Page 1 of 16 12311 ... LastLast
Results 1 to 10 of 158

Thread: DIY Refractory Compositions

  1. #1
    Senior Member
    Join Date
    Sep 2005
    Ancient Persia

    DIY Refractory Compositions

    Well, here it is. I won't have to keep on typing the same post over and over again! Thanks for the new forum Lionel! Oh, yes it's my first post here!

    Here's something I wrote for an online book project, explaining the basics of ceramics, refractories, and other high temperature goodies.


    Ceramic materials

    1) Nomenclature
    Before discussing various aspects of high temperature furnaces and other equipment, it is helpful to understand how ceramics actually work. Simply put, ceramics are mostly metal oxides, such as aluminum oxide (Al2O3), silicon dioxide (SiO2), calcium oxide (CaO), etc. To specify an oxide of a metal, commonly the –um, –ium, or -on at the end of the element’s name is dropped, and replaced with –a. eg; alumina, zirconia, silica, calcia, etc. Ceramics can also be composed of more complex mixtures, such as kaolin, a type of large grained clay, Al2O3 . 2SiO2 . 2H2O. A list of commonly used ceramic materials and their properties will follow soon enough.

    2) Acid, Neutral, and Basic compounds
    Ceramic compounds fall into these three broad categories. Where R represents a metal and O represents oxygen, chemicals with the formula R2O and RO are bases or fluxes (eg calcia), chemicals with the formula R2O3 are neutral compounds (eg alumina), and chemicals with the formula RO2 are acids or glass formers (eg silica). At high temperatures, which depend on the compound in question, fluxes attack acids or glass formers, lower their melting point, and together form a glass. Soda lime glass, for instance, is easily melted in a furnace because of the large amounts of soda and lime, while pure silica is much harder to melt. Neutral compounds do not flux other compounds and are not easily dissolved by fluxes, with some exceptions such as boria (B2O3). While glass is great for making household glass items, pyrex, and pottery glazes, in the furnace itself glass is undesirable.

    3)Green strength
    Clay particles will adhere to one another when wet and dried, but most particles will not. If clay is used, the ceramic will probably not need another binder, but if pure alumina, for example is used, some sort of binder will also be needed to keep the powder together until it has heated enough to form a ceramic bond.

    Ceramic objects and ceramic bonds are created by high temperature firing of powders. The powders are usually slip cast, pressed, or extruded into the shape needed, and then dried and heated. At high temperatures the mobility of molecules and ions within ceramic objects increases, and eventually gain enough mobility to diffuse across various grains of the ceramic powder, fusing the separate grains into one monolithic object. This is a ceramic bond created by what I will call solid state sintering. As the grains fuse together, they shrink towards one another, decreasing the porosity of the ceramic object, causing it to shrink. Another type of bond that occurs is called liquid state sintering (by me at least). Here, some of the components of the mixture melt into a glass, which envelops the non-melted particles, and begins to dissolve them. Eventually the solution saturates, and sometimes higher melting point crystals form within the solution, also knitting the ceramic together.

    1) Binders and aggregates
    Refractory compositions have two essential parts; binders, which are the glue used to hold the composition together, and aggregates, which are the main bulk of the composition. One could make a composition composed solely of a binder, of course, but as binders are usually more expensive than aggregates, and aggregates add desirable properties such as thermal insulation they are almost always used. Certain aggregates work best with certain other aggregates and binders; large amounts of both acidic and basic components should not both be used in the same refractory composition, as the composition may melt. A small percentage of components which are fluxed and melt are sometimes actually beneficial because they form a glass gluing the rest of the particles together (perlite for example).

    2) Commercial refractory compositions
    Commercial refractories mainly use binders based on three categories of anions; aluminates, phosphates, and silicates. Calcium aluminate, or ciment fondue, is created by the heating of calcia and alumina by charcoal in a blast furnace; the two components melt, are poured into ingots, crushed, mixed dry with the aggregate, and later cured by the addition of water. Calcium or aluminum phosphate binders are created by adding reactive alumina, or calcia to the refractory aggregate, and then adding orthophosphoric acid solution, forming a metal phosphate and water. Some compositions using alumina and phosphoric acid remain puttylike and unhardened until being fired. Furnace cements sometimes use alkaline silicates (eg sodium silicate) as a binder. Other refractory compositions, such as are used in arc furnaces for steelmaking, use phenolic resins which decompose to glassy carbon as a binder. There are many different commercial refractory aggregates; most are neutral compounds because of their greater compatibility, such as alumina, silicon carbide, etc. Where the binding agent is not ceramic (such as phenolic resins) the acidity or basicity of the aggregate is not important, and thus basic oxides with very high melting points such as CaO or MgO can be used.

