Castable Refractories and the Salt Kiln


This article first appeared in Studio Potter, Volume 4, Number 1 (June 1975).
Copyright © 1975 by Studio Potter. All rights reserved.

Some Thoughts on Refractories and Salt

by Tom Turner

When one starts thinking of building a salt kiln and wondering about refractories, three things will determine the outcome. Refractories already available, finances available for purchasing new refractories and the intended lifespan of the salt kiln. At first I used whatever hard brick I could get my hands on and I coated the interior with alumina hydrate. This retarded the reaction between the bricks and the salt vapors, but eventually the salt won. Then I tried coating the interior with a high alumina cement that I trowelled on the surface, but again the salt won. More recently I've been involved with high alumina refractories that would hopefully resist sodium destruction. I much prefer making pots to building and repairing kilns. The chief disadvantage of high alumina refractories is their cost. A. P. Green Mizzou bricks are selling between $1.25 and $1.50 and the Mizzou castable is a little over $20.00 per 100 pound bag.

Several years ago I decided to build a high alumina kiln because all indications were that high alumina was the answer for long lasting salt kilns. The kiln had a complete interior of A. P. Green Mizzou. The floor and walls were Mizzou straights cemented together with A. P. Green #36 High Alumina Cement. The fireboxes and arch were cast with Mizzou Castable and the entire kiln was G-23 Insulating brick on the outside. The Mizzou material ranges from 60-63% alumina and in brick form is very dense, strong, resistant to chemical attack and very refractory, PCE 36-37. In the castable form, it allows you to mix and use it like concrete to cast monolithic units.

The reasoning behind the castable was to cast monolithic units that would be crack free, therefore eliminating sodium and chlorine seepage into my studio. The same reasoning applied to the fireboxes in preventing liquid sodium chlorine seepage through the floor structure deteriorating the refractories. This happened in the past because prior to vaporization, there is always a period of liquefaction even with super hot burners. The liquid can cause more trouble than the sodium vapors. The reasons for using the castable proved to be as weak as the castable.

First, the arch suffered a construction crack that opened in the firings and sifted aggregate onto the ware. This crack resulted from casting from one side of the arch to the center and then from the opposite side to the center. There was enough time lapse to prevent the material from fusing or sealing together and it acted like a scratch on a piece of glass, a natural breaking line. Aside from the construction crack there developed a network of cracks which I've learned are quite natural with castables. They will form their own stress cracks if not cast in sections. My concern at that time was twofold. I couldn't allow the large crack to dribble debris onto my pots. I was also concerned that the arch might collapse. As for the fireboxes, they became very soft and crumbly although they were containing the liquid sodium chloride. So I removed the arch with a sledge hammer to learn several things.

The arch would have never fallen on its own accord. During the removal I could also see that there was a good deal of vapor penetration, even though the arch interior surface revealed no reaction. About an inch inside the arch there was a zone of crumbly Mizzou where the vapors were actually deteriorating the castable chemically. The A. P. Green representative explained that almost all castables are bonded by cement which is fluxed by calcium. The calcium reacts strongly with any alkali, especially sodium vapor. This brings me to the reason for this article. Several disadvantages of castables are: (1) they will crack in use which permits vapor penetration, (2) they are very porous because they are seldom fired to maturity in a cone 10 kiln, and (3) they are fluxed by calcium which reacts strongly with any alkali present. This applies directly to Mizzou castable but relates to probably most others. If you choose a lower temperature castable that will become denser, it will overreact with the salt vapors. If you choose the high alumina castables, they will be extremely porous. Therefore, maybe our concern should be with a refractory's density more than its alumina content. High alumina with high porosity is bad news. High alumina and high density is good. Alumina is a very porous material by itself and firing a high alumina refractory to cone 10 is like bisque firing a stoneware body.

At the recommendation of the A. P. representative I replaced the Mizzou arch with a refractory plastic. A plastic is like an unfired brick with a moisture content of about 10%. Industry uses pneumatic hammers but you can get by with a three pound sledge hammer if you don't mind beating your brains out. We used the Super Hybond Plastic because it was less refractory and we thought it might become denser and seal its surface from the sodium vapors. A real bad mistake! I spent about six hours beating the plastic into place, removed the arch form as directed and fired the arch in place with a 32 hour curing cycle plotted by the engineer. First problem was the arch sagging and the second problem was the sulphur given off by the plastic. But the biggest problem was the many cracks that resulted from insufficient beating or ramming of the plastic into place.

The sulphur chemically bonds the plastic refractory and the A. P. Green representative thought it would burn out in the curing cycle. Unfortunately for me it did not. Instead, it ruined a load of stoneware. The sulphur chemically reacted with the salt vapors and the iron in the body and reduced the pots so much that two subsequent oxidation firings could not lighten the color. The pots were resalted in another firing with little change and finally put through a cone 9 electric kiln cycle which finally made them look nice. The sulphur has been reacting each time I fired so I'm planning on replacing the arch in the near future. I may try a couple vaporized copper red salt firings to see if the sulphur might be of some help instead of just a hindrance.

