Answer 1:
Regular and irregular objects do not appear to
be standard terms. However, a
Platonic solid (3D) could be what is meant
by a regular object. Such solids have
faces which are identical and regular polygons.
Regular polygons means that the lengths of
all sides are identical and the angles between
sides are identical. There are only 5 Platonic
solids. If we take those to be "regular"
objects, then irregular objects are any which
do not meet those criteria. The question
then becomes why cracks in glass do not
propagate along directions which produce those
faces. By definition,
atoms in glasses are not arranged on planes
but are randomly positioned. One
consequence of this is that their properties do
not depend on direction. We can therefore expect
that the material itself does not have a preferred
direction for crack propagation. Random
propagation directions are unlikely to
produce the regular shapes described above.
The direction in which a crack grows also
depends on the stresses present in the
material.
Many factors contribute to this, from
residual stresses introduced during forming,
chemical treatments to change the structure in
some regions, differences in temperature across
the piece of glass, etc. These effects are
typically not spread uniformly throughout the
entire piece of glass, so the direction of
propagation will change with location. That
being said, these principles can be used to cause
glass to break into (relatively) regular pieces.
Tempered glass uses special processing
to produce a stress state that causes glass
to break into small pieces that can be roughly
cuboid. [Although not part of the original
question, manipulation of the stress state to
change the crack direction is being
investigated due to implications for certain types
of small-scale fabrication.]
| 
