Ceramic Recipes logo

Custer Feldspar

Feldspar is an important material for potters, as we use it in both clay body and glaze recipes. Because of this widespread use, we need to have an accurate analysis of the materials in order to get the results we want.

An Accurate Analysis

Custer feldspar can be described as a natural frit because it provides us with relatively insoluble sodium and potassium oxide. Sodium and potassium oxides are important because they start their fluxing actions at 1472°F (800°C)—well before the other fluxes we commonly use and, if present in the right amounts, control the production of cristobalite.

It is important that we have accurate analysis of the materials we use in order to deal with the technical requirements of our craft. It has become relatively easy to contact the mines where our materials are produced. That, and the availability of low-cost analysis of raw materials at many labs, gives us the tools to avoid the many problems of firing clay and glazes.

For instance, looking up Custer feldspar on the web leads to many published analyses and everyone seems to assume that the historic analysis for Custer spar is accurate for all samples of the material. This particular analysis was downloaded from the Pacer Minerals web page September 1, 2013. (Pacer Minerals is the company that mines Custer Feldspar near Custer, South Dakota.)

*The information and data contained herein are believed to be accurate, but the manufacturer makes no warranty with respect thereto and disclaims responsibility for reliance thereon. This data relates only to the specific material described herein, and does not relate to use in connection with any other materials or in any process.

Over the last few years, various potters began to experience problems with glazes and clays fluxed with Custer spar; some glazes did not look or work the same, and some clay bodies that worked well before developed cristobalite dunting. Gradually a pattern evolved that led some of us to suspect that the formula for Custer spar had changed.

We gathered samples of Custer spar from various ceramic studios and had them analyzed. Much to our surprise, the samples of Custer purchased after 2000 differed from those purchased before—the recent samples had about 2.5% less potassium.

We compared two samples of Custer spar purchased before 2000; sample A from a supplier in Toronto, Ontario, Canada; and sample B from a supplier in the Philadelphia, Pennsylvania area.

The lab analysis results were as follows:

When compared with the official analysis from the Pacer Minerals web site, the average of these two samples compares essentially with the Pacer typical analysis, especially in the amounts of potassium and sodium (K2O and Na2O).

Samples were also gathered from all over North America—samples 1, 2, and 3 came from the Pacific Northwest, samples 4 and 5 came from the US Southwest, sample 6 came from the Toronto, Canada area, and sample 7 came from Florida. These samples were all purchased between 2000 and 2012 by individual potters.

Trace amounts of TiO2, MnO, CrO3, and V2O5 that appear on the lab reports have not been listed.

Results: So a quick glance tells us that SiO2 is up, Al2O3 is down, iron is up, K2O is down and the rest have not changed much except the amount of P2O5, which is not mentioned in the Pacer Minerals typical analysis.

Using Custer Feldspar Now

To illustrate what is necessary to equalize two glaze recipes, one using the pre-2000 A and B Custer and one using the current Custer being mined, I have developed a cone 10 glaze with 30% A+B Custer then have adjusted the recipe using glaze calculation software, which shows lower potassium to match the original recipe.

The adjustments do not show a great difference and many potters may not even have noticed, but the calculated expansion went from 411 down to 384 in the unadjusted version. What if this glaze had started with an expansion of 380 and dropped 27 points down to 353 into the area where shivering and glaze dunting can happen? This is especially critical in high-fired stoneware clay, which may be on the edge of cristobalite development because it already has marginal amounts of sodium and potassium. This would be even more serious if the Custer spar was used in both clay and glaze.

(These are example recipes only. I have not fired or tested these fictional glazes.)

The following three examples show what is needed to make the adjustment for the new Custer spar formula, which is lower in potassium to make sure you get enough alkali fluxes to avoid cristobalite. (These are example recipes only. I have not fired or tested these fictional bodies.)

Percent of KNaO (sodium and potassium) in #1 = 2.2, in #2 (low potash Custer) = 1.98 and % of KNaO in #3 = 2.21. When I made that adjustment, I noticed that the alumina went down (in clay bodies the amount of alumina is a good indicator of how much melting will happen), so I had to raise that as well. The silica is up a bit and the iron is down. At this point, knowing that I would need 3% more of the low K2O Custer to keep any cristobalite under control, I would probably reformulate to hopefully have a recipe that was either 50-pound bags or half bags.

Porcelain Clay Body Adjustments

In a typical porcelain cone 10 body, there never is a concern with cristobalite in feldspar-fluxed porcelains. All the feldspar easily melts any of the micro-sized cristobalite crystals of silica that develop. The problem that we would have to deal with would be not enough flux because of the shortage of K2O in the current Custer. So a typical porcelain recipe could be adjusted as follows, (note the ratio of non-plastic to plastic clay materials is somewhat changed but not by much in either stoneware and porcelain).

(These are example recipes only. I have not fired or tested these fictional glazes.)

What About Cone 6?

What does all this mean to those firing at cone 6? Not a lot. There may be some noticeable differences in glazes using the current Custer, if there is enough in the glaze.

When I calculate the difference in the Glossy Base 2 from Mastering Cone 6 Glazes (a well balanced glaze that uses 22% of the old Custer), I don’t see a big difference except in the lowering of the expansion from 392 down to 374. Upping the low K2O Custer to 26.0, reducing the silica to 22.5 and the EPK down to 16.5, brings the recipe into line with the original recipe. Notice how close the ratio and the calculated expansion are to the original. Cristobalite is only rarely a problem at cone 6 and then only because it is introduced in a raw material.


The burning question for anyone using Custer spar is, what is the real analysis of the material you have? It looks like the only way you will be able to find that information is by sending a sample to a lab. Unfortunately there is no information on the bags of Custer—when it was packaged, batch number, or any information that could be helpful in establishing the makeup of the contents. We did find information about when the bags were made and that did help somewhat in establishing some sort of time frame.

If you would like a PDF copy of the lab reports, contact the author at ronroy@ca.inter.net.

the author Ron Roy lives and works in Ontario, Canada. Additional information on Custer feldspar can be found at his website, http://ronroy.net. He co-uthored the book Mastering Cone 6 Glazes with John Hesselberth. Their book can be found at www.masteringglazes.com. The book can also be purchased as an iBook from Apple iTunes.

This article was published in the November 2013 issue of Ceramics Monthly.


Click Image to Enlarge



This item will be permanently deleted and cannot be recovered. Are you sure?

Once you submit this recipe for moderation you will no longer be able to edit its content.


Leave a Reply

You must be logged in to post a comment.

Enter Your Log In Credentials
This setting should only be used on your home or work computer.

Larger version of the image

Send this to a friend