Sciencemadness Discussion Board

castable refractory mixtures

Rosco Bodine - 29-11-2004 at 22:38

There have been other topics about the
construction of high temperature furnaces
and I also have a need to build a furnace .
One of the things which can facilitate the construction is castable refractory materials . There are a few materials
which I am considering as candidates for
a refractory mixture which can be dampened and then tamped into place .
Some of these may have been used in some combination before with good results by others and I am asking for any
proven formulas or suggestions or links
that others may offer .

Diatomaceous earth is one material I plan to use as a component of the mixture . It has a very high silica content and is a refractory material , and it is available at reasonable cost as a swimming pool filter material .

Perlite is another material I plan to use in the mixture , and this has already been used by others .

A fusible clay material like fireclay will be an ingredient . I am considering using
porcelin slip , or some other ceramic slip
for this ingredient .

A portland type cement will be used . I have some type I portland cement , but
I noticed the white sanded grout used for tile setting , and also the white mortar mix
for setting glass blocks , and I wonder if these more highly refined low iron and light colored mortars would be better .
I also spotted a surface bonding mortar mix which is light colored and contains chopped glass reenforcing fiber .

Sodium silicate solution , magnesium oxide , hydrated lime , chopped fiberglass
"kitty whisker" , cabosil , and glass microballoons are also available materials .

Any suggestions for a mix containing DE and perlite and perhaps any of these other materials ?

Organikum - 30-11-2004 at 02:47

Usual Portland-cement cannot withstand high temperatures for a prolonged time.
There are special furnace cements ("Trass-Zement" in german) available for cheap for this though.

Phosphor bonded cement was suggested by Tacho and used successfully by Axehandle IIRC.

Rosco Bodine - 30-11-2004 at 06:38

Hmmm...about the portland cement then ,
I wonder if the tile grout or the glass block mortar would be better , or if it contains a phosphate and is different from regular portland based mortar .
Portland is a dark gray material , but the
others are absolutely white in color .

Also I have a half gallon of 80% phosphoric acid and I wonder if perhaps
that may be used to create a phosphate
bonding component in situ , perhaps with
dolomite or calcium carbonate or zinc carbonate . The byproduct CO2 would act as a foaming agent and give the material a cellular structure and insulating properties .

[Edited on 30-11-2004 by Rosco Bodine]

Tacho - 30-11-2004 at 08:24

There has been extensive discussion on this subject, mostly in furnace & foundry threads. A search will help you.

Phosphate bonded cement is sold in dentistry supply shops as "dental cast investment" or something like that. Very available.

I could never get any bonding using 85% phosphoric acid and CaO. This also was discussed elsewhere.

Axehandle has discussed this topic a lot in other threads.

Eclectic - 30-11-2004 at 09:12

There is an aluminum silicate cement called "cement fondu" made by Fondag that is used in foundrys. I think it is good for up to 1600 deg C. A little phosphoric acid in the mix should react with the alumina in the clay to make a good bond. Use maybe
0.5% of the total mix weight It might need firing to develop the bond.

Rosco Bodine - 30-11-2004 at 09:47

Regarding the phosphoric acid , I was thinking perhaps some ammonium salt ,
or some partial neutralization with ammonium hydroxide may facilitate the
process . If there is also used some magnesium oxide or perhaps dolomite ,
along with some chloride containing material , there is the potential for forming
a magnesia type cement component , such as the oxychloride . A similar effect might occur for a related sulfate magnesia
type cement . Anyway the chemistry of what happens to the cast refractory at firing temperatures is complex and unpredictable to an extent . So the mixtures are something of a trial and error "witches brew" .

The use of diatomaceous earth with perlite , perhaps with some scheme for foaming the mixture , and use of the white
mortars like grout or glass block cement ,
and ceramic slip is not anything I saw in
the other threads about furnace construction . The topic I started here is
an effort to focus on the refractory material mixture itself , in particular what
method of bonding cement would be best for a filler material of diatomaceous earth and perlite , and hopefully to find readily available OTC materials that will serve the purpose . I could make a six hour drive to an industrial supplier and buy any of the standard refractories and pay high prices if I wanted to build a professional grade furnace . What I seek is a method
which uses the most mundane materials
from the local building supply and hardware store , not medical or industrial
specialty suppliers . If a satisfactory mixture can't be made without specialty
cements , then it would be easier to simply build a stacked box furnace from firebrick and seal the joints with that fireplace grout which comes in tubes for a caulking gun . I have on hand some angle iron and stainless steel strapping
to build a permanent frame for a stacked firebrick furnace , if I want to build a long term use and very sturdy forge type furnace . More the interest of my topic
begun here is a refractory mix which would suffice as something relatively easy
to use as an insulating liner for a more improvised furnace like would be made inside a metal five gallon bucket or a large clay flower pot .

The operative principle here is *improvised* refractory mixture . I am not looking for recomendations for commercial refractory products , but for
common OTC products which may be pressed into use in formulation of a good refractory mixture .

Refractory mixes and chemical compositions

chemoleo - 30-11-2004 at 11:03

Hmm, I am not sure whether to merge this with Cyrus's thread on metal casting molds. Let's see how this thread evolves I guess.

Anyway, I am (like most others) also quite interested in refractory mixes.
I have various oxides (i.e. CaO/SiO2/ZrO2/TiO2/Al2O3/MgO) and am thinking of using one of the commercially know mixtures.
For instance, tables on this, this, this , or this page I find very useful - although the first one in particular uses too much ZrO2 for my liking.

I wondered 1) whether it would make sense to mix all these oxides according to the published amounts, and fire it up with a bit of water? and 2) whether anyone knows more tables such as the above, with various properties & firing temperatures, thermal expansion and so on?

[Edited on 30-11-2004 by chemoleo]

Rosco Bodine - 30-11-2004 at 12:37

As a bubble producing agent for creating
a cellular structure , perhaps a bit of hydrogen peroxide or one of the persulfate or percarbonate chlorine free bleaching materials would be useful added
to the mix .

Twospoons - 30-11-2004 at 13:11

Voids in the cement are usually created by adding a combustible material like sawdust, then burning out once the cement is hard. Gaseous voids in wet cement cannot be 'rammed', and if the cement is pourable then all your bubbles will rise to the top.

Rosco Bodine - 30-11-2004 at 13:34

I have read about the use of sawdust but I don't favor that method because it would not produce a microcellular structure like a gas entrainment methid would provide . There are "air entrained"
concretes which are common . So long as the mix is thick , the bubbles do not migrate . And the mix I plan to make is
going to use minimal water , just wet enough to hold itself together like a damp and sticky mixture as would be used for a sand casting mold , maybe just a little wetter but not much . About like cookie dough , if I threw a ball of it against a wall it would stick there and dry where it hangs without falling off .

Rosco Bodine - 30-11-2004 at 21:28

Well this is very interesting . See the attached pdf file . Sodium metasilicate powder is cheaply available as a cleaning detergent , used for degreasing surfaces to be painted , a cheaper substitute for trisodium phosphate . It sits on the shelf in the paint and wallpaper department in building supply stores .

Attachment: soluble silicates in refractory mixtures.pdf (112kB)
This file has been downloaded 5616 times

Rosco Bodine - 30-11-2004 at 22:21

Here's more info on soluble silicates as binders and coatings from the same source

Attachment: soluble silicates as binders.pdf (121kB)
This file has been downloaded 6636 times

Rosco Bodine - 30-11-2004 at 23:24

Here's the page where I have been getting this interesting data on silicates .
From what I have read it seems the higher silica to sodium ratio silicates would be preferable for use as a binder to the more basic metasilicate commonly used as a detergent . The higher ratio product is likely the OTC product sold as an engine block sealer at automotive stores , or sold as "water glass" in some hardware stores . Possibly some concrete sealers are also the higher ratio
sodium silicate preferable for use in refractories .

Here is an additional file describing some of the history and general chemistry of soluble silicates .

Attachment: soluble silicates general chemistry.pdf (108kB)
This file has been downloaded 2407 times

tokat - 1-12-2004 at 01:05

I think water glass is sodium hydroxide and HCL and water. I think its a gel.
Is it a gel?

Tacho - 1-12-2004 at 04:45

Water glass is sodium silicate.

IIRC, if you add NaOH or HCl to it's solution, you end up with silica GEL, which, btw, never looked like a gel to me. Curious...

rift valley - 5-12-2004 at 07:03

I don't know if this has been mentioned but i was just at home depot and noticed that they sold refractory cement, it runs 14 bucks for a gallon (about 4 liters). A little expensive for my blood

Rosco Bodine - 5-12-2004 at 08:26

Sounds interesting , if only for use as part of a blend or as a bonding agent . What department and was there any brand name , was it dry or liquid ? Refractory materials are a specialty item and many of them have limited shelf life so as a rule they are all expensive , which is a fact that motivates the desire to find an improvised mixture which would substitute
for the manufactured products .

