about what that body is doing above and below that temperature. Is it being employed in a volatile range of the porosity or firing shrinkage curves? Zero porosity bodies of this type are best when fired to a point near where the porosity curve descends to reach the x-axis. However that curve remains at zero while the shrinkage one tops out and reverses direction. At some point the porosity curve sharply rises. Only by firing and testing at a range of temperatures in your own lab can you where your body is on the curve.

The porcelain on the left is highly vitreous, it cracked in 12 seconds, but the other lasted until 25! Why? Only theories, both of these should be failing sooner. The vitreous nature of the one on the left could be giving it the sheer strength to endure longer. The one on the right is Crystal Ice (from Plainsman Clays). It has very high quartz, which should have accelerated the failure. However it is not completely vitreous, the voids within the matrix could be giving it resilience. Another factor could be that the glaze is under more compression, adding strength. This whole thing is counter intuitive, porous bodies like earthenware normally endure best and vitreous porcelains the worst. Yet this same test on my earthenwares cracks them in less than ten seconds?

A recommended flameware recipe from a respected website (equal parts of 35 mesh grog, talc and ball clay). Looks good on paper but mix it up for a surprise. The texture is ridiculously coarse. Recipes like this often employ fire clays and ball clays, but these have high quartz contents (in a test like this a ball clay vessel could easily fail in 5 seconds). But this one is surviving still at the 90-second mark. Or is it? While porcelain pieces fail with a spectacular pop of flying shards, these open-porous bodies fail quietly (note the crack coming up to the rim from the flame). There was an intention to create cordierite crystals (the reason for the talc), it is hard to say whether than happened or not. But the porosity of 12.5% would be difficult to deal with. On the positive side, you could likely continue using this vessel despite the crack.

It seems impossible but that is what happens with this one at cone 03. This is a native material that was found on the banks of the South Saskatchewan river near Hayes, Alberta (and brought to me for testing). Even when fired to maturity (around cone 2) it still has 10% porosity! This specific sample has even been ball milled for hours and it still does not shrink. And it still feels sandy on the potters wheel. It also has incredible dry strength, the highest I have ever seen. Yet its drying shrinkage is still less than 7% (that of a typical plastic pottery clay). Plus it has very high plasticity. This behavior defies logic, I have found a good explanation.

These two glazes have the same chemistry but different recipes. The F gets its boron from Ulexite, and Ulexite has a high LOI (it generates gases during firing, notice that these gases have affected the downward flow during melting). The frit-based version on the right flows cleanly and contains almost no bubbles. At high and medium temperatures potters seldom have bubble issues with glazes. This is not because they do not occur, it is because the appearance of typical glaze types are not affected by bubbles (and infact are often enhanced by them). But at low temperatures potters usually want to achieve good clarity in transparents and brilliance in a colors, so they find themselves in the same territory as the ceramic industry. An important way to do this is by using more frits (and the right firing schedules).

A 3 inch by 3/16 thick tile is flamed and timed until it cracks. The crack usually is audible and sometimes spectacular (the tile flying apart). However if the body is heavily grogged it can grow slowly (often from the edge inward).

Feedback, Suggestions

Your email address


Your Name


Copyright 2003, 2008, 2015, All Rights Reserved