Re: Pyrex "Explosion"

["John S. Denker" <jsd@MONMOUTH.COM>]

At 01:36 PM 2/20/01 -0500, Polvani, Donald G. wrote:
> Last week I witnessed what I can only call an "explosion" of a Pyrex baking
> dish. .... My
> chemist daughter tells me that Pyrex glass should be capable of going from a
> warm oven into a freezer.  She would heat quartz (not Pyrex) tubing up to
> 600 deg C and then thrust it into ice water.

I've done things like that.  Once upon a time I was trying to seal a toxic
volatile explosive substance into a glass tube.  I was melting the top
shut, while a few millimeters away the tube was soaking in liquid nitrogen
(so that the substance would stay put).  No problem.  In one case a clamp
slipped, and the glowing glass dropped right into the liquid nitrogen.  It
didn't break.

> I really don't understand the mechanism which caused such a violent event
> to occur.

1) The warranty the comes with pyrex utensils says that heat won't break
them if used according to instructions.  However, the instructions say "do
not use if cracked".  Pretty sneaky, huh?  I betcha there was a smallish
pre-existing crack somewhere.
  -- Thermal stress is OK if there is no way for cracks to nucleate.
  -- Cracks are OK if there is not too much thermal stress.
  -- The combination (cracks + stress) may well cause problems.

2) The coefficient of thermal expansion for Pyrex is less than half that of
ordinary glass.
  http://hyperphysics.phy-astr.gsu.edu/hbase/tables/thexp.html
This means you have to provoke it quite a bit.

For fused quartz, the coefficient is fantastically small.  You can get away
with really amazing things.

3) The conversion of thermal stress into kinetic energy might proceed along
these lines:

The inside is hot.  The outside is cold.  The inside is expanded relative
to the outside.  If all you had was a flat piece of glass, it could
accommodate this by _bowing_ upward, to make a concave-downward spherical
shape.

However, the walls of the bowl complicate this picture.  If (for
simplicity) we assume the wall is just a bystander, there will be
tremendous bending stresses where the bowed bottom meets the wall.  Suppose
there is a tiny crack, and it starts to spread.  Once it starts, it will
propagate very rapidly.  Stress will be relieved.  The wall-to-bottom angle
will be restored to its normal value, causing the bottom to fly downward
and/or the wall to fly outward.

Also, at each point along the path of the crack, there will be some glass
under tension and some under compression.  The crack relieves the
tension.  The compressed pieces maintain contact for a moment longer,
pushing the pieces apart.

We are talking about huge forces here, as you can see by multiplying the
coefficient of thermal expansion times an elastic coefficient e.g. Young's
modulus.