I have been reading some interesting patents about refractory mixes which are
perhaps satisfactory . I will post the numbers here for some of those patents which I have been reading , in trying to understand the strategies others have used , and adapt or combine such information for experimental compositions .

There is an amorphous fibrous Aluminum Silicate product which is used as a spill cleanup and sweeping compound , and may be a valid substitute for "rock wool"
or loose ceramic fiber as a reenforcing material . The product is called Abzorbit .

An interesting fact which I did discover is that portland cement can be used as a component in refractory mixtures , but only if the chemistry of its setting up is modified to produce different intermediates and/or high temperature products of the reactions than would be obtained from the ordinary reactions which are produced in concrete or mortar .
So , to a limited extent portland can be used as as component in true refractory mixtures , but its usage in that regard is highly qualified by the conditions and the other ingredients also used with the portland .

GB1449484 Refractory Compositions

Some of my thoughts about foaming agents have also been confirmed valid ,
for example use of hydrogen peroxide has been used successfully .

In no particular order here are some patents which I have been reviewing in connection with this topic .


Have fun reading !

Especially interesting to me are the compostions involving the reaction between hydrated lime and diatomaceous earth , along with sodium silicates which
could be produced in situ as a preliminary
reaction with a caustic alkali . The aluminum phosphate compositions are also very interesting .

[Edited on 5-12-2004 by Rosco Bodine]

Geomancer - 5-12-2004 at 14:44

Obviously, the choice of refractory depends on the temperature range you want. Glass, of course, softens at relatively low temperatures. I wouldn't recommend it for anything hotter than aluminum casting. You mentioned porcelain slip, so I assume you ave access to a pottery supply. The folks there probably can give you good info.

My furnace actually used regenerative cooling instead of any refractory. This was enough to melt copper easily, though the steel wall oxidized somewhat badly. The device is now in disrepair, having self-destructed in an attempt to melt iron.

Alumina based refractory....

kazaa81 - 7-12-2004 at 15:17

Hallo to all,

I would suggest that alumina (Al2O3)based refractories are good.
The problem may be get alumina.
It can be made in a simple method:

-make a soluble Al salt
(ex. 2Al + 6HCl ---> 2AlCl3 + 3H2)

- add aqueous NH3 to this solution
AlCl3(aq) + NH3(aq) -----> Al compound
the addition of aqueous NH3 to Al salt will yield a gel salt of a formula that in this moment I don't remember.

-calcination of gel
the hot heating of this gel will yield powder alumina

then you can made refractories in what shape do you want.

Have fun ;)

neutrino - 7-12-2004 at 16:53

I’d guess that the following happens:

AlCl<sub>3(aq)</sub> + 3NH<sub>4</sub>OH<sub>(aq)</sub> -> Al(OH) <sub>3(s)</sub> + 3NH<sub>4</sub>Cl<sub>(aq)</sub>

Upon drying and heating:

2Al(OH) <sub>3</sub> -> Al<sub>2</sub>O<sub>3</sub> + 3H<sub>2</sub>O
NH<sub>4</sub>Cl<sub>(s)</sub> -> NH<sub>4</sub>Cl<sub>(g)</sub>

BromicAcid - 7-12-2004 at 17:25

Alumina could be had quite easily if one wanted it, take a large container, make a trip to a scrap yard and buy 15 kg of aluminum scrap, take a small amount of mercury (3g) and spread this over aluminum in different places, give it a few weeks and you should have converted most of your aluminum to the oxide.

I remember reading about using aluminum powder as a foaming agent in Portland cement, it increases the volume of the cement 2x and the cellular structure adds the necessary strength to accommodate that expansion. Fumed silica also adds more strength then even it's most powdered of brethren.

Twospoons - 7-12-2004 at 19:23

Your local ceramic supply shop is once again your friend! Alumina, and a lot of other refractories are sold in pure form (relatively speaking), as fine powders for use primarily in glazes. There's really no need to go to the effort of making the stuff yourself.

Rosco Bodine - 7-12-2004 at 19:53

A major ceramics supply distributor would likely have such components available to the minority who formulate their own special mixtures . But most of your local ceramics shops do not have and do not use pure component ingedients , but use proprietary mixtures and products that have already been mixed for some particular use .

The use of mercury is something I actually have considered as a way of producing alumina in a finely divided form for use in refractory mixtures . There would probably be some percentage of the hydroxide from the effect of moisture in the air mixed in with the oxide , which is evolved fairly rapidly from the surface of amalgamated aluminum exposed to the air . Actually amalgamated aluminum is one of the best dessicants known for this reason . But anyway , if a chunk of structural aluminum was amalgamated and placed upon a grate supported over a tray to catch the loose effluent oxide falling from the aluminum as it steadily disintegrates , that should work well ,
and would yield a more pure product in terms of the oxide , if the process was done during dry weather . The best way of doing this would be to first degrease and etch the aluminum with some oven cleaner spray of the sodium hydroxide type , then rinse away the cleaner after
a few minutes and dry the aluminum .
The aluminum is then dipped into a solution of mercuric chloride which is made acidic to a slight degree with HCl .
The aluminum will rapidly , almost instantly take on a mirrored appearance as it amalgamates . A fairly heavy layer of mercury will be desirable for good reactivity with air . The chunk of aluminum
will be evolving hydrogen at a steady rate and the HCl will be converting the reaction product to a soluble chloride to
keep the surface clean and facilitate an even amalgamation . The chunk is removed and placed upon the grate .
If the chunk is removed before it is reduced to a size small enough to fall through the grate , nearly all of the mercury can be recovered by dropping the
remaining piece in dilute HCl . When all the aluminum remaining has been converted to soluble aluminum chloride ,
the mercury metal will be found on the bottom of the beaker where it may be saved for recycling . I wonder what the actual composition of the product would be in terms of oxide to hydroxide . This could indeed be a good quick and dirty method for aluminum oxide . A low intensity vibrator acting on the support grate would perhaps be necessary too ,
since the oxide comes off the amalgam
like a white snake , and it is very lightweight . The reaction is very interesting to watch , as it produces an
unusally large volume of oxide from a small volume of metal .

The amalgam might be useful also for producing aluminum phosphate gel which could have value . Perhaps an aluminum silicate gel could be also made by way of aluminum amalgam . These materials could have great usefulness as a binder component in refractory mixtures . The aluminum alcoholates like aluminum isopropoxide which forms rapidly and easily might also have use in refractory mixtures as a reactive anhydrous intermediate .

[Edited on 8-12-2004 by Rosco Bodine]

chemoleo - 8-12-2004 at 09:02

Yes, I ordered some Al2O3 lately... it came at 15 £ (25$) per 5 kg - not that cheap. Got some SiO2 powder as well, for that very purpose too. This, if mixed & heated at proportions given in the links from my last post might hopefully make a decent refractory material.
In fact, some of the mixtures contain SiC (silicon carbide), which can be also obtained at pottery supplies, at about the same price as above. This is used for high temp refractories I believe.
ZrO2 (zirconium oxide) is also used for extremely high temp applications, I managed to obtain it for 15 £ for 500 g - definitely not cheap, so for most this isn't an option. Zircosil (ZrSiO4) is presumably anothe good material, and it is available at much cheaper prices.

One thing that I yet have to find is perlite.

I am thinking of adding sodium silicate solution, to make the binding better, and to possibly calcinate it a little with CaO.

Pottery supplies rock - wouldn't know what I'd do without them!

[Edited on 8-12-2004 by chemoleo]

Rosco Bodine - 8-12-2004 at 11:36

There are likely some very good refractories which can be made from relatively common OTC materials without even having to go to the commercial suppliers , and of course that is my primary interest , improvising a decent quality product from mundane materials .
Once you have a working furnace , you can actually use that furnace to process by calcination the raw materials available into refractories for higher temperature refractories , to use in building a better "second furnace" . So a small furnace even having a lining of a limited useful life , allows you to process materials for building a much better and higher temperature furnace as a subsequent project . There are hundreds and more hundreds of patents concerning refractories , and I am about halfway through the task of picking out a few of the more interesting ones , in terms of their adaptability to use of OTC materials for experimental "home made"
refractories . Diatomaceous earth is one of the raw materials which interests me greatly , but it may present problems due to expansion and crystalline transitions . I haven't tried any actual mixtures yet , so it may be workable or not , using the uncalcined , pool filter grade of D.E. as a component for refractory mixtures . A patent for a D.E. and perlite aggregate refractory bonded with aluminum silicate is


Another improved silicate binder which may work for similar mixtures is a trisodium phosphate / water glass mixture


Something as simple as pure quartz sand containing very fine quartz flour and perhaps some colloidal silica can be bonded by phosphoric acid alone , which upon heating forms a binder matrix of silicon phosphate . Fumed silica , cabosil , may perhaps work in such a scheme .


