search  current discussion  categories  glazes - misc 

how accurate do glaze measurements need to be?

updated tue 27 oct 09

 

Paul Lewing on fri 23 oct 09


On Oct 23, 2009, at 5:15 PM, Bob Johnson wrote:

I know that precision--to some degree--depends on the type of
ingredient,
with cobalt seemingly more sensitive to measurements than, say,
feldspar.
But is there any rule? Can I round off all my glaze recipes to the
nearest
whole numbers?

Almost always, yes. As you say, the colorants are the most sensitive,
but even then most of them don't matter that much. Even then it would
only be the more pastel versions of the color where you'd notice.
There will be a difference between 0.25% and 0.50% cobalt carbonate,
but between 2.25% and 2.50% you can't tell the difference.
With the base glaze chemicals, rounding off to the nearest whole
number won't make any difference. And even a whole number either way
usually won't be noticeable. Of course, if a recipe calls for 2% of
something and you put in 4, it will make more of a difference than if
it calls for 22 and you put in 24.
You can often tell how a potter originally came up with a glaze
recipe. If it's in whole numbers, chances are that potter originated
that recipe by repeated tests with the materials only. If it's
carried out to two decimal places, chances are calculation software
was involved. As you have figured out, you can't weigh out that
precisely anyway.
Paul Lewing
www.paullewingtile.com
www.paullewingart.com

Bob Johnson on fri 23 oct 09


Apologies if this is an old horse being beaten anew--but as I was mixing
glazes today I began to wonder how accurate glaze recipes need to be. My
gaze program defaults to two decimal places, and that seems like a level of
precision that wouldn't be noticeable in a glaze.

Lots of things can throw the measurements off by way more than 1/100th of a
graham. E.g., chunks that get stuck in the sieve or not getting the glaze
materials centered on the scale.

I know that precision--to some degree--depends on the type of ingredient,
with cobalt seemingly more sensitive to measurements than, say, feldspar.
But is there any rule? Can I round off all my glaze recipes to the nearest
whole numbers?

Thanks!

Bob Johnson
Southern Oregon

John Hesselberth on sat 24 oct 09


On Oct 23, 2009, at 8:15 PM, Bob Johnson wrote:

> Can I round off all my glaze recipes to the nearest
> whole numbers?

Hi Bob,

I wouldn't do it that way. It depends, of course, on how big your
batch is. A better way might be to maintain 2 significant figures--3
at the most. So 1.5 grams of cobalt would remain 1.5, but 2367.8 grams
of feldspar could certainly be rounded to 2370 and possibly 2400.

Most of these recipes that show things like 34.56 grams of something
got that way because we use dumb computers to reset the recipe to a
total of 100. It certainly was not because the glaze developer
measured them that accurately.

And when a person likes me writes a program like GlazeMaster we have
to be a bit arbitrary on how many figures we display trying to
accommodate a wide range of people's work practices.

Regards,

John

Lee Love on sat 24 oct 09


Bob,

The importance of decimal places depends upon your batch
size. They are important if you are doing 100 gram tests and not so
important if you are filling a 25 gallon glaze bucket.



--
Lee, a Mashiko potter in Minneapolis
http://mashikopots.blogspot.com/

"Ta tIr na n-=3DF3g ar chul an tI=3D97tIr dlainn trina ch=3DE9ile"=3D97tha=
t is, "T=3D
he
land of eternal youth is behind the house, a beautiful land fluent
within itself." -- John O'Donohue

Snail Scott on sat 24 oct 09


On Oct 23, 2009, at 7:15 PM, Bob Johnson wrote:
> Apologies if this is an old horse being beaten anew--but as I was
> mixing
> glazes today I began to wonder how accurate glaze recipes need to be..


Most glazes are written to 100th of a percent
because someone developed the recipe in
increments, saying to themselves, "Well, that's
a good start, but next time I'll add 2 grams more
of 'this', then reduce 'that' by 3 grams..." and so
on. Then, when they get a mix they like, they
normalize it to state it as percentages. Often, if
you fiddle the math, you find that percentages
which end in weirdly-exact numbers to the
second decimal place actually turn out to be
nice round integers for a batch that wasn't
designed to add to 100. It was designed to
work well as proportions, then turned into
percentages later.

Most materials are OK measured to a percent
of accuracy. Some are fine with an even bigger
margin of error. Feldspars, frits and clays are
often the largest part of a recipe, and getting 40%
instead of 38% may make no perceptible difference.
On the other hand, some materials are powerful,
and a half-percent error with cobalt, will change
things radically.

Accuracy to a 100th of a percent is totally unnecessary,
though. Recipes are only written this way so that
they will add to 100, making for easy comparisons,
not because that degree of accuracy is needed.

-Snail

paul gerhold on sat 24 oct 09


Bob,
The real answer is that it depends on the glaze and the ingredient. I have
glazes where you can vary individual ingredients by plus or minus 20% ( of
ingredient amount) and not see any difference. And then there are others
where the ingredients can be varied almost no amount.