Disregarding the bad experience with the sulphur, the trial with the plastic refractory substantiated my feelings that density as well as alumina content is necessary for the resistance of sodium vapors in salt kilns.

My next arch will be 4-1/2" of dense Mizzou hardbrick cemented together with #36 high alumina cement and backed up with block insulation, equals 14 inches of hard brick relative to insulating ability. An arch like this would be high in alumina and density, light weight, and I think as permanent as can be at this point in time.

There are other brands of refractories available as well as other types of refractories. This hasn't been an ad for A. P. Green, these are simply the materials I've been using and am familiar with. If you go below the 60-65% alumina range, the refractory will react with the sodium to form glaze on the interior and eventually deteriorate. If you go above 60-65% range you will have problems with porosity as well as increased cost. There is a KX-99 brick that A. P. Green makes that is about their densest high temperature brick but it's been cracking because it can't take quick firings and coolings that studio potters practice.

These are my ideas and experiences with but a few materials and I'm sure many people are experimenting with other refractories and maybe having good luck. My concern is that people starting out might avoid the problems that some of us have experienced. Special refractories are expensive, as is down-time for repairing kilns and sore backs.

Tom Turner teaches at Clemson University and operates a pottery in Liberty, North Carolina.


by Wally Smith

As a professional potter the first thing that comes to my mind in planning a salt kiln is what the costs will be to build, maintain, and fuel the kiln. Never having had the money to construct the "ultimate" salt kiln, I have made compromises and explored refractory materials which perform well and are reasonably priced.

I was able to find a free source of hard firebrick for my present salt kiln, but I didn't know how they would react in salt. I had seen a salt kiln which incorporated refractory castables in its design, so I decided to do the same.

I constructed a 50 cu. ft. sprung arch kiln with 9" hard firebrick walls and floor. Plywood forms were then constructed inside the kiln and a 2-1/2" thickness of high alumina castable was poured against the firebrick. During construction, openings were left in the brickwork to allow the castable to flow into the walls and prevent it from pulling away from the walls. The arch of the kiln is a 6" thickness of refractory castable, with vermiculite poured on top of that for extra insulation. Burner ports, damper lintels, bag walls and several other special shapes were also cast. It is much easier to cast and shape perfectly than to try to chisel the shape from hard firebrick.

The only bricks exposed to the sodium vapor are those in the door of the kiln. They are Babcock and Wilcox firebrick number 80 and are approximately 45% alumina. Their cost two years ago was 57 cents each. They have never been coated with alumina and do not stick together after salting. The soaps for shelf supports are the same brick.

The castable used was also a Babcock and Wilcox product. Three thousand pounds of Kaocrete 28-LI (50% alumina) were used in the kiln. The cost then was $140.00 per ton, which is very reasonable.

Stainless steel rods can be used to reinforce refractory castables if they are first coated with wax, which will burn off and allow the steel to expand slightly without cracking the castable.

The kiln has been used professionally for two years and is completely free of sodium deposits except for the fire boxes, where there is some build up. I do wash the fireboxes of my kiln each firing with a mixture of 80% alumina hydrate and 20% Georgia kaolin. This is also used on my shelves and posts. My pots are also wadded on this same mixture in plastic form.

Just a word about castables themselves. There are several methods of applying castable. They can be gunned, rammed, or poured. Pouring is by far the easiest for the studio potter without elaborate equipment. Castables come in a coarse aggregate form and is mixed with water. The more water used, the weaker the castable. I found the best results came from adding water until the aggregate almost lost suspension and started to settle. It required a couple of minutes to mix 50 pounds in a five gallon bucket with the aid of a jiffy mixer and a 3/8 inch electric drill. The castable begins to harden in about 30 minutes, so entire sections must be mixed and cast in that period of time to be free of seams - this is imperative in the arch! Castable separates readily from wooden forms so there is no need to coat them with a release medium.

I do not know how this castable will hold up over the long run. It is in excellent shape now with 90% of the surface free of any salt deposits at all. Of course, if the castable does break down at some point in the future, the present lining can be taken out and another cast in place, yielding an essentially new kiln for about 10-25% of the cost of high alumina brick. Or, the kiln could be taken down and the brick used in some other type of kiln because they are free of salt.

Another big advantage to the professional potter is that a kiln of this type uses the same amount of salt each firing, yielding uniform results.

The only disadvantage in using castable is the extra time required to construct the forms.

There are many other types of refractory castables available, ranging from 40 to 95% alumina. The higher the alumina, the higher the cost.

Insulating castables are also available for use in reduction kilns. Arches and doors can be cast in one piece. They will have tremendous strength, be free of sag, and have high insulating properties.

I have always found the refractory company engineers to be very helpful in calculating amount of castable required per job, arch thickness, etc. Before calling a company for information, always have exact dimensions of your proposed kiln, including span of arch and arch rise.