The technical dificulty to be overcome with refractory mixtures is to address the structural instability which appears in the
temperature range which is intermediate ,
as the material is being heated sufficiently high for the slight melting and sintering which accomplishes the permanent and stable "ceramic bond" by vitrification of the matrix which cements all the particles together firmly . It has to be a carefully chosen mixture which will hold together as it it is heated through the temperature range upwards , so that a smooth transition occurs from the chemical bonding of the "cold bond" which holds the unfired material together , to the ceramic bond produced at much higher temperature . The trick to be accomplished is that the zone of transition is a stable formation which doesn't result in crumbling or separation between the cold bonded material and the ceramic bonded material . Ideally there should be no discontinuity in the bonding through the transition zone from the "hot face" through the thickness of the refractory all the way to its cooler outside surface , which never really gets hot enough to form a ceramic bond . If this technical difficulty cannot be addressed , then a furnace must be built from separately made and fully fired brick ,
as opposed to using a "fire in situ" refractory composition serving as a furnace liner . The intended use of the refractory has much to do with whether a particular composition will be suitable or not , not just the upper temperature limit ,
but the manner it will be subjected to that temperature , as applied to one surface or as a surrounding environment .
Refractory compositions are indeed a science all to themselves . A patent which describes the problem and a solution for the intermediate temperature zone difficulty is


Update : With regards to the use of aluminum amalgam , there is another useful material which may be obtained by treating the metal first with hydrochloric acid to form aluminum chloride solution , and then subsequently to slowly add phosphoric acid to the stirred solution to
form aluminum chloro-phosphate hydrate ( ACPH ) , which is a highly water soluble and effective refractory binder . ACPH is described in several patents ,


The aluminum chloride starting material could probably also be obtained more easily from aluminum sulfate and calcium chloride , or conversion of the sulfate to a carbonate and then treating with hydrochloric acid . Anyway these are some alternative routes for producing an ACPH refractory binder from either the aluminum scrap metal , or the commonly available soil acidifier aluminum sulfate .

[Edited on 9-12-2004 by Rosco Bodine]

rift valley - 11-12-2004 at 07:51

I made my first mold today. It came out better then expected but I don't think the Al was hot enough. Do you guys know of any inexpensive ways I can make a thermometer that could withstand molten Al. It would make the whole process much easier if I knew the them. of the melt. I have been taking a bunch of pictures of the whole process, I can post them if anyone else is interested.

BromicAcid - 14-12-2004 at 11:01

Check out this video from popular science.

You can click on the link on the page to play the video, an aluminum I-beam gets ate up badly over the course of an hour, the link is about how aluminum amalgamates with mercury to form aluminum oxide which flakes off, just as we discussed earlier in the thread.

Rosco Bodine - 14-12-2004 at 11:26

Probably so as not to greatly publicize the truth of what is an obscure and highly destructive knowledge which would not be wanted to be better known , the authors fudged on the technical details in their misinformation about "mercury paste" . The active agent is not any such fictitious "mercury paste" at all . The amalgamating agent is in fact a liquid which consists of corrosive sublimate ( they don't call it that for nothing ) which is mercuric chloride , dissolved in hydrochloric acid . The liquid is effective as an aerosol , a stream from a "water pistol" or syringe , or dripped or painted onto the aluminum object to be destroyed . The reaction zone migrates into fresh unreacted material , and the white oxide comes off rapidly like a white "snake" similar to a pyrotechnic snake . Five cents worth of the corrosive "weapon" can very quickly destroy millions of dollars worth of aluminum whatever , and do it silently ,
covertly , reducing the target to a large pile of fluffy white powder .

Polverone - 14-12-2004 at 14:41

Interesting. Just how quickly does it work? It's something to keep my mind nervously alert on my next trans-Atlantic airline flight.

I've seen HgCl2 applied as a methanol solution make aluminum do the flaking into powdery oxide trick, but the effect didn't seem truly catalytic. The reaction stopped after a time, while there was still plenty of (beverage can) aluminum left. I supposed that the mercury was being carried away mingled with the aluminum oxide.

Rosco Bodine - 14-12-2004 at 15:02

The air reaction is definitely a visible rate , and to my eye looks more rapid than the rate of attack of concentrated NaOH upon unamalgamated Al . Would have to take a fairly heavy piece of aluminum barstock and time the reaction to see exactly what the actual rate is and it is likely dependant upon temperature and humidity . IIRC the white powder snake would come off the surface of the Al at a length of the ashlike material of maybe one inch per minute , so it isn't all that slow . It is definitely fast enough to consume an aircraft in a time measured in hours not days .

BromicAcid - 14-12-2004 at 15:51

I think the time lapse on that video at that site is 1 hour. As long as the inital concentration of mercury is low this would be a good way to make the aluminum oxide as the starting material for refractories, although I doubt it, does a similar thing happen for magnesium? I've never heard about it though.

[Edited on 12/14/2004 by BromicAcid]

Rosco Bodine - 14-12-2004 at 16:15

Never tried magnesium but it would probably be even more vigorous based upon the activity , however the solubility rate of the metal in the amalgam is the limiting factor . Also any impurities in the alloy could cause complication , so the more pure the metal the better . The mercury layer should be fairly heavy for best results , a good fluid layer of liquid mercury in excess of the amount required to maintain the ongoing solution of the metal , so that the amalgam maintains a freely fluid consistency and doesn't form a sludge with the metal being dissolved and transported through the layer . The mercury layer is actually in fluid motion as it completes a loop from the metal being dissolved upwards to the air where it is dumped as oxide , and then returning to the surface of the metal to pick up more , sort of like a conveyor belt for metal ions .

rift valley - 15-12-2004 at 20:19

I need to make a lid for my furnace, I am going to use a terra cotta flower pot base and I will fill it with some fireplace refractory crack repair cement mixed with mostly perlite, just enough cement to hold it together. I saw this on another website for a refractory mix but it didn't look too sturdy, for a lid it should be fine though. I'll let you know how it works out

Rosco Bodine - 15-12-2004 at 21:39

There is a cheap material that is used for landscaping on baseball fields and golf courses , to surface running tracks , maybe tennis courts , ect. to keep rain from making the ground muddy and slippery . It is a fullers earth material sort of like kitty litter , but it has been high fired in a furnace until it vitrifies partially , and forms a porous ceramic gravel like material . It is called Quick Dry , made by a company called Profile , and sells for about six dollars for a fifty pound bag .
A variety of the material is sold as soil for aquatic plants , because it won't disintegrate in water like ordinary unfired clay . This grog like material may have usefulness as an aggregate in refractory materials . It is principally magnesium aluminum silicate granules which are 75% empty space , about half open pores and half closed pores , a hard gritty material of density 2.5 . I haven't been able to find anything about it being used as a refractory component by googling it , but it would seem like a good candidate refractory since it is 74% silica and 11% alumina , along with other high melting point materials in the remainder . And it is cheaper than perlite or vermiculite , plus it has structural and dimensional stability , in that it does not swell from moisture . There are several mesh sizes available , but the Quick Dry is the finest mesh . I will probably try this material as at least part of the aggregate and insulating material in experiments with refractory mixtures . My idea here is that it should give strength and good thermal properties cheaply , as well as providing porosity for escaping water vapor when the "green" material is baked out and brought up to vitrification temperature .

Cyrus - 16-12-2004 at 18:09

It has fluxes in it. (MgO)

Fluxes aren't good in refractories.

The furnace will probably melt at a lower temperature than you'd like.

(Yes, I'm back.);)

Rosco Bodine - 16-12-2004 at 18:59

It may have fluxes , but I have no information to suggest it does . The labeling says it is 100% blended clays which have been calcined at high temperature . Anyway MgO is not a flux . MgO is a very high refractory used for embedding the nichrome element inside metal sheaths for hotplates , and also used to line high melting alloy electric arc furnaces .

Maybe you are thinking of PbO or some of the alkali oxides or fluorides .

BTW , I spotted something interesting on the shelf in the pet section , a kitty litter that is pure silica gel , snow white material having refractory properties , melting point 1700 C . It is sort of coarse
about like aquarium gravel .
I have seen much finer mesh silica gel sold in bulk in hobby shops for use in drying flowers for dried flower arrangements . A mixture of these silica gels maybe with perlite and perhaps DE along with a compatable binder would make a good refractory .

[Edited on 17-12-2004 by Rosco Bodine]

Cyrus - 17-12-2004 at 17:40

Here's a quote from

MgO (Magnesium Oxide, Magnesia)
Family: Flux
Weight: 40.300
Expansion: 0.026
Fusion: 2800C

(Sources: Talc, Dolomite, Magnesium Carbonate)

-Together with SrO, BaO and CaO it is one of the Alkaline Earth group of oxides. It has a cubic crystal structure.

