search  current discussion  categories  glazes - misc 

glaze fit practicalities

updated sat 13 dec 97

 

Owen Rye on sat 29 nov 97

Questions and answers about glaze fit appear often on Clayart. As
someone who took a first science degree I admire the work of Tony
Hanson and others who use science well - except in one respect.
Careful calculation require accurate analyses, which are mostly not
available; so beneath all the figures the results are more or less
empirical and we still need to choose the 'best' result by eye. Ditto
for thermal expansion values; anyone, like one of my MA students who
has studied this area closely realises that there is a level of
approximation involved because of the range of ways in which thermal
expansion is determined, and that there is no one scale for all of
our materials.

On to another problem. Glaze recipes can be highly misleading if you
only get the recipe, maturing temperature and a one-word 'reduced' or
'oxidised'. Leaves out most of the firing schedule and the most
important factor if you are worried about glaze fit: the body on
which it works. Puzzling about this years ago, I realised that much
of the research which led to the wisdom about adding or subtracting
silica to vary glaze fit was based on original research into English
white earthenware fired at relatively low temperatures, and did not
necessarily apply to stoneware or porcelain. Especially since the
silica variation worked best in the body rather than the glaze (a big
change of silica could totally change the characteristics of a
glaze), and most people could not change this since they bought
bodies made up and ready to go.

Solution. To use stoneware glazes on porcelain, or vice versa, change
the amount of felspar in the glaze. If using a stoneware glaze on
porcelain, take out about 10% of the potash felspar (ie if the
original called for 40% felspar, use 30 instead). If using a
porcelain glaze on stoneware, add about 10% felspar. Reason? The
porcelain body has much more felspar which reacts with the glaze, so
porcelain glazes need less. The stoneware body has less felspar, and
so the glaze needs more. The higher the temperature the more this
applies; by 1500C the amount of felspar in a porcelain glaze and body
can be identical.

Implication: if you have a porcelain/stoneware glaze fit problem,
which you may have if you use recipes from books etc without full
information, do a line blend on your glaze varying the amount of
felspar up and down about 10% each way. At one end of the line blend
the glaze fit will be better. Sometimes you may need to go further
but not often.

No felspar in your glaze? Shucks, shuffle, shuffle - try asking
someone who knows? How many stoneware or porcelain glazes do not have
felspar?

Owen Rye (with aplogies for the length of this message).
Owen.Rye@Arts.Monash.Edu.Au

Tony Hansen on thu 11 dec 97

> Careful calculation require accurate analyses, which are mostly not
> available; so beneath all the figures the results are more or less
> empirical and we still need to choose the 'best' result by eye.

Not so. Although those white powders look the same they are
dramatically different chemically. Calculations and analyses
explain those differences so well. Lithium carbonate and
gerstley borate both are powerful melters but completely
different in so many ways and calculation helps us grasp and
exploit these differences.

Modern materials and frits are fantastically consistent
compared to bygone days. The analyses differences
from batch to batch in a material are minscule compared to
the profound differences between materials.
Even for materials that are variable calculation is
invaluable in adjusting glaze mixes to insulate them from
variations in a constituent material.

Material analyses are easy to find on the web and are
available from all material wholesalers. Every truckload of
custer feldspar, for example, comes with an analysis report.
Even if you can't get a specific batch analysis generic ones
still beat the heck out of a trial-and-error materials level
approach.

I can't imagine not being able to differentiate all the
available frits if it were not for ceramic calculations.

> Ditto for thermal expansion values

Again, thermal expansion comparisons between different oxides
are significant and even if there are errors the calculated
results still generally point in the right direction. The main ones
like SiO2, Al2O3, Na2O, K2O, CaO are fairly accurate and
they usually make up 98% of a glaze.

> Glaze recipes can be highly misleading if you
> only get the recipe, maturing temperature and a one-word 'reduced' or
> 'oxidised'. Leaves out most of the firing schedule and the most
> important factor if you are worried about glaze fit: the body on
> which it works.

Naked glaze recipes are one of the biggest reasons for value
of ceramic calculations. It is much easier to predict the
kiln performance of a formula than a recipe.

> The research which led to the wisdom about adding or subtracting
> silica to vary glaze fit was based on original research into English
> white earthenware fired at relatively low temperatures, and did not
> necessarily apply to stoneware or porcelain.

Stoneware glazes can usually tolerate additions of silica
much better because they have twice as much.

> (a big
> change of silica could totally change the characteristics of a
> glaze)

Exactly. But with ceramic calculations you do not need to do
this bull-in-a-china-shop approach to glaze adjustment.
Knowing the approximate relative expansion of all the oxides
offers many strategies for reducing a glaze's expansion and
each has well understood implications.