It is impossible to explain everything about this kiln in a short explanation like this. Visitors are welcome to visit my studio, The Potters' Mark, in Gatlinburg, Tennessee, or I will accept telephone calls if there are questions.

Wally Smith is a production potter working in Gatlinburg, Tennessee.

Cast kilns
Building a kiln with castable refractories at Franklin Pierce College, Rindge, N.H. Erected by David MacAllister and friends. While not a salt kiln, this illustrates the building method used with castables.

Notes on Castables for a Salt Kiln

by Richard Leach

Kiln building for the ancients wasn't as simple as with the plans, kit kilns and quality controlled commercial refractory products available today. When I saw a scove kiln constructed of the same material being fired and clay being mixed, blended and formed and the rows of drying adobe brick at one site, I began to wonder about the necessity of depending or having to depend on rather expensive refined resources for kilns. The technology for building the kiln is present in the knowledge of blending clays and forming pottery. Primitive peoples tempered their clay with a grog of whatever was available and with an organic binder, usually finely chopped manure or straw and grass. The primitive pottery dries quickly, is fired fast and has excellent thermal shock resistance when used directly in the fire.

After having come in contact with Southwest Indian pottery and primitive potteries in Mexico, I decided to try some mixtures of inexpensive (cheap) locally available materials to construct or cast a kiln suitable for stoneware temperatures. Bricks of locally available clay, refractory filler and organic binder were made and bisqued for testing. The drying and bisquing period became an anxious wait so I fired tests green, some air dried, some merely leather hard. These tests were brick shaped blocks of clay placed in the wall of a test kiln-the tests were not fired in the kiln.

There was some difficulty in handling due to crumbling some of the test blocks, so with the addition of a small amount of Portland cement to hold things together, the bricks could be moved and inserted in the test kiln wall. Realizing the calcium silicate of Portland cement could be a powerful flux at higher temperatures, I checked out the proportion of alumina in the refractory cements and added a measure of alumina hydrate to the mix to approximate a prefired refractory concrete such as Luminate refractory concrete.

Of the four climbing test kilns I built at Albion in 1970, the first two were without alumina. Kiln #1 is still functioning after 5 years, but shows signs of wear. Kilns 1 and 2 are used only for stoneware. Kilns 3 and 4 were built with 4% alumina or about 1/2 measure by volume in a 1:6 or 1:7-1/2 ratio (1 cement to 2 clay, 2 grog, 2 or 2-1/2 sawdust-vermiculite). Kiln 3 is for stoneware and kiln #4 for salt. After over 100 firings, kiln #4 is in better shape than chamber #3. Neither has secondary insulation and both retain heat as well as a K23 I FB.

Some suggestions for a successful cast kiln: Incorporate a ratio of Portland cement to aggregate.

For a salt chamber, 1/2 to 1 part of the aggregate is alumina hydrate.

50% of particle size of the aggregate should be 1/4" or smaller with the maximum size aggregate roughly 1". Crushed bisque ware screened through hardware cloth works well for stack sections or floors.

Use a fairly dry mix. We used about 5-1/2 gal. water to a batch of 160# with the volume equivalent of 2-1/2 gal. pail: 1 part crushed cobs (or sawdust), 1 part vermiculite, 2 parts fireclay, 2 parts grog, 1/2 part cement and 1/4 to 1/2 part alumina.

If a portion of the mix placed next to the form be dampened more and the dry mix rammed from the outside, a more smooth and dense surface will result on the inner lining of the kiln. Too wet a mixture makes for a dense arch that will dissipate heat too quickly, prolong the firing, use more fuel and break up easily. Too dry a mix will not possess initial dry strength to maintain shape. A correctly mixed batch will maintain its shape after being squeezed. (Cast against oiled metal or tar paper on the inner form).

A high alumina fire clay is to be preferred over a fireclay high in silica if the kiln is to be used for salt (A. P. Green DM Mo. fireclay). High heat duty grog, high in alumina (A. P. Green P Grog) is preferred for the same reason. It may be worthwhile to make your own grog by high firing crushed bisque to cone 6-7-8 or by calcining a rough ground fireclay. Check manufacturers for analysis and mesh size of local materials.

Curing the kiln doesn't make too much difference. Some kilns have been fired within 12 hours of ramming the arch, back and doors.

Up to a point, the more times the kiln is fired, the better insulation the wall becomes. After 100 firings in one kiln tested, the temperature was 300 deg F less at a point halfway through a 4-1/2" wall than in the original six firings.

The cast arch can be further insulated with a mixture of clay, vermiculite and cement after it has been fired enough times so the chemical water and carbon have been driven off the original casting.

Luminite has been used in several of these kilns. There does not seem to be any difference in the lasting quality whether used for stoneware or salt.

Sawdust from a chain saw or saw mill is better than table sawdust. It provides a more porous texture for better insulation.

Oat or wheat or rice chaff is an excellent substitute for sawdust and is preferable if it is available.

Dick Leach teaches at Albion College in Albion, Mich.