    3) Homemade refractory composition
    Obviously one simply duplicate a commercial refractory composition, but it is usually more feasible to use inexpensive compounds, such as clay. Various recipes have been made which use Portland cement, which contains a high percentage of calcium silicates. The calcium oxide in Portland cement is obviously a flux at high temperatures, and thus lowers the melting point of the composition, but cement does allow the composition to cure at room temperature, whereas pure clay only reaches its full strength after being fired to its maturing point. The outer parts of a clay based refractory furnace will never reach full strength, but this is usually not a problem. Essentially the recipe consists of approximately 15-30% clay; ball clay, kaolin, or fireclay work, but not bentonite, which looses its strength at high temperatures. The rest of the composition is composed of aggregates such as sand, whose main advantage is its low cost, and perlite, which decreases the strength of the composition because of its porous nature, but greatly increases the insulation value. Paper, wood chips, sawdust, or other combustibles can also be added to increase the insulation value as they burn out. Obviously the composition used will be based on what is available; if one can obtain alumina or other higher quality refractory materials, they should of course be utilized.


    Now that you've waded your way through that, or breezed over it, here's the reason for this post. I'd like to have everyone post all of their refractory formulas that they know of, and describe in a scientific manner how they have worked. This will give us a tried and true collection of good formulas, techniques, suggestions, and ways of obtaining the needed compounds, which will hopefully be a good reference for those who are building furnaces.

    Here's my "first" suggestion for someone to try, which I probably will try myself, but multiple perspectives always help.

    The Three Component Refractory: by volume-
    1/3 clean fire clay
    1/3 fine white clean silica sand
    1/3 chunks of clean white diatomaceous earth

    The clay is the binding agent, the silica sand should decrease the shrinkage and the cost, because it's less expensive than clay, and the diatomaceous earth provides some insulation because of its light and porous nature. The mix only would need to be dampened until it loosely held together, then rammed into place.

    *You must use a clay that will mature at a very high temperature, and the higher the temperature, the better. Otherwise, the clay may melt, when really you were trying to melt metals instead. A good clay is kaolin. To check if the clay will work, look at its oxide contents, and make sure it contains no more than a few percent of fluxes. Do not use beach sand, because it probably isn't silica. If the sand isn't white, it's not pure silica, and has impurities, which may flux the sand and cause it to melt. The diatomaceous earth can be found as kitty litter in dollar stores, or as an absorbent floor drying product of some sort. Powdered d.e. will not provide insulation nearly as well as d.e. that comes in small pebbles.

    I'm on a mental holiday.

  2. #2
    Vermiculite-insulated compositions

    2-4 parts loose vermiculite (1 part by weight?)
    1 part ball clay, any refractory clay will suffice
    - Mix and moisten until stiff. Vermiculite will tend to break up due to shear forces. Let dry and fire.
    - Results: weak unless vitrified. Shrinks about 10%, useless for rammed-in-place refractory. Iron and potassium content of the vermiculite melts at a low temperature (perhaps 2200°F) and readily bleeds into surrounding uncolored refractory (furnace cement patching, for instance).

    2-4 parts loose vermiculite
    1 part portland cement
    - Mix and moisten until crumbly (portland cement isn't sticky). Let cure for a week, then dry.
    - Results: weak, even weaker after firing. Portland cement has bad bonding characteristics, especially with temperature.

    8 parts loose vermiculite
    2 part clay
    1 part portland cement
    - Mix and moisten until moldable. Ram in place, let cure over a week, then dry.
    - Results: ultimately not very refractory, but has strength, low shrinkage and insulates. Haven't fired to maturity yet.

    Fired clay compositions

    65% bauxite
    25% kaolin
    15% ball clay
    - Mix to appropriate wetness (stiff for molding, slip with deflocculant for casting). Mold, dry and fire at up to cone 30. Be mindful of high shrinkage when using hydrated alumina or bauxite. If using bauxite, beware of impurities such as iron and calcium. To use alumina, reduce to 50-60%.

    35% kaolin
    25% 100+ mesh flint or silica sand
    20% ball clay
    15% spodumene
    - Same as above; fire to perhaps cone 15 maximum. Melting point unknown; guessing cone 20. Due to high silica content, more suitable as a generic stoneware clay than a refractory product.