-Like CaO, MgO is refractory at lower temperatures, so much so that it can be used to increase opacity, to perform as a matting agent (i.e. magnesium carbonate), and act as a check to glaze fluidity in a manner similar to alumina (e.g. to prevent devitrification or the tendency to produce crystalline surfaces). When mixed with CaO, it is not as refractory.

-It can act as a catalyst in low temperature bodies assisting the conversion of quartz to higher expansion cristobalite (which reduces crazing).

-In high temperature glazes it acts as a flux (beginning action about 1170C) producing viscous melts of high surface tension and opaque and matte glazes. Like CaO, its melting action drastically accelerates at high temperatures.

-The surface tension of MgO-containing melts is less of a problem in reduction.

-Zircon and Magnesia melt at 2800C making them the highest melting oxides. Remarkably, MgO readily forms eutectics with other oxides to melt at surprisingly low temperatures.

-It is valuable for its lower expansion and crazing resistance. When introduced into a glaze it should preferentially replace calcia, baria, and zinc before the alkalis to maintain surface character. Adding too much will generally move the surface texture toward matte or dry.

-MgO is a light oxide and generally is a poor choice for glazes to host bright colors. However, it does work well in earthtone and pastel glazes, especially in high temperature reduction firing. Likewise, it may be harmful to some under-glaze colors.

-Does not volatilize.


I don't have my oxides confused. :P

In other places the fluxing action is not mentioned at all so I understand the confusion, and yes it's used in bricks. Here's a lil' trick to remember which oxides are glass formers, which are fluxes, and which are neutrals.
R= metal O=oxygen

R2O- flux
RO- flux
R2O3- neutral
RO2 - glass former

Of course there are going to be some exceptions, but in general this works.

That silica gel you mention looks interesting- It's got to be pretty porous, and that's good for insulation.

[Edited on 18-12-2004 by Cyrus]

Rosco Bodine - 17-12-2004 at 19:09

I think the special case where MgO would act as a flux would be those specific proportions where the eutectic effect is the cause . A similar thing can occur with aluminum phosphate and alumina if the proportions favor the low melting eutectic , but in proportions which are comfortably away from that eutectic the same materials are refractory . The ratios are everything , the same for silicates .
Anyway at some point in the vitrification there are refractory compositions formed from magnesia and other materials , whether the magnesia is a flux or not , it does not simply melt and solvate the other materials but forms higher melting complexes which get even more solid with rising temperature . Perlite does the same thing and so do several other materials which first begin to flow and then resolidify as they react with other materials .

The factor I have been looking at carefully is the acidic or basic or neutral nature of the materials present and how the reactions proceed in the presence of moisture to produce a cold setting effect , and then what follows during intermediate heating as moisture is driven off , and vitrification occurs . The different reactions in each stage have to proceed in a way they overlap so that the composition doesn't crumble . And the products of the final vitrification reactions need to be unchanged by falling temperature and re-exposure to moisture .

There is some magnesia content in nearly all the common clays . I don't think it will be any problem in a phosphate or combination phosphate and silicate complex bindered refractory . Even DE has some magnesia content and DE is most definitely refractory . What is going on in glazes is similar but I think the difference with magnesia as a refractory is all about the proportions and the nature of the other materials being quite different . Even powdered borosilicate glass which has a lot of fluxing action and low melting point can actually be used as a binder in high temperature refractories ,
within certain proportions . Just because the powdered glass is the first thing to melt doesn't cause the aggregate with which it "alloys" to fall apart in the fire .
Actually I am trying to avoid magnesia in the refractory compositions I am considering , but for reason of pH considerations causing a short working life for the cold setting of wet mixtures containing basic refractories . I am trying to keep to neutral and acidic components to have extended working life for the wet mixture , and to accomplish the overlapping sort of binding schemes I mentioned earlier . Unless magnesia has been "dead burned" and granulated fairly coarse , it is too chemically active when wet and reacts too rapidly with binders so that many of the compositions set up much too quickly . Anyway , even though MgO is listed as a constituent in certain clays , I am reasonably certain that this is the type of mineral analysis expression as is common usage in the fertilizer industry , referencing Magnesium content expressed as the oxide , even though the magnesium is likely present as a silicate complex or some other compound that is anything else other than actual MgO , if you follow what I mean . I am experimenting with this material anyway , and it either works or it doesn't . A good old blow torch test on some of the granulated material side by side with some sand might be revealing . I can check the pH of the wetted material too and see what that reveals .

rift valley - 18-12-2004 at 07:21

These are the products that they sell at home depot, that I was talking about.

The castable refractory goes for 13 bucks a bucket and a 32oz bucket of the furnace cement is only three bucks!

This product is made in nearby Vermont so I dont know if it is a nationwide home depot product of just the ones nearby.

P.S. What do you guys know about kaowool (wool made out of ceramic shards I think?) I have seen pictures of furnaces where there is just folded kaowool and no castable refractory.

[Edited on 18-12-2004 by rift valley]

Rosco Bodine - 18-12-2004 at 08:26

Yeah that Home Depot stuff is strictly regional availability only . There are other things the same way .

The availabiliy issue is exactly why I am sorting through patents looking for the most useful mixtures of OTC materials which can be combined in "off label use" compositions which will serve the purpose . There are three or four compositions which I believe are workable in that regard , and one of the most interesting involves a phosphate modified portland cement binder . It is tedious reading through the patents and picking out the few that seem most useful . When I complete the reading and making notes , a summary of what I find will be posted here .

Regarding the ceramic blanket , it is good stuff , particularly in thick layers and even better if coated and sealed into a rigid structure . Prices vary widely . eBay is your best buy from what prices I have seen . The fiber I believe is aluminum silicate . There is another material called rock wool carried by some building suppliers for similar purposes as fiberglass but has twice the fire resistance . Rock Wool might be useful as a fiber reenforcement in refractory or as a backup insulation for the ceramic blanket .
Rock wool can be hard to find though because it has almost been compltely replaced by fiberglass for ordinary uses where the higher temperature rating for rock wool isn't needed .

FrankRizzo - 20-12-2004 at 15:44

Rockwool can be found cheaply at ANY online hydroponics store and even some local gardening centers. It's used as an inert wicking medium in soiless gardening.

axehandle - 20-12-2004 at 17:54


P.S. What do you guys know about kaowool (wool made out of ceramic shards I think?) I have seen pictures of furnaces where there is just folded kaowool and no castable refractory.

Ah yes, Kaowool. I recently aquired a substitute for it called Superwool ( I tried to get Kaowool, but apparently it's phased out here due to safety problems with it (silicosis etc).

It's a blanket, 610mm x 7320mm x 25mm.... very reminiscent of white rockwool. Physical properties are in PDF above. Cost me SEK1000 (~= €100).

I intend to line a large metal bucket with it for a big propane based aluminum melting furnace.

Twospoons - 20-12-2004 at 18:26

I have heard that ceramic fibre really must be coated, for safety. The fibres can break down through thermal cycling, and can be very nasty if they get into your lungs. Something called 'ITC100' seems to be a common coating. Its also supposed to improve furnace performance by re-radiating heat from the walls.


Cyrus - 20-12-2004 at 21:45

Yes, ITC-100 seems like a near monopoly on rigidizers, and rigidizers are necessary.
Satanite is another product, but I haven't been able to find much on that except that it contains crystoballite and quartz.

IIRC ITC-100 is made of zircon/zirconia, (colloidal?)silica, alumina, sodium silicate, magnesium silicate, and water, but I have no exact proportion data, just general ranges, (ie 5-15%) and have done a test on this, it's not perfect yet, but moderately ok so far.

Rosco, I like the idea of the silica gel spheres, but won't they tend to absorb a lot of water- I mean that's what they are made for. :D. This may make mixing/casting difficult or it may need to be dried/fired very slowly to let off all of the water. Also, I take it you plan on using calcium aluminate or a phosphate as a binder? Where can you get this stuff?

[Edited on 21-12-2004 by Cyrus]

axehandle - 21-12-2004 at 03:41

This is confusing.... the information I have is that rigidizers don't have to be used <i>unless</i> the fiber tends to disintegrate yielding harmful particles. The insulator I bought is supposedly normally used without any rigidizers what so ever. Seems more research is in order.

Cyrus, I found this PDF with Satanite technical data:

EDIT: There is an exact chemical analysis in there that should make it possible to mix something similar after a trip to the pottery supplier...

[Edited on 2004-12-21 by axehandle]

Rosco Bodine - 21-12-2004 at 16:33

Originally posted by Cyrus

Rosco, I like the idea of the silica gel spheres, but won't they tend to absorb a lot of water- I mean that's what they are made for. :D. This may make mixing/casting difficult or it may need to be dried/fired very slowly to let off all of the water. Also, I take it you plan on using calcium aluminate or a phosphate as a binder? Where can you get this stuff?

