Dave Bross.com

Home About Contact SiteMap/Contents Media

Testing and Measuring -

Let's start with the quick and easy test and then go to the more complicated, difficult or more time consuming ones.

The one I always use is called the...

Thread Test

Two equal amounts of glass are melted together.

Pull the melted glasses into a thread.

Let it cool

Measure the bend (if any) in the thread.

If the thread is straight when cooled, the glasses match.

If the thread is bent, the glasses don't match.

The more the thread bends, the bigger the mismatch.

The higher expansion glass is on the inside of the bend.

There are some glasses this doesn't work very well on.

It's not extremely accurate.

1. Two equal amounts of glass are melted together.

The easiest way I've found to do this is cutting about 1/2" wide strips of clear sheet glass from a known and tested expansion (COE) sheet of Spectrum 96 for 96, Bullseye for 90, Moretti for 104. Spectrum and Bullseye test their glasses, I assume Moretti doesn't test their glass because it can vary a little, but not too bad. This way we have one glass that we are sure what it is.

Note: I think Frantz Art Glass is about the only place to get Moretti sheet anymore.

the strips of sheet glass

Melt some of the unknown glass to be tested on to the strip of known COE glass. You're trying to get an amount on there that will be the same size and depth as the part of the strip it's covering.

melting the mystery glass on

Mash it flat with the tweezers so you can get a good idea how much unknown glass is on the strip of known glass. It's much easier to judge how much when dealing with the same shapes on each part.

mashing with tweezers

Be sure you've got the same amount and thickness.

mashing with tweezers

2. Pull the melted glasses into a thread.

Melt these two together . They need to be very liquid so that we have enough time to pull them into a thread.

mashing with tweezers

Grab the melted glass with the tweezers and pull.

I always turn the strip of glass at a right (90 degree) angle to the thread so I can remember which section of the melted glass was the strip and which was the glass I added to the strip.

Keep pulling on the thread until it cools so you're not putting any extra bend into by letting it droop while hot.

We need a 200 mm or 7.87" long section of this thread to measure the expansion so the thread has to be pulled a bit longer than that.

sorry I don't have a pic of the actual pull. I couldn't hold the camera and do the pull.

pulling the thread

3. Let it to cool.

Break the blob off the end where the tweezers were so its weight doesn't throw us off when we go to measure the thread.

4. Measure the bend (if any) in the thread.

I've cut this piece of aluminum angle iron to exactly 200mm, or 7.87 inches. This is what I use to measure the bend from from.

the thread and the angle iron

I use an inexpensive set of measuring calipers. They're accurate enough.

I measure the distance between the bent thread and the angle iron by eye, being sure I'm looking straight down. You know you're right when you see only the top of the angle iron, not any of the side, like in the pic below.

measuring the thread

5.If the thread is straight when cooled, the glasses match.
6.If the thread is bent, the glass do not match.
7.The more the thread bends, the bigger the mismatch.

Anything more than a 1.5 mm bend in the thread means that glass can cause mismatch trouble in a finished glass piece.

There are some interesting exceptions and workarounds to this. We'll go over some of these in a minute, and some others later, at the end.

You can't read the calipers in the pic but this one is at 1.5 mm, so it's just OK.

8. The higher expansion glass is on the inside of the bend.

The higher expansion glass is going to expand further than the lower expansion glass. It's also going to shrink more when it cools. This is going to force the lower expansion glass to bend along with the higher expansion glass as it shrinks.

This is what's making stress inside a piece of glass when the glasses don't match. Deep inside a piece of glass there is nowhere for the lower expansion glass to go to compensate for the mismatch. These threads are thin enough to bend to compensate. This also gives visual proof that glass is elastic.

9. There are some glasses this doesn't work very well on.

The two glasses tested need to have about the same amount of elasticity. Technically that's called/measured as the modulus of elasticity.

I know, "elastic" is not a word you think of with glass, particularly when it's cold, but it's there and it's important.

Lead makes glass a LOT more elastic. it's been used for many years (centuries?) in large quantities for colored glasses so they could get away with larger mismatches and still have the glasses hold together. It works unless you try to saw the glass, then it will break.

If you try the thread test with one lead glass and one regular soda lime glass your results will be off. Way off.

This is also true for all the striking glasses, fluorine whites, phosphate whites, reds and yellows and silver glasses.

You can get these troublesome glasses to work with a thread test, you just have experiment and find out how far the different glass throws the result off and factor that in. That's done by multiple trials and checking the results for stress with a polariscope.

With many of the striking glasses you can figure out the number of strikes needed to make the piece you have in mind and strike the combined glasses melted together for the thread test that many times before you pull the thread. This helps increase accuracy a lot.

10. It's not that accurate.

Not accurate like laboratory tests, no, but it's fast, costs next to nothing and it's more than accurate enough to check what you're doing. More important, you can find and fix trouble before it happens or use it to figure out what has to be done to fix a known problem. We'll get into that when we go into E&T factors and using them to fix problems.

Some other common tests for glass mismatch are:

1. For furnace glass, blowing out a cylinder with an equal thickness of both glasses, one inside the other. You anneal the cylinder, score the cylinder, then crack it open the score. The two ends will either stay put, open wider, or try to close on themselves depending on how much mismatch there is. No movement of the split means the glasses match.

Not many people are skilled enough to get the two layers of glass equal thickness. It's tough to measure how much you're off. It takes a long time.

2. Fusing the glasses together in a kiln and then looking at them under a polariscope. A final check with a polariscope is always a good idea for any of these methods.
Running the glasses through a cycle in the kiln takes a long time.

There's an article about homemade polariscopes here