If you're not interested in all the technical details of "why",
or you just want
to know how to check if one glass will fit another, skip this first
part and head over to Page 2 for "Testing and
Ever notice that most commercial glassware is only one color? That's because it's hard
to put two different colors of glass together hot without
having problems. Art glass is usually the only place you'll
color combinations done hot.
The problem is stress where the different glasses join.
This stress will cause cracks, breakage or even explosions in the glass
cools, or, sometimes much later, when it's sitting in someone's house.
There are a number of things that create stress.
First on that list is...
Coefficient of expansion or COE
COE means coefficient of expansion.
Big words but a simple idea.
It's a measurement of how much a glass expands (heating up) or
contracts (cooling down)through a certain range of temperature. COE is
measured in ten thousandths of an inch .
In art glass, almost everyone uses the English & Turner system for
finding, calculating and correcting expansion. Much more later about
The temperature range used to measure the expansion of the glass in the
English & Turner system is from 20 to 300 degrees
A 96 COE glass is one that expands or contracts 96 ten thousandths of
an inch as it
heats up or cools down through the range between 20 and 300 degrees
These are the most common COEs seen in art glass today, and the
manufacturers best known for them:
90 COE = 90 ten thousanths of an inch expansion - Bullseye Glass
96COE = 96 ten thousandths of an inch expansion - Spectrum, Gaffer,
104 COE = 104 ten thousandths of an inch expansion - Moretti
If you're batching and melting your own glasses you can change the COE
of your glass.
In case you didn't know, "batch" is what glass melters call the mix of
coloring agents and other odds and ends that gets melted down into
We'll start with....
What is Glass?
Basic glass is sand, alkalai to lower the sand's melting
temperature, and calcium (lime) to help harden the glass and keep all
the alkalai from dissolving out of the glass when it gets wet.
Soda ash is the most common alkalai used so that's where the term "soda
lime" glass came from.
Add more sand and the COE goes down
Add more alkali or calcium and the COE goes up.
There are a lot more chemicals that can end up in glass to change its
characteristics. If you're going to add these others you need to know
if they raise the COE or lower it.
That's where the E&T system begins to shine. It tells you how much
COE change a certain amount of a certain chemical will cause.
Many think that the only time two glasses will create a mismatch, and
the resulting stress, is when the COEs aren't the same. Unfortunately,
Why the Big Emphasis on COE?
Due to a lot of misinformation left over from an earlier era of
US art glass and over-simplification by some glass
marketers over the years.
That's why so many people think COE is the only thing that has to
match for different glasses to fit together without stress.
Sometimes you have to make the COE of one glass very
different from the other to get them to fit together without problems.
All the glasses like System 96 or Bullseye 90 are not all 96 or 90 COE
glasses. They are a system of glasses that will fit each other within
that system without problems. There will be different COE glasses
within that closed system.
It's a long story, get comfortable, and we'll start with...
Annealing and Strain Point
Is the temperature at which the glass is just about to
turn from liquid to solid, or the other way around if you're heating it
up instead of cooling it. It's just liquid enough to flow at this
Whatever you are annealing isn't going to collapse or distort because
it's close enough to a
solid to hold its form.
Is where the glass becomes a solid.
Hot glass above the strain point is liquid and plastic. It will flow
and has no stresses.
Once glass cools to the strain point it does not flow any more and it's
done expanding, contracting or flowing.
When cooled off and solid, after passing through the strain
point, glass is both elastic and brittle.
Elastic meaning it can bend but will return to it's original shape,
like a spring.
Brittle meaning it breaks cleanly with no distortion at the break, like
Let's start with the problems you can get into with annealing
Any glass worked hot must be cooled slowly enough to anneal
Means cooling the glass slowly enough to be sure the temperature all
the glass stays almost equal while cooling. This gets tricky because
very slow to take on or give up heat. The thicker the glass, the slower
the heat penetrates or leaves the glass.
Even without mismatched glasses combined in the same piece,
you can put
big stress into a piece of glass by not annealing correctly.
Let's say we're annealing a piece of glass and it's all the same glass.
- If the temperature drops too fast, then the outside of the
piece cools to the strain point and sets up before the
- The inside and the outside of the glass will be trying
to pull away from each other because the outside quit contracting while
the inside was still shrinking.
Another way to think about that...
- The outside is being held at a larger than normal size by
of all the glass, inside and outside.
- The outside becomes solid at this
- Now the inside of the glass is trying to shrink down to
its normal smaller size as it cools but it can't.
- Huge stress builds up as the inside tries to tear away from the
Mismatched glasses stress
of the two glasses goes from liquid to solid (gets to strain temp.)
the other as they are cooling, Once you get at or below the strain
temperature of one of the glasses, there's no flow and no more
shrinkage in that glass.
- The other glass joined to the first glass hasn't hit it's
strain point temperature yet so it's still shrinking.
- Stress builds up as the second glass tries to pull
away from the first. when the second glass hits its strain point
temperature (becomes solid) then the stress from it trying to pull away
the first glass is locked in.
Even though it's all about the strain point in both cases,
annealing problems and mismatch problems are often mistaken for each
other. If you're having problems that you think are mismatched glasses
then one of the first things you need to check is your annealing cycle
and times. Run a piece of clear of the same type and size through the
anneal process and then check it for stress with a polariscope when
it's done. If this comes out OK then you can be sure you've got a
mismatch and not an annealing problem.
Expansion and contraction happen in very tiny increments, which is why
we measure them in ten thousanths of an inch.