[Edited on 21-12-2004 by Cyrus]

With a very few exceptions , ( for example US6284688 ) , nearly all of the refractory mixtures will contain some amount of water to be baked out slowly , before the material can be gradually brought up to vitrification temperature . The more porous the aggregate the more of a problem that can be , although the rate at which the moisture is absorbed can also be slow enough that it is desirable in helping to set the binder . The order in which the aggregate and fillers and binders , liquids ect. are mixed is important to regulate the setup time and minimize the water requirement for some compositions . Some advantage may be taken of first dampening slightly the most porous materials and then coating them with another finer dry ingredient which tends to seal them before that aggregate is waterlogged .
The slower absorbtion will then help the cold setup which occurs later , and reduce the amount of physically held water to be baked out . Silica gel formed in situ can be a portion of the refractory binder material in silicate bindered refractories where water glass is the used in the "cement" portion of the composition .
The combination trisodium phosphate / waterglass binder looks more interesting because of simplicity .
I am more in favor of the phosphate binders because they are cheaper and provide better performance . The phosphate binder that is looking interesting there is the ACPH . The phosphates which go through a sol/gel scenario and chemically change through rising temperature seem most promising .
As to where to get them , I don't know . But I have on hand precursors which may allow me to synthesize the materials . For Aluminum Chloride solution I will probably try the method of US4447351 , and then treat the concentrated solution with 85% phosphoric acid to make the ACPH , or I will make the ACPH directly from aluminum amalgam and the acids .
Some ACPH related patents are GB1435988 , GB1322724 , GB1429555 , US4046581 .

But the most simple of all the refractory binders which I found in all the patents is to modify the high temperature chemistry of portland cements , using phosphates , such as described in GB1449484 . The difficulty to be overcome there is the quick setting nature of such compositions . I want a composition which has a good working life , perhaps a half hour to one hour after mixing . So I am studying the patents and looking to adapt the techniques and mixtures for my purpose .

Regarding the ceramic wool , it definitely should be sealed with some solid refractory coating . Several places I have seen it stated not to use the bare material as a hot face , but rather as an insulation layer , even as an outer covering for a cast liner , but definitely not to use the bare ceramic material to line a forge or kiln . It is dangeous because of particulates , and there is mention that the uncoated material also physically deteriorates badly over time , making its replacement more expensive than coating it to prevent that deterioration . It is also reported that fluxes attack the bare hot material like gasoline attacks styrofoam .

fiberfrax, etc.

S.C. Wack - 21-12-2004 at 17:24

Not to disagree, just thought I'd say that I've seen the untreated blanket material used on gas furnaces and as the refractory for the combustion chamber of oil burning furnaces. It holds up OK to a jet of burning #2 fuel oil. It is a little fragile to the touch - even new, though.

Rosco Bodine - 22-12-2004 at 08:40

The ceramic material will degrade after time and how fast probably depends on the intensity of the conditions , but the danger of respirable particulates carries warnings even for handling the cold , new material . Given that certain hazard is increased for the uncoated hot material , especially in contact with a steady draft in a fuel fired furnace , it would seem prudent not to use the ceramic blankets without some sealant . Some of the blanket materials may even contain a binder component which conveys flexibility and helps hold the fibers together during installation , but after the installed material is brought up to high temperature , that binder degrades and with it the physical integrity of the blanket is degraded in the absence of some additional sealant which coats and rigidizes the structure in its post fired permanent installation . To me ,
coating the ceramic fiber material with a heat setting refractory sealant makes good sense from a safety and durability standpoint , and probably also enhances the refractory properties .

Some additional patents of interest regarding aluminum phosphate binders are US3999999 , US4059669 , US3899342 The last of these three is interesting because it mentions something which I have also seen referenced in some other patents relating to phosphate binders . The addition of even a small amount of citric acid to a phosphate bindered refractory , for example one quarter of one per cent of the dry weight of the refractory composition , can result in doubling the fired strength of the virified composition . The solubility of the phosphate is enhanced by formation of an organic phosphate complex which decomposes at high temperatures and modifes the crystallization at high temperatures of the phosphate binder . My speculation about this is that a fine dispersion of graphite probably results from the thermal decomposition of the small amount of organic phosphate , and this modifes the crystallization of the phosphate binder at high temperature , and may even result in the formation of refractory carbides which also contribute to the strength of the refractory .
update: There is another patent which simplifies further the preparation of phosphate binder systems GB595768 . Earlier in this thread the term sol/gel binder system was mentioned , and there is a good description of the concept in US2522548 which explains how phosphate binders work , and how this is analogous to other binder systems which also work on the sol/gel principle .

[Edited on 23-12-2004 by Rosco Bodine]

Cyrus - 5-1-2005 at 16:32

I've been looking at the ACPH binders, and the preparation seems simple enough, but isn't boiling phosphoric acid bad for glassware? :(

If glass can't be used, what can?

neutrino - 5-1-2005 at 17:04

Some glass is eaten by hot phosphoric acid, some is not. Pyrex is to some extent (not too much, I think), Simax is said to be inert. How about steel? I know that the acid is used for passivating it, but I’m not sure how well it would actually hold it at high temperatures.

Rosco Bodine - 5-1-2005 at 20:07

Originally posted by Cyrus
I've been looking at the ACPH binders, and the preparation seems simple enough, but isn't boiling phosphoric acid bad for glassware? :(

If glass can't be used, what can?

That high of a temperature is not required for making the ACPH , all you are doing is forming an aqueous solution of the complex by adding phosphoric acid to a solution of the aluminum salt . Glass or even a stainless steel cooking pot should do fine . Think of it as making a pot of mildly corrosive ACPH "salt solution" . I'm not really sure it will even be viscous when freshly made , at the dilution it is to be used . I have an impression of the ACPH solution as sort of being the "special binder water"
you use to wet down your aggregate mix or "mud" . If the other components are
chosen correctly , it should cold set like mortar . But it will also be a sort of "greenware" from which the water will have to be first slowly dried out by evaporation at normal temperature , then slowly baked out to further dry , and brought up to firing temperature to vitrify and complete the hardening .

Twospoons - 19-1-2005 at 12:38

Don't know why I haven't posted this link before :

Castable refractories for a Soda kiln

Must be a pretty tough hot-face mix to stand up to salt glazing!

BrAiNFeVeR - 10-3-2005 at 12:10

Today I made my very first refractory mix:

-33% portland cement
-65% Perlite (I think it is perlite, it's white, doesn't melt in blue flame and it's very light)
-2% 400 mesh Al powder
(measurements are ballpark ranges, I may have used quite a bit more portland cement now that I think about it ...)

It's curing now ... hoping for beginners luck :cool:
It foamed up pretty good, like chocomousse :P air bubbles in it no bigger then 1mm diameter.

EDIT: I'm confident that it was a very bad conductor for heat, but it was as weak as chalk, even after 30 hours of curing.
It broke in several pieces as I tried to remove it from it's form :(

For this relalatively low temperature application (RBF heater), I think it's best to stick to a normal concrete mix with SiO2.
Would 50/50 portland cement to sand do?

[Edited on 12-3-2005 by BrAiNFeVeR]

Cyrus - 18-4-2005 at 18:48

Ok, this is kind of a late reply, but if by RBF you are referring to round bottomed flask, than any real refractory is overkill.

Pyrex (this is what you are using, correct?) melts/softens/oozes/because there's no real melting point at around 800 deg. C. That's barely even glowing red (at least my pyrex). Concrete contains the flux CaO, which doesn't really even become active until much higher temperatures, maybe 1200 or 1300 deg. C. That's why axehandle's concrete based furnaces don't melt. So, pretty much anything you use will work fine unless you are planning to melt the flask.

I would suggest a mixture of sodium silicate and perlite. Extremely insulating, cheap, hard at low temps, easy, and it won't melt at your low temperatures. Now, you may have some spiffy setup where the heating wire itself gets really hot (as in bright yellow glowing hot) but even then I think you may be ok.

Ok, I've "finished" my study of kaolin (china clay type stuff) based ceramics, using talc, graphite, alumina, silica, and grog as additives. (and some other stuff) I've got a paper and graphs mostly written up. They should be coming soon. It's kind of abstract, but very interesting. Oh, BromicAcid, that will be part of the paper for you.

Now I should get slip casting some crucibles.

I'm also building a new furnace with diatomaceous earth, perlite, and aluminum chlorophosphate (I'm pretty sure I'll use that binder).

BrAiNFeVeR - 20-4-2005 at 06:28

The concrete is holding up pretty good, besides the fact that it cracked.
After 3 days curing I placed the resistor wire and during the first firing I noticed the crack, but it is still very strong.

I think it just needed some flexibility ...