> Solution. To use stoneware glazes on porcelain, or vice versa, change
> the amount of felspar in the glaze.

It is the very high expansion sodium and potassium in the
feldspar that needs adjustment. Changing feldspar also
changes the alumina and silica it contributes. Viewing the
recipe as a formula via calculations lets you substitute
lower expansion fluxes for the sodium and potassium only.


> porcelain glaze on stoneware, add about 10% felspar. Reason? The
> porcelain body has much more felspar which reacts with the glaze, so
> porcelain glazes need less.

Not really. Porcelains have a lower thermal expansion so
glazes need to have a lower expansion to fit. That means
minimizing sodium and potassium in favor of CaO, MgO, Li2O,
etc. This is the significant side effect of reducing feldspar
and it is the reason it sometimes works.

> Implication: if you have a porcelain/stoneware glaze fit problem,
> which you may have if you use recipes from books etc without full
> information, do a line blend on your glaze varying the amount of
> felspar up and down about 10% each way. At one end of the line blend
> the glaze fit will be better. Sometimes you may need to go further
> but not often.

Not many glazes can take big changes in the feldspar without
significant changes in surface character, melting
temperature, etc. There is no doubt in my mind that the
best way to adjust thermal expansion is by formula rather
than recipe adjustment.

The material blending approach requires something extra to
make it work. I recommend prayer.

--
-------
T o n y H a n s e n thansen@digitalfire.com
Get INSIGHT, Magic of Fire at http://digitalfire.com

Owen Rye on fri 12 dec 97


> ----------------------------Original message----------------------------
Rye: Careful calculation require accurate analyses, which are mostly
not available; so beneath all the figures the results are more or
less empirical and we still need to choose the 'best' result by eye.
>
Hansen : Not so. Although those white powders look the same they are
> dramatically different chemically. Calculations and analyses
> explain those differences so well. Lithium carbonate and
> gerstley borate both are powerful melters but completely
> different in so many ways and calculation helps us grasp and
> exploit these differences.

Rye replies: sorry, but you missed the point which was no matter how
the glaze is formulated and fired the result is evaluated as an
aesthetic judgement primarily ie which is the best colour, the best
'quality '- we work for aesthetic reasons basically, otherwise we
would all be using plastic cups.Even if glaze fit is part of that
evaluation most everyone estimates it rather crudely (not a lot of
autoclaves or vacuum chambers out there in potteryland).

Hansen said: Modern materials and frits are fantastically consistent
> compared to bygone days. The analyses differences
> from batch to batch in a material are minscule compared to
> the profound differences between materials.
> Even for materials that are variable calculation is
> invaluable in adjusting glaze mixes to insulate them from
> variations in a constituent material.

Rye replies: OK, but you make two assumptions. The first is that all
our materials come from a supplier with an analysis (what would you
do for example with woodash - I have a series of lineblends
completed and fired before I would even find a 'typical' analysis
which applies only to a specific part of a specific tree grown in a
specific soil). I do not disagree with the value of calculations
and would certainly recommend to any beginner in pottery that this is
the quickest way to learn, and to any production potter who must have
good glaze fit, absence of toxicity and reliable repeatibility of
results to go with your approach. What about us old-time woodfirers
who deliberately change things every firing searching for something
we've never seen before, and where glaze composition is a minor,
almost irrelevant consideration in the face of deliberately wildly
variable kiln handling?

Hansen: Material analyses are easy to find on the web and are
> available from all material wholesalers. Every truckload of
> custer feldspar, for example, comes with an analysis report.
> Even if you can't get a specific batch analysis generic ones
> still beat the heck out of a trial-and-error materials level
> approach.

Rye replies: depends entirely on your years of experience with 'trial
and error' and your approach to it. Trial and error can be dumb or
smart; I subscribe to half of it, the trial half; don't think much of
the concept of error, preferring to think of all results as
contributing to learning. I will make the same deal with you as I
make with students doing glaze testing: give me your worst result (in
your judgement) and I will given a little time show you where it can
be used effectively. It will probably not be on a coffee mug; amazing
what can be used in sculpture.......Sometimes it can be forgotten
that we are all in this for the art, which is no simple thing.

Hansen: I can't imagine not being able to differentiate all the
> available frits if it were not for ceramic calculations.
> > Ditto for thermal expansion values
> Again, thermal expansion comparisons between different oxides
> are significant and even if there are errors the calculated
> results still generally point in the right direction. The main ones
> like SiO2, Al2O3, Na2O, K2O, CaO are fairly accurate and
> they usually make up 98% of a glaze.
Rye reply: agree.....my original comments about felspar content in
stoneware and porcelain glazes equally points in the right direction.

Rye: Glaze recipes can be highly misleading if you
> > only get the recipe, maturing temperature and a one-word 'reduced' or
> > 'oxidised'. Leaves out most of the firing schedule and the most
> > important factor if you are worried about glaze fit: the body on
> > which it works.
>
> Naked glaze recipes are one of the biggest reasons for value
> of ceramic calculations. It is much easier to predict the
> kiln performance of a formula than a recipe.

Rye reply: Perhaps you could predict the results of a specific
formula fired in my anagama; damned shame really, that would take
away all elements on which my work is based but I feel safe. In what
you say you are correct; I was adding some further variables which
you must agree need consideration.

Rye: The research which led to the wisdom about adding or
subtracting
> > silica to vary glaze fit was based on original research into English
> > white earthenware fired at relatively low temperatures, and did not
> > necessarily apply to stoneware or porcelain.

Hansen: Stoneware glazes can usually tolerate additions of silica
> much better because they have twice as much.

Rye reply: Seems like a big generalisation (the part about
stoneware glazes tolerating additions of silica, that is. Some do,
some don't). Part of my original point is that the English
eartheware bodies contained high silica which allowed flexibility in
glaze body composition much more so than in stoneware bodies.

Rye: a big
> > change of silica could totally change the characteristics of a
> > glaze)

Hansen: Exactly. But with ceramic calculations you do not need to do
> this bull-in-a-china-shop approach to glaze adjustment.
> Knowing the approximate relative expansion of all the oxides
> offers many strategies for reducing a glaze's expansion and
> each has well understood implications.

Rye reply : why does the bull-in-a china shop comment hurt my
feelings a little? I should be old enough to handle it.......Guess
if we're getting personal I could observe that most of the Hansen
arguments are based on the need to sell calculation programs, whereas
mine are based on the need to follow my curiousity as an artist.

Rye: Solution. To use stoneware glazes on porcelain, or vice versa,
change the amount of felspar in the glaze.

Hansen: It is the very high expansion sodium and potassium in the
> feldspar that needs adjustment. Changing feldspar also
> changes the alumina and silica it contributes. Viewing the
> recipe as a formula via calculations lets you substitute
> lower expansion fluxes for the sodium and potassium only.

Rye reply: I hold my line, which is based on the relative felspar
contents of porcelain and stoneware bodies. Anyone else reading this,
try my suggestions before you discard them; they make more sense than
the reply suggests, because they consider the felspar content of both
bodies and glazes. If you know what you are doing the time factor in
both approaches is similar, but in my approach you get to see the
trends in real life and learn from them....

Rye: porcelain glaze on stoneware, add about 10% felspar. Reason?
The
> > porcelain body has much more felspar which reacts with the glaze, so
> > porcelain glazes need less.
>
Hansen: Not really. Porcelains have a lower thermal expansion so
> glazes need to have a lower expansion to fit. That means
> minimizing sodium and potassium in favor of CaO, MgO, Li2O,
> etc. This is the significant side effect of reducing feldspar
> and it is the reason it sometimes works.

Rye reply: Both methods achieve the same result: a glaze and body
that work together. I cannot believe that you are discounting
glaze-body reactions, especially in high-felspar porcelains by your
'not really'? Especially at higher temperatures - 1450C and
above...but also at 1300C
>
Rye: Implication: if you have a porcelain/stoneware glaze fit
problem,
> > which you may have if you use recipes from books etc without full
> > information, do a line blend on your glaze varying the amount of
> > felspar up and down about 10% each way. At one end of the line blend
> > the glaze fit will be better. Sometimes you may need to go further
> > but not often.

Hansen: Not many glazes can take big changes in the feldspar without
> significant changes in surface character, melting
> temperature, etc. There is no doubt in my mind that the
> best way to adjust thermal expansion is by formula rather
> than recipe adjustment.

Rye reply: You overlook the fact that I was talking about using
stoneware glazes on porcelain, or porcelain glazes on stoneware, not
just looking at changing a glaze component in isolation. Be
interesting for you to do your stuff, me to do mine, and see how
close they are......in the end it would probably come down to the
method of assessing glaze fit (by eye, autoclave etc).
>
Hansen: The material blending approach requires something extra to
> make it work. I recommend prayer.

Rye reply: got nothing to pray to. I rely mainly on experience
(having started 35 years ago with the methods you advocate before
gradually moving to the approach I use now) but over the years have
come more to rely on very careful observation. Having seen the
results of an (estimated) couple of million student glaze tests,
along with some more of my own, helps. It may help to point out that
in my work I use very few glaze compositions, and that the
complexities of variations in firing the anagama kiln are much more
interesting. Anyone got any calculations for them? There are usually
30 different glaze results on each piece of work; all rising from the
same glaze composition applied overall. Much more stimulating....

Owen Rye (sweating through an Australian summer).......
Owen.Rye@Arts.Monash.Edu.Au