Semi- matt glazes are the least tolerant followed by matt followed by gloss=
y
as a general rule. That said you occasionally see glazes shown with the
ingredients to two decimal places which is just stupid ( except on really
rare occasions ) and usually indicates computer or calculator abuse.

As a general rule when I am done developing a glaze I do the math to put th=
e
ingredients in a 100% batch. Then round each ingredient to a tenth of a
percent, and then fudge the silica or clay so the batch adds up to 100.

On Fri, Oct 23, 2009 at 8:15 PM, Bob Johnson wrote:

> Apologies if this is an old horse being beaten anew--but as I was mixing
> glazes today I began to wonder how accurate glaze recipes need to be. My
> gaze program defaults to two decimal places, and that seems like a level =
of
> precision that wouldn't be noticeable in a glaze.
>
> Lots of things can throw the measurements off by way more than 1/100th of=
a
> graham. E.g., chunks that get stuck in the sieve or not getting the glaze
> materials centered on the scale.
>
> I know that precision--to some degree--depends on the type of ingredient,
> with cobalt seemingly more sensitive to measurements than, say, feldspar.
> But is there any rule? Can I round off all my glaze recipes to the neares=
t
> whole numbers?
>
> Thanks!
>
> Bob Johnson
> Southern Oregon
>

Bob Johnson on sat 24 oct 09


Thanks, all, for your comments on the accuracy of glaze measurements. I was
thinking along the same lines, but thought I ought to confirm with people
more expert than myself. I'll reset the default in my glaze program to one
decimal place. I gather that I can also forget about compensating for the
gunk that won't go through the sieve, too!

Best,

Bob Johnson

Roseburg, Oregon

http://www.flickr.com/photos/psychoceramics/

Dana & Chris Trabka on sun 25 oct 09


Bob,

When I use glaze calculation software, I will often tweek a recipe to get a
coeficient of expansion that fits my clay (for a base glaze - copper reds
are a different story). After tweeking, I often have a total of 143 or
something like that. When I retotal to 100, the amounts have too many digit=
s
to the right of the decimal point. I've used the side by side feature of
most glaze calculation software to "fix" the numbers to the right of the
decimal point. The recipe with the correct coeficient of expansion
(retotaled) is in one column, the one to adjust is in the other column.
Adjust things so that the Mole equivalents in the one you are adjusting sta=
y
the same as the retotaled one. Some times you can change a value a great
deal, some times you have to live with one decimal point. In a 100 gram
batch it is unlikely that you can measure to two decimal points. However, i=
n
a 1000 gram batch two numbers to the right of the decimal point is not much
of an issue and in a 10,000 gram batch it's not a problem at all.

When I look for a new recipe, I will often test a 100 gram batch with a
multitude of variations, just to see where it is going with various changes
(changing the amounts of one or two of the base chemicals). After selecting
a few of the "better" ones, I try them in a 2000 gram batch on some of my
pieces. If I like the results and the new glaze is "better" than any I
currently use, I add it to the collection (and get rid of my least
favorite).

As for colorants, cobalt is quite sensitive and so is iron and copper. In m=
y
brighest copper reds I use 0.5% copper. In my celedons a change of iron by
0.25% gives a dramatic change in the color.

Chris

Dana & Chris Trabka on mon 26 oct 09


Bob,

When I use glaze calculation software, I will often tweak a recipe to =3D
get a=3D20
coefficient of expansion that fits my clay (for a base glaze - copper =3D
reds=3D20
are a different story). After tweaking, I often have a total of 143 or=3D20
something like that. When I retotal to 100, the amounts have too many =3D
digits=3D20
to the right of the decimal point. I've used the side by side feature of =
=3D

most glaze calculation software to "fix" the numbers to the right of the =
=3D

decimal point. The recipe with the correct coefficient of expansion=3D20
(retotaled) is in one column, the one to adjust is in the other column.=3D2=
0
Adjust things so that the Mole equivalents in the one you are adjusting =3D
stay=3D20
the same as the retotaled one. Some times you can change a value a great =
=3D

deal, some times you have to live with one decimal point. In a 100 gram=3D2=
0
batch it is unlikely that you can measure to two decimal points. =3D
However, in=3D20
a 1000 gram batch two numbers to the right of the decimal point is not =3D
much=3D20
of an issue and in a 10,000 gram batch it's not a problem at all.

When I look for a new recipe, I will often test a 100 gram batch with a=3D2=
0
multitude of variations, just to see where it is going with various =3D
changes=3D20
(changing the amounts of one or two of the base chemicals). After =3D
selecting=3D20
a few of the "better" ones, I try them in a 2000 gram batch on some of =3D
my=3D20
pieces. If I like the results and the new glaze is "better" than any I=3D20
currently use, I add it to the collection (and get rid of my least=3D20
favorite).

As for colorants, cobalt is quite sensitive and so is iron and copper. =3D
In my=3D20
brightest copper reds I use 0.5% copper. In my celedons a change of iron =
=3D
by=3D20
0.25% gives a dramatic change in the color.

Chris=3D20