To prevent it from developping a second crack and thus fall apart, I wrapped it in glaswool isolation with some metal wire strapped around it to keep the glaswool in place and give the entire contraption something to hold it together.
Even the concrete against the red hot wire seems to hold up good.

Heat transfer efficiëncy is about 80% so that's pretty good too :-)
It takes some fidling with the concrete , the wire and some electronics, but I'm very happy with the results.

Rosco Bodine - 20-4-2005 at 10:00

I haven't had time to pursue my experiments with the ACPH binder
and the various refractory mixtures in which it would likely be a good performer .

I did find it mentioned in one of the ACPH related patents that it performed the best of any of the known binders which were tested for their bindering properties on particles of fullers earth type refractory materials . That statement lends support to my idea of using the calcined fullers earth porous absorbent type materials which are used as artificial soil for aquatic plants , and for maintenance of athletic fields . Some smaller percentage of the
lower fired " kitty litter " type of material which disintegrates when saturated and helps to fill the small spaces between the larger particles of the calcined material supposedly strengthens the bond of the entire mass
which is bindered with ACPH .

Also gleaned from the patents regarding portland mixtures , the white portlands that are used for tile grout and other finish work are better performers in refractories where portland is to be a
part of the mixture , than are the dark gray colored portlands .

[Edited on 21-4-2005 by Rosco Bodine]

Cyrus - 20-4-2005 at 15:27

I've found diatomaceous (fuller's) earth at the $1 store as kitty litter. It doesn't seem to fall apart at high temperatures. Is that what you were referring to as the kitty litter material which fell apart at high temps?

I found some phosphoric acid also at the dollar store... its only about 1.3 M H3PO4 from what I can tell though.

Rosco Bodine - 20-4-2005 at 16:48

DE and fullers earth are different materials . DE filter aid for pool filters is
a white high silica porous material like microscopic snowflakes made of quartz .

Fullers earth is used for kitty litter of the clay type , but this is simply dried clay and
will get pasty when saturated with water .

The calcined material has been high fired and it retains its structure when wet , and
it is refractory , to what extent I'm not sure , but it should be a good candidate
in a refractory aggregate since it is porous and has some strength of its own ,
where some other porous materials like perlite are not really structural and could weaken the composite if used in too great a proportion , especially if never fired hot enough to actually melt the perlite to leave glass lined voids where once was the perlite .

Magpie - 15-5-2006 at 20:11

Roscoe did you ever settle on a satisfactory castable refractory?

Axehandle used Portland cement as a binder in his castable refractory, and he estimates that his furnace reached about 1300C. But as Organikum has stated Portland cement is not supposed to hold up at high temperatures. Can anyone explain axehandle's success with Portland cement?

I have done some rough calculations on the power demand of a simple tube furnace, neglecting heat loss at the ends. I used this equation, taken from a heat transfer text:

P = 2(pi)kL(Ti-To)/(ln Ro/Ri)

with k = thermal conductivity of the insulation
pi = 3.14
L = length of furnace
ln = natural log
Ri = inside radius of furnace insulation
Ro = outside radius of furnace insulation

This formula is interesting to play with. What struck me was that the conductivity of the insulation is a bigger factor than the thickness of the insulation.

The power calculation results confirmed to me how important it is to get that conductivity as low as possible.

[Edited on 16-5-2006 by Magpie]

12AX7 - 15-5-2006 at 21:18

Where's Axe's post saying so?

It's certainly possible if you don't use much, or use a lot of alumina or other refractory oxide.


Magpie - 15-5-2006 at 21:33



4th post down.

garage chemist - 21-7-2007 at 12:34

I am going to need some castable (or rammable) refractory as well in the near or less near future (electric furnace! I have 100m 1mm Kanthal A-1 wire, good for 1400°C!).
I have been told that perlite is not refractory enough to be used for fire-facing refractories since it is essentially a sort of glass and melts below 1000°C.

I am thinking about using vermiculite (the stuff that is often used for packaging dangerous liquid chemicals) as the porous constituent of the refractory since Ullmann says it withstands up to 1200°C which would be adequate for my temperature requirements. The higher price and lower availability of vermiculite compared to perlite does not matter since I'm going to build one or two furnaces and not a series.

What do you think about that? Can vermiculite be used as a superior replacement for perlite?
My refractory mix would then consist of vermiculite, portland cement or fire cement and maybe bentonite (cat litter), and water.

12AX7 - 21-7-2007 at 13:12

Vermiculite is about as bad as perlite. It just stains more (due to Fe(II) content, which is a flux on par with Ca).

Don'tm even bother with PCC!


garage chemist - 21-7-2007 at 13:27

What do you mean by PCC?

What does the Fe content of the vermiculite have to do with its refractory properties? References suggest vermiculite is more heat resistant than perlite. Obviously the Fe does not act as a flux or its amount is not sufficient for this if the vermiculite is so heat resistant.
I have even read that vermiculite is used as a constituent of insulating refractory bricks.

12AX7 - 21-7-2007 at 14:17

You've never dealt with iron if you don't think it's a flux!

I tried vermiculite with fireclay and it got nice and gooey up in the yellow to yellow-white range (around 1200C).

"Insulating" and "refractory" are two words that mean different things in different contexts. Plain old fiberglass wool is an excellent insulator, and refractory enough to insulate a cooking oven. Sure as hell isn't suitable for a kiln though.

PCC = Portland Cement Concrete.


Magpie - 21-7-2007 at 18:47

GC I also have some plans for building a tube furnace. I was going to do that this summer but I am way behind schedule.

My maximum temperature was to be 1350C. What is your maximum design temperature?

I have most of the parameters worked out but have not yet purchased any materials. I had also selected 1mm Kanthal A-1 for a heating element.

For a castable refractory I had identified 3 possibilities that are available to me:

Harbison-Walker Mizzou, 1650C
Harbison-Walker Kast-O-Lite, 1428C
Thermal Ceramics Kaolite Light Weight, 1372C

Thermal conductivity for these are, respectively, 1.04, 0.41, and 0.22 W/m-K.

There should be similar products available in your part of the world. They run about $40/25 kg.

From the reading I have done it is important that any refractory touching the heating element be low in iron (<1%).

I can provide more design "rules of thumb" if you are interested.

[Edited on by Magpie]

IPN - 21-7-2007 at 19:02

Not quite on topic of refractory mixtures but for a small tube furnace you should consider buying a ceramic tube. It's easy to insulate it with glasswool and you're nearly done. Just arrange the heating elements and set up the controllers.
Of course it is a bit expensive solution but IMO it makes the construction and future maintenance a lot easier. :)

Also, garage chemist as you live in Germany you can get ceramic tubes with ease from Haldenwanger.

garage chemist - 22-7-2007 at 01:15

I already have built a super-low-cost tube furnace for up to 1100°C (it uses NiCr 80/20 as the wire). I have not mixed or purchased any castables for it, I simply built it out of "Ytong", which is a form of highly porous brick that is used to build entire houses out of due to its good insulating properties. It is sold in large blocks very cheap and can be cut easily with a woodsaw.

It is not meant to be made hot, but it does not get soft at all even at 1100°C and insulates reasonably well.
The downside is that it shrinks at such a heat and develops cracks. The tube furnace broke in two after a few firings. I put strong wire around it to hold the pieces together. It holds up fine like that now.
Although I still dont have the quartz tube for it! I cant use the tube furnace for anything at the moment. Its going to take at least two weeks from now for the quartz tube to arrive.

Here is the link to the pictures of my tube furnace:,9256,0,-Bau+eines+Rohrofe...

As you can see, the fire-facing side has been covered with a layer of fire mortar sold at the home store for repairing ovens. It was difficult to apply, and I certainly wouldnt do it again if I had to rebuild the tube furnace. The Ytong had to be drenched in water in order to not suck the water right out of the fresh mortar, and all that water had to be baked out again before the first firing which took a whole day at medium power setting.

The heating spirals at the ceiling of the furnace are held in place by a lot of hooks which are made of the same wire and glued into the Ytong with waterglass. (Dont use waterglass with Kanthal wire, it will ruin it! Kanthal must not contact any materials containing SiO2 or silicates when hot, the manufacturer warns against that. Nichrome seems to hold up fine to waterglass and SiO2 though.).
As seen in the last picture, the spirals have sagged considerably in some places which led me to use even more wire hooks for the ceiling spirals. They dont sag any more now.

IPN, I plan on making a new tube furnace for at least 1200°C using Kanthal wire wound around a ceramic carrier tube. Unfortunately, Alsint tubes are extremely expensive (over 100€). I am hoping to get one cheaper via ebay.
Haldenwanger also doesnt seem to deal with individuals.

The wire around the Alsint tube will be embedded in a mass made of MgO and Al2O3 (as Brauer recommends it) and thermally insulated by loosely poured Vermiculite in a box. So no castable required for this one either.
Read up in Brauer, they have an excellent section on building all sorts of laboratory furnaces!

