[Stoves] Clay Stoves - Columbia University

[Crispin Pemberton-Pigott]

Dear Tae

You ask an interesting question to which there are numerous answers, none of
which seem to be perfect.

The highest quality material is boron silicate glass which can makes stoves
with zero thermal expansion and nearly unlimited thermal shock in the
temperatures which stoves run at.  I have not yet see a small stove made
from this material however it is only a matter time before someone makes
one.

There is so little known about the vast, almost a universe of diversity in
the formulation of natural clays, that one wonders why more names nave not
been given to them. They can be characterized as kaolinitic and high-mica
and so on.

Something recently discovered by Bruce Berger in Johannesburg is that the
availability of free silicon in the clay is a very important indicator of a
propensity to crack.

Silicon changes from a Phase A to phase B with a 2% larger volume as it
passes through 573 degrees.  That is why glass has to be cooled very slowly
through that temperature so that the entire object changes size at the same
time, or it will fracture.

Free silicon in clay does the same thing. If it is taken through 573
degrees, up or down, there is a size change in the silicon, which can induce
fractures which can joint together after a few more thermal cycles and the
component fails.

JIKO Stoves, which have a ceramic liner with a surrounding metal shell are
made so that when (not if) the ceramic fails, the metal will hole it
together. The metal also prevents a certain amount of mechanical shock. The
reason they are not made entirely of clay is because it simply would not
last.  On the global scene, 1 year is considered very good lifetime for a
clay stove.

That said, it depends on the amount of thermal shock the clay receives.  It
is it a charcoal stove lit with kerosene it is likely to be faced with
sudden changes in physical size on the inside of the stove. The inside
expands, cracking the outside surface, if the compressive strength is much
higher than the tensile strength (which is usually is). Making the clay
porous can weaken the structure, making it 'spongier' and this can reduce
the fracturing for two reasons: it has a lower compressive strength and it
is slightly stretchy in tension.  The problem is that getting it into that
conditions usually means not having something strong enough to make a stove
from.

The Baldosa tile made traditionally in Central America is an interesting
exception to this general approach. It has nearly no free silicon in the mix
- quite how is not clear, but I had it tested by Bruce recently.  The
expansion coefficient is quite high, more than 5 x 10^-6, but it exhibits
very little evidence, if any, of free silicon changing size at 573 degrees.
It is important for researchers to find out why.

The Baldosa tile is known for being able to take quite high temperatures as
a flat tile, I think provided that it is given room to expand on one side
and not on the other, which would make it curve a bit.  It is a little big
spongy, quite strong, but quite expansive.  Normally this would be death to
a stove part, but the lacy of a sudden change in size on one side as the
surface goes through 573 degrees points to a possible way to avoid the
development of low thermal expansion materials.

The clay stoves fail mostly because of thermal stress from differential
expansion.  The is expectable because you have a material that is quite
rigid, brittle and changing dimension on the hot side faster than the cold
side.

Making a ceramic 'insulative' (reduce the heat conduction coefficient)
invites additional stress because there is a high temperature difference
between the two sides, however the pores introduced to reduce heat
conduction often make it spongy enough to partially of fully counteract the
increased stress.

Some natural clays have a thermal expansion coefficient of 3.0 x 10^-6 or
less however even these, unless fired to a low temperature so they do not
form a lot of ceramic bonds, will fail quite quickly when made into a
bucket-shaped vessel.  Baldosa tiles, formed into a ring, are likely to have
a short life, unless they are heated very slowly.

Another possibility is to increase the heat transfer coefficient so the
entire stove heats evenly or nearly so.  This loses heat so the stove should
have a heat-recycling design where the heat is picked up and taken back to
the fire.  However it reduces the physical stress by making the two sides
more or less the  same temperature.

The features of a clay that will make a stove are:

Low thermal expansion coefficient
Low free silicon content
Low porosity including quasi-porosity
High tensile strength relative to the compressive strength
High heat conduction coefficient

There are many 'refractory' ceramics that have the characteristic of having
a low heat conduction coefficient (for building kilns etc) but these are
usually very soft and can even be cut. Others have high 'temperature
resisting' properties but conduct heat very well.  Others can take massive
thermal shock, but tend to be very low thermal expansion mixes.

The problem is that most of these are fired at very high temperatures,
usually far above their 'working temperature'.  That proves to be impossible
to produce in a low tech environment.

I am very interested to see what you discover in our work.

There is a very interesting paper showing that the clay:non-clay ratio is
very important to finding a mix of clays that will survive well. One the
members of this list sent it to me so it might pop up again.

Best regards
Crispin