The stress they create when they mismatch or are not annealed correctly
are NOT tiny, even at just a few ten thousandths of an inch.
These stresses make huge pressure, as in tens of thousands of pounds.
Which is why artwork or any glass with mismatched glasses and/or poor
annealing can explode like a bomb.
All those plate glass table tops and other pieces that have been
exploding at random lately are caused by melting the glass with
inexpensive ingredients that contain something that will cause tiny
in the glass.
The stones are a combination of nickel and sulfur that melt together
and form little stone-like pieces of incompatible glass inside the
finished glass. These stones have a VERY different strain point from
the rest of the glass. Even though those stones are tiny, they make
enough stress to shatter a piece of 1/2" or larger plate glass.
Strain Point is where the stress gets set into the glass
things aren't right. Right means any different glasses have been
correctly matched and the annealing was done correctly.
If you are combining different glasses, you might think the perfect
situation would be for both glasses to have the same strain point, and
We'll get into exactly how to get that done and what it is we're really
changing as we look into...
You're usually trying to adjust a colored glass to fit the clear glass
If there is a mismatch, what you have to do is change the COE
of the colored glass to get it to fit the clear.
You end up with two glasses that are different COEs but that fit each
other with no stress. This happens because the strain point temperature
of both glasses is the same and that's what really makes a go or no go
out for combining glasses.
Bear with me...It's story time again, starting with the history of
trying to get glasses to fit each other.
Before some very generous glass scientists started helping hot
glass artists by teaching them the finer points and "why" of annealing
and matching glasses, a lot of glass matching was done by a
"rule of thumb". That rule said that as long as you didn't add more
than 1% of
anything that made glass color to one of the glasses, that the glasses
two glasses would fit together well
Superstition has always been a problem with glass info. Coming into the
twentieth century many factories were still melting glasses by the
phase of the moon.....which has no effect.
You still hear this "1% rule of thumb" today so let's look at
where it's right and where it's soooooo wrong...
The rule is partially right because the transparent colors aren't
usually a mismatch problem. They take well under 1% of the oxide used
them to get the desired result. This 1% or under is usually put into
the same formula batch recipe that was used to melt the clear so the
strain points and everything else stay close enough that they fit
The 1% rule is wrong because some glass color additives will cause
plenty of trouble at well under 1%.
Fluorine, Phosphorous, Silver, selenium and/or cadmium (red and
yellows) and many other chemicals used to make striking glasses will
cause problems at low percentages, way below 1%.
Striking glasses are the ones that change color,
color density, or opacity when they are cooled and reheated. Mostly
opaque glasses, silver glasses, reds and yellows.
The 1% rule is also right if you're melting very dense transparent
colors, like the color bars furnace workers
use. The percent of added colorant used has to go well over 1% and then
there's trouble...big trouble.
So how did hot glass artisits have as much success as they
did in matching glasses in the past?
Well, they didn't actually solve the problem, there was a "patch" to
get around the problem...
Many of the dense colors for furnace glass workers in the past, and
even now, are full of lead...as in they're 40% lead.
Lead does incredible things to increase the elasticity of glass. As a
nice side benefit, it makes it pretty too.
There can be a large mismatch in strain points and the lead will allow
the glass to
stretch enough to relieve a lot of the stress....unless you try to saw
it, that is. There's enough stress left in the glass to cause it to
break if sawed.
The problem was "patched" but wasn't really solved this way and lead is
one nasty toxin.
So why all the trouble, workarounds and strange problems?
Because not many people then or now understood the effects of changes
Viscosity is how easily a liquid will flow, or, the more correct
definition, reistance to flow.
If the two glasses you're trying to match have the same COE
but have different viscosities, there
will be a different strain point for each of the glasses.
There will be permanent stress left in the glass when it's cooled down
and done because they didn't hit the strain point together.
Annealing has no effect on this, it can't change the strain point so it
can't relieve this type of stress.
A lot of the problems with the troublesome color additives mentioned
above are because they make BIG changes in
the viscosity of the glass when tiny amounts are added.
The same problem happens with over 1% regular color additives, the
viscosity starts to go crazy at that point.
These viscosity changes are not linear or predictable.
As in...doubling the amount of color additive doesn't necessarily
double the amount of viscosity change.
It gets even crazier with the striking glasses because the
structure AND the viscosity of the glass change radically at the point
cool enough to strike. This is usually right close to or at the strain
point. Each seperate strike after the first (some glasses need more
than one strike to bring out the color) changes the
structure and the viscosity
Complex problem, right?
Yeah, it is, so let's get into some more detail on how to fix it,
assuming you're melting your own glasses and can change the chemical
makeup to change the COE.
Fortunately, changing either COE or viscosity change will change the
Changing the COE changes the strain point in a gradual,
linear and predictable way.
Remember all the crazyness in viscosity we talked about before with a
lot of the additives that make color? No tracking that, right?
You need something that's predictable to test and correct with.
To put it another way...
- There are two things you can change to get a match at the
strain point, viscosity and COE.
- You can't know or measure strain points without expensive
- You can't measure or track viscosity changes without
expensive lab equipment. Even if you could, it wouldn't be linear or
- You can predictably measure and change COE with simple
inexpensive thread and polariscope tests. You can track the
changes with the E&T factors. You can also know if
your strain points match so that you've eliminated stress with these
When you get it right you can have two glasses with different
different viscosities but a match on strain points...and they will fit
each other without stress.
Let's go over to Testing & Measuring for some how-to.
Want to read more, and I mean a LOT more about this?
Read a LONG discussion among many of the old hands in the art glass
world about how the COE "mistake" happened here