    65% ball clay
    19% magnesite
    16% alumina [hydrate]
    - Ball clay is very plastic; some can be replaced with a fine (100+ mesh) grog of the same material. To use some or all kaolin, reduce alumina appropriately, or use some kyanite. Same wet process as above. melts sharply around cone 20 (depending on impurities) so fire carefully above cone 10. If a quality product is made, it should be quenchable from dark red heat repeatedly without cracking.

    - A half-breed between cordierite and mullite may prove to be very stable. With even more alumina content, a three-phase product might be possible, the third being free corundum, which is conductive (cordierite and mullite alone are relatively good insulators). This may make a very good crucible, capable of melting at least bronze and possibly nickel and cast iron.

    (I haven't fired these yet. Soon as the kiln starts burning...)


  3. #3
    TIM: I also want to see a thread (somewhere) on the use of slag/dross in refractories!! I TOTALLY missed that one on the old forum... Thanks..

  4. #4
    Senior Member
    Join Date
    Sep 2005
    Ancient Persia
    Here's the recipe I used for my furnace, back when I wouldn't have known that it was a bad recipe. This is the exact recipe used by Axehandle ( member) in his furnaces.

    4.8 L Silica sand
    7.2 L Portland Cement
    7.2 L Pearlite
    1.5 kg bentonite
    10 L H2O

    Ah, where to start critiquing...

    The SiO2 is ok, and should just sit there. The portland cement is supposed to provide most of the strength or bonding between particles, but the trouble is that it looses all of its strength at high temperatures, because most of the strength depends on hydrated crystals, IIRC, and the water is driven off by heat. The pearlite is a poor material because of the fluxes it contains, but this recipe was so bad that it was probably the best component in there. I guess the bentonite is supposed to provide fluidity or help bind things together, but I used kitty litter as a source of the bentonite, and so it provided very little actual strength, and bentonite is a lousy clay anyways for refractory stuff, though it is just the thing for greensand. (I can melt little slivers of GOOD bentonite in the open with a plumber's soldering torch.) There's also way too much water in this mix. It is more of a sloppy liquid than a rammable cookie dough, which detracts from the final strength even more.

    In conclusion, don't use this recipe, or anything with roughly the same proportions as this one. The furance was weak, crumbled so badly that I couldn't smear on a top coat because it would just fall off, and melts at cone 10. (approximately 2330-2380 degrees, because the exact temperature cones soften at depends on the rate of heating.)
    I'm on a mental holiday.

  5. #5
    a mixture of:

    6 cups portland
    8 quarts perlite
    120 lbs of sand
    15 lbs of kaolin fireclay
    2 quarts water

    worked well for my original furnace. it melts at a mid yellow heat, for the most part. the lid has sustained yellow-white incandecance 3 times, and is in very bad shape. melted, glazed, cracked, crumbleing. the furnace did not survive the testng of the dual fuel burner, it melted completey through the wall, and begain to burn off the outer steel shell.

  6. #6
    Junior Member
    Join Date
    Sep 2005
    Auckland, New Zealand
    STICKY !

    dont have time to read this propperly at the moment but just skimming though it looks good and i think it should be a sticky so its at the top and everyone can see it and refer to it easy :wink: if it does get made a sticky delete my post lionel, so it doesnt get clogged up :P
    "Hey, does this rag smell like chloroform to you?"

  7. #7
    hey tim /cyrus hows about some chemical symbol for these recipes as well as the names you know just in case its known as something else over here

  8. #8
    Senior Member
    Join Date
    Sep 2005
    Wheeling, MO
    I just mixed some up that is 2 gal fire clay, 2 gal mizzou, and 2 gal silica sand. I mixed all that together then decided to add about 2 gallons of pearlite then I added about a 3lb. coffee can full of saw dust. I think I should have doubled the saw dust. Haven't fired it yet though.


  9. #9
    Small point... pearlite is the eutectoid phase in iron alloys, perlite is the puffy volcanic glass. 8)


  10. #10
    by volume:
    50% expanded perlite
    30% fireclay
    20% sand from beach (it was fairly light in colour)

    The fireclay was quite expensive so for the outer and bottom peice of refractory I used:
    50% perlite
    25% portland cement
    25% sand from beach

    It seems to hold up OK, but you can sort of scratch it with your fingernail

    I burn wood and coal and stuff in the furnace, there are no major cracks or anything, it seems to hold up ok...........

    Thanks to lionel for setting up a proper forum ;-) haha


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts