search  current discussion  categories  glazes - chemistry 

glaze coefficients of expansion (fwd)

updated wed 11 feb 98


Tom Buck on tue 10 feb 98

Sorry, I didn't cc this to Clayart. So here it is now.

Tom Buck ) tel: 905-389-2339
& snailmail: 373 East 43rd St. Hamilton ON L8T 3E1 Canada
(westend Lake Ontario, province of Ontario, Canada).

---------- Forwarded message ----------
Date: Sun, 8 Feb 1998 23:34:01 -0500 (EST)
From: Tom Buck
To: Bob Cyr
Subject: Re: Glaze Coefficients of Expansion

Bob Cyr: If there is general agreement on linear COEs, this agreement is
not reflected in the texts that are cited as reference works for potters.
FYI, the units of the coefficient of thermal expansion (linear)
are delta L/L/deg C, that is, the change in linear length of a standard
length per degree Celsius. The usual units used in technical literature
are: micrometres change per metre per degree Celsius, and this is usually
cited as a specific value, namely, eg, Ferro Frit 3110 is said to have a
COE of 0.000 0101 metres/metre/deg C (or 10.1x10-6).
Generally, workers in the field of thermal expansion of solids
have adopted the notion that individual metal oxides exhibit a fairly
constant COE value throughout the usualy firing ranges that potters use.
This notion is a rather crude approximation but it is the best available
(as far as I can find in the literature, others may know better). And,
further, that when several metal oxides combine to form a glaze, that the
values of each oxide are added toegther to yield the final COE of the
glaze melt itself.
So, in reality, the COEs one often finds cited with recipes are
only indicative (a guide) of glaze contraction/expansion and one should
not expect the actual value cited (sometimes to three significant figures)
to prove right in practice. If a glaze calc program reports the glaze's
COE as 7.3 x 10-6, there is every chance the "real" value is somewhere
between 5 and 8 x 10-6, although Ron Roy's tests with a dilatometer on
prefired clay rods (unglazed and glazed) came up with a claybody (C6 white
stoneware) value of 3.3 x 10-6 and the glazed rod was slightly less than
Which all goes to show that only by firing a test pot can one have
some confidence that a particular glaze mix will properly fit a certain
claybody. The glaze calc program is meant to get you to the ballpark so
you can run a few tests and find a good glaze for your pots. Otherwise
you may end up like a local potter who ran 100 separate tests to find a
good copper red; one or two looked promising but there was no solid
winner. The potter had fun though, and mounted the test tiles on a large
wooden panel for all to see.
Til later.

Tom Buck ) tel: 905-389-2339 & snailmail: 373 East
43rd St. Hamilton ON L8T 3E1 Canada (westend Lake Ontario, province of
Ontario, Canada).