I also want to build a conventional, top-loading furnace similar to axehandles and Davsters design, and for that I will need some sort of insulating refractory castable. The fire-facing side will probably be made of chamotte bricks, but those dont insulate and will have to be surrounded by castable.
Though I see that the castable wont be subjected to the full heat of the furnace, and perlite will probably do the job with that design...

[Edited on 22-7-2007 by garage chemist]

garage chemist - 23-7-2007 at 11:07

There should be similar products available in your part of the world. They run about $40/25 kg.

Unfortunately, that does not seem to be the case. The is no seller for such products in germany that sells to individuals.
Comparable products are called "Feuerleichtbeton" and are used exclusively for industrial size furnaces in the steelmaking industry and by furnace builders.
No individual builds their own high-temperature furnaces in germany, and those who do dont have castable refractories at hand and have to use replacement materials like Ytong which are not meant to be heated, or refractory bricks which however are not insulating.
Hence I have to mix my own castable refractory.

I again pose my question: Has anyone successfully used vermiculite as the insulating component of such mixes?
On the site of a manufacturer of refractory materials I have again found evidence that vermiculite is more heat-resistant that perlite and would therefore be a preferred ingredient.

garage chemist - 14-9-2007 at 17:15

Another question: I seem to be having some language problems concerning the definition of some of the raw materials. Dictionaries say that fireclay=chamotte. But the chamotte that ceramics stores carry is pre-fired, e.g. grog. In germany only the name chamotte is known.
I bought 10kg of fine chamotte (0 - 0,2mm) for use as grog in admixture with Ciment Fondu that I got from france. The Ciment Fondu says on the sack that it has to be mixed with refractory aggregates in order for the cured mortar to be refractory. So I bought the chamotte.

But what is the fireclay that is used in the first bucket furnace here:
Is that pre-fired chamotte as well? Because I have never seen unfired fireclay offered for sale anywhere.

12AX7 - 14-9-2007 at 17:21

Fireclay is unfired clay with a high vitrifying point (> cone 30). Try looking for raw clays, especially china clay (kaolin).

Chamotte = fired clay = grog.


garage chemist - 14-9-2007 at 17:40

OK, so the stuff I got is prefired.

Is fireclay supplied as a dry powder or as a moist paste like normal clay?
Do you happen to know the german term for fireclay? Dictionaries translate is as Schamotte, which obviously only means the prefired grog. I need to know the term for the unfired material.

12AX7 - 14-9-2007 at 19:28

I don't.

Fireclay can be had dry or pugged. Often the pugged material is a blend for specific pottery uses (e.g., stoneware), so do ask your supplier if it's raw.

Other terms that come to mind: "lute" (as used to seal flasks in Ye Olden Days) was sometimes a clay composition I think, and you may also encounter the term "loam" (IIRC, Bessemer used the term), which is a sand and clay type of soil.


Eclectic - 14-9-2007 at 19:53

Something like a china clay or kaolin would make a very high grade fireclay. Any clay with a fusion temp over 1600 C. You can add alumina to make it even more refractory.

You can get these clays either dry in bags, or wet from a pottery supply.

[Edited on 9-14-2007 by Eclectic]

Vermicullite and Perlite have fairly low melting points. Diatomite or Celite are supposed to be a good refractory insulation as a loose fill. If you added a bit of colloidial silica binder to slurry of diatoms and had a way to cast it while vacuum filtering the water out, you could make bricks and sheets of the stuff.

(Sorry if I'm being redundant 12AX7)

[Edited on 9-14-2007 by Eclectic]

garage chemist - 15-9-2007 at 05:14

Silica is bad news when using Kanthal wire, look it up in the Heating alloys handbook from the Kanthal website- they say embedding masses containing free SiO2 rapidly destroy the heating wire at high temperatures. Any SiO2 must be bound as silicate or compounded with lots of Al2O3.

So the embedding mass for the heating wire in my planned tube furnace can't have SiO2 in it. Perlite both contains SiO2 and only resists temperatures up to 800°C, so a castable mix containing perlite can only be used for back insulation, the mix directly surrounding the heating wire will consist of equal parts calcium aluminate cement (Ciment Fondu) and ground chamotte.

I will first build a small test furnace anyway, that uses Nichrome wire. Nichrome resists SiO2.

Eclectic - 15-9-2007 at 05:58

If you want a mix to embed the heating elements, alundum cement works well. AFAIK, it's just alumina with 1-3% phosphoric acid binder. You could use a clay binder if you don't like phosphoric acid, but after firing the acid will be bound up as aluminum phosphate.

I'm thinking of experimenting with milled zircon (Zircopax) with a colloidial alumina binder, maybe with a bit of magnesium formate so as to get a transient corderite phase while firing.

I was recommending Celite as a loose backfill insulation.

For a castable insulating refractory, you can mix in sawdust. It burns out leaving a porous structure.

[Edited on 9-15-2007 by Eclectic]

garage chemist - 15-9-2007 at 06:06

The pottery suply only had hydrated alumina, not calcined alumina. Can hydrated alumina be used as well or does it have to be calcined first?

I'll probably use the calcium aluminate cement/chamotte mix anyway, its cheaper and cures rapidly at room temperature.

Eclectic - 15-9-2007 at 06:09

Hydrated alumina is fine if what you want is to mix it with clay to make a higher melting refractory. You can mix calcium aluminate with clay if you want to have something cheaper that will set up at room temp. You could experiment with powdered aluminum, calcium hydroxide, and clay, to see if you can come up with a high temperature version of the foamed concrete building material.

Calcium Aluminate (Ciment Fondu) is good for up to 1600 C?
Higher then any temps you will get with a Nichrome or Kanthal heating element.

[Edited on 9-15-2007 by Eclectic]

angelhair - 11-6-2009 at 01:55

How about silicon carbide and Al oxide? Will Nichrome and kanthal do well in that? I've seem a formula for one that contains;

25.4% Al2O3
9.3% SiO2
0.4% FeO3
3.5% CaO
59.8% SiC
0.3% Alkalines

The thing I like about this is it' s thermal conductivity of 8.1W/mK where as the high aluminates are around 2 and below. I'm assuming this a good thing if using it as a mantle.

Vegemeister - 4-7-2009 at 00:40

A few ideas about refractories:

1. Considering the amount of material needed for a furnace, you should try to minimize the use of things that you can't get in 50 lb. sacks at a reasonable price.

2. If you can improvise the facilities, pre-fired bricks are much easier to work with than castable or rammed mixes. Due to their surface area, they can be dried much more evenly and more quickly. Thoroughly fired bricks are more structurally sound than fired-in-place mixes, and more importantly, most of the shrinkage is done with, so unpleasant cracking is avoided.

3. For insulating refractory, you want some organic filler to burn out. Although perlite will work, its low melting temperature will severely limit the service temperature of the refractory. Sawdust is probably cheapest, if you can find a source (I have not yet succeeded.), but if you don't mind a headache you can shred the polystyrene foam blocks used for packaging. The shavings weigh almost nothing, so they easily become airborne, and static effects cause them to stick to everything. I suggest shredding with a motorized wire brush and collecting the results in a freshly emptied shop-vac.

4. Unless you have calcium aluminate cement or something of that sort, the best thing for your fired refractory to turn into is 3:2 mullite, 3(Al2O3)2(SiO2). Here's an alumina-silica phase diagram from Tim's site. Straight Al2O3 has a higher melting point, but it tends to lose more strength as it approaches its melting point. It is best to err on the side of alumina, however, because free silica goes through phase changes with changes in temperature, especially if it soaks for a long time, and the associated volume changes fuck things up.

While it is possible to make mullite by firing finely milled alumina and silica in the appropriate ratio, Al2O3 is very expensive. It is best to start with one of the high pressure stable aluminosilicate minerals (kyanite, sillimanite, and andalusite) and add alumina to make up the difference. The aluminosilicate minerals are all unstable at atmospheric pressure, so they decompose to mullite and silica at high temperature. Because kyanite is the highest pressure form, it decomposes at the lowest temperature (around 1350 C ish). It also undergoes a volume expansion, but finer mesh kyanite changes its dimensions less because the expansion instead goes to reducing porosity. Here's a PDF put out by Virginia Kyanite that explains more:

If, like me, your local pottery supply shop doesn't stock kyanite, you can use a cheap kaolin clay such as Edgar Plastic Kaolin, but more alumina will be necessary in that case.

5. I have recently discovered that the recycled cellulose fibers sold as building insulation carbonize but do not burn in the flame of a propane torch, and show incredible insulation performance. The side of the wad of fiber directly heated by the torch was carbonized, while the other side remained unchanged. Considering the very low cost of this insulation, it may work well as a sacrificial loose fill around a firebrick inner wall. If the loose fill space is well sealed by a casing and perhaps high temperature silicone sealant, the fiber should combust just enough to maintain a neutral/reducing atmosphere in the space. It probably wouldn't work well as organic filler for firebrick, however, as what I've read suggests that it is treated with boric acid to provide fire resistance, and this would probably flux the refractory.

6. I've run across references suggesting that complex carbon shapes may be converted to silicon carbide by heating to extremely high temperature in a bed of silica (sand or fumed?) and loose carbon. Both SiC and carbon are very effective microwave susceptors, so I think it might be possible to form silicon carbide structures (crucibles, burner tubes, resistive heating elements) in a domestic microwave oven from carbon objects formed from powdered charcoal held together with an organic binder (wheat flour?)

Pourable Alumina

shaft - 17-5-2017 at 15:57

I want to make my own Alumina tube, 600mm long, 40mm OD, 32mm ID with a 4mm wall. I made a polyurethane mold and bought some Rescore Al2O3 from Cotronics in a mad shopping rage.

It is frightfully expensive for me. What are the chances of making my own mix than I can pour or somehow ram down a 4mm gap?

I saw someone who combined Calcined Alumina with Veegum (Kaolin) but how do you make that pourable?
made a kiln shelf from 96.5% calcined alumina and 3.5% Veegum

In this patent, someone combines an aqueous phosphate and silica sol as a binder but it doesn't look pourable.

Refractory binder
EP 0824092 A1

A mixture of 1 volume Clinochem P8 at 50% wt polyammonium phosphate (36% "P205") is mixed with 4 volumes of 40% wt silica sol, the mixture is then mixed with tabular alumina grains and powders such as are readily known to users and manufacturers of ceramics at between 60 ccs per kilogram to 80 ccs per kilo, (depending upon the application) and then placed in a mould. At normal working temperatures, 15° - 21°C, gelation occurs in approximately 30 minutes and can be removed within one hour. No accelerator in the form of a finely divided grain is used in this example; it appears that the silica sol causes the phosphate to gel as well as vice versa.

So what are my chances of making something pourable? How does cotronics do it?

Chemetix - 18-5-2017 at 03:32

I've found myself needing to make a castable refractory for a large oven, it's only going to a max of 1000C but settling on a mix has been confusing to say the least.
Everyone has a different recipe or a variation on a recipe. And when there is a comment steering someone away from using a certain compound or product there's just as many declaring it's vital in their formula. Sodium silicate seems to be a staple, and yet there's so many that will swear you off it because it ages badly or is incompatible with other chemistries.??

So in the end I decided to just use a formula given to me by a refractory company. 4 parts vermiculite, 1 part ciment fondu, 1 part fireclay.
They say add 2 parts water, but it's not nearly enough I found, I needed at least 3 parts and some more to get it to flow into all the parts of the mold and not trap air. And it set so quickly that I barely had time to empty the wheelbarrow with a shovel before it was unable to trowel smooth. A spray bottle with water helped get the top smoothed over. Even when given a decent amount of water it's still a very plastic mix and heavy to work by hand. A few good bangs with a rubber mallet drove off the larger trapped inclusions of air and made the edges crisp and well formed. Adding a bit more water makes it perfectly pourable, but work quickly.

I used linseed oil as a mold release and anything coated in it repelled the concrete mix perfectly well. I suggest getting a bit on your hands before getting any of the cement on them. The Fondu is hellishly drying on the skin. I'd suggest you use a respirator and coat the paper filter with oil too. Nuthin's gettin through that!

The volume reduced by a little more than half when you add water. The original suggestion of 0.6 it's dry volume was quite a bit off.
After setting for a day the mold was opened and the block has a good hollow thunk sound to it even though it's really quite hard. It's 20l by volume but probably closer to 13 to 15 kg even without the loss of water from drying. It should provide the necessary structure and insulation without the specific heat capacity of a sand filled mix.

Elemental Phosphorus - 18-5-2017 at 04:00

There is another forum online dedicated to furnace and foundry construction. I read it sometimes, since I built a metal-melting furnace, they helped me find a brick-and-mortar castable refractory source near me. The topic of refractory compositions, and where to acquire refractory material has been covered in almost mind-numbing detail there. I suggest any member interested in furnace design read these two webpages:

shaft - 18-5-2017 at 05:50

Thanks for the link. I will definitely have a look.

In the mean time, I found this;

Keratab Fine Special
Description; A self flowing self de-airing, fine tabular alumina pourable hydraulic setting ceramic.

Applications; Pipes, thin walled tubes, lead-outs, crucibles funnels, complex shapes, burner quarls.

Maximum service temperature; 1800°C

Maximum particle size; 0,5mm

Density; 2,85g/cm³

Air dry;          35MPa
Fired 1200°C; 80MPa

Mixing; Add +9% water, mix for 2 minutes and pour. Pot life 12 minutes. Chemical analysis;

Al2O3   97%
CaO     3%
SiO2     0.01%
B2O3    0.008%
Fe2O3   0.01%
Na2O    0.25%

Seeing that I don't know much, I have to ask, Does this need an added binder like Phosporic acid or is Calcium Oxide at 3% the binder? Surely there is something not listed here.

shaft - 16-6-2017 at 07:18

I'm pretty sure the missing item is a dry phosphate salt.

People in this thread have mentioned keeping the "free silica" content low to avoid damage to the element.

I'm still not clear on what "free silica" actually means. I am looking to follow this patent;

where the binder consists of a wet or dry phosphate and silica sol or silica fume.


A mixture of 1 volume 50% wt polyammonium phosphate is mixed with 4 volumes of 40% wt silica sol, the mixture is then mixed with tabular alumina grains and powders such as are readily known to users and manufacturers of ceramics at between 60 mls per kilogram to 80 mls per kilo and then placed in a mold.

At normal working temperatures, 150 degree - 21 degree C, gelation occurs in approximately 30 minutes and can be removed within one hour. No accelerator in the form of a finely divided grain is used in this example; it appears that the silica sol causes the phosphate to gel as well as vice versa.

This doesn't seem like free silica to me. So what kind of ingredients would result in a "free Silica"?

yobbo II - 22-6-2017 at 07:04

Quote: Originally posted by shaft  
Thanks for the link. I will definitely have a look.

In the mean time, I found this;

Keratab Fine Special
Description; A self flowing self de-airing, fine tabular alumina pourable hydraulic setting ceramic.

Applications; Pipes, thin walled tubes, lead-outs, crucibles funnels, complex shapes, burner quarls.

Maximum service temperature; 1800°C

Maximum particle size; 0,5mm

Density; 2,85g/cm³

Air dry;          35MPa
Fired 1200°C; 80MPa

Mixing; Add +9% water, mix for 2 minutes and pour. Pot life 12 minutes. Chemical analysis;

Al2O3   97%
CaO     3%
SiO2     0.01%
B2O3    0.008%
Fe2O3   0.01%
Na2O    0.25%

Seeing that I don't know much, I have to ask, Does this need an added binder like Phosporic acid or is Calcium Oxide at 3% the binder? Surely there is something not listed here.

The lower % ingredients are not actually 'ingredients' at all but rather impurities that have to be tolerated IMO.

High temp. bricks are very expensive because the ingredients have to be very pure and there are not so many produced like the more normal (1350C) ones.

Somewhat contradictory to this is the fact that you can purchase very high temperature refractory cement for a reasonably price.


shaft - 10-7-2017 at 23:16

In my country, Australia, things are hard to get. We don't make things anymore.

I did find a castable mix called Shiraflo 90 SB made by Shinagawa, but it's still only 88% Alumina with a lot of silica rocks. I filtered those out and managed to cast a 600mm long tube with a 4mm wall.

I had to use 10% water to make it flowable enough to seep into the gap. They recommend a max of 6.5% water.

I have been trying to make a castable using Calcined Alumina 300 mesh from the pottery store and various phosphate binders as outlined in these patents;

Also this where they make Aluminium Phosphate in situ;

Here in one example, they use 9.6% Phosphoric acid solution @48% wt concentration. I couldn't do that with Calcined Alumina, it was still a dry powder. I had to use about 60-70%. All results were failures.

I believe I worked out why it doesn't work. I didn't know that Calcined Alumina had such a massive surface area. So now, I want to use Tabular Alumina.

Where the hell do you find a 25kg bag of that? Not in my country. Not that I can see.

yobbo II - 11-7-2017 at 15:37

Quote: Originally posted by shaft  
In my country, Australia, things are hard
I believe I worked out why it doesn't work. I didn't know that Calcined Alumina had such a massive surface area. So now, I want to use Tabular Alumina.

Where the hell do you find a 25kg bag of that? Not in my country. Not that I can see.

I would think that calcined anything will have a smaller surface area than the stuff that is not calcined?
Perhaps I am not correct.

I suppose that links to stuf is of no use as you want to actually make your own.


[Edited on 11-7-2017 by yobbo II]