🚀 go-pugleaf

In article <1992Oct28.223610.203965@ns1.cc.lehigh.edu>,
ps02@ns1.cc.lehigh.edu (PENELOPE E. SMITH) wrote:

> Question:  Suppose I want to build glass bookcases anyway.  How fast will
> those cracks grow?  Can I make my shelves heavy enough, say, to last at least
> 100 years?  Does it help to smooth the stress concentrations at the supports?
> -- 


I once saw a rather amazing display of the effect in question.  A professor
of mine took two 1/2 inch glass rods.  One he slammed against a table top
which caused the rod to obviously shatter into a million pieces (yes, he
did it in a contained area behind a shield).  The next rod he immersed into
an acid solution (concentrated sulfuric I believe) then rinsed the rod.  He
slammed it agianst the table.  It did not break!  

Thinking back, you could conclude that the presence of moisture is not the
sole cause of crack propogation, but also crack geometry.  The acid would
round the crack tips and prevent propogation.  Correct me if I'm wrong out
there (and I'm sure you will) but this may be the one result of tempered
glass, in addition to getting rid of crystalline defects.

Obviously the success of glass shelves is very dependant on the loading. 
Glass is extremely common in shopping malls, so its not that it shouldn't
be used for shelves.  "Smoothing" stress concentrations at the supports
will do no good.  If "smoothing" would work you would have to do it at the
greatest surface tensile concentration, which occures on the underside of
the shelf at the point farthest (in general) from any support (assuming
even loading), ie at the greatest bow or bend of the shelf.



kevin weber      weber@jedi.ssd.lmsc.lockheed.com
==============================================================================
The greatest sign of intelligent life in the universe is the fact that they
have not yet tried to contact us.
                                     Calvin, Calvin and Hobbes
==============================================================================

Kevin Weber writes:
>I once saw a rather amazing display of the effect in question.  A professor
>of mine took two 1/2 inch glass rods.  One he slammed against a table top
>which caused the rod to obviously shatter into a million pieces (yes, he
>did it in a contained area behind a shield).  The next rod he immersed into
>an acid solution (concentrated sulfuric I believe) then rinsed the rod.  He
>slammed it agianst the table.  It did not break!  
>
>Thinking back, you could conclude that the presence of moisture is not the
>sole cause of crack propogation, but also crack geometry.  The acid would
>round the crack tips and prevent propogation.  Correct me if I'm wrong out
>there (and I'm sure you will) but this may be the one result of tempered
>glass, in addition to getting rid of crystalline defects.

The acid technique does work, for the reason listed, but once you
scratch (by handling in any way) such a treated piece of glass, you have
the same problems again, so it is not a general way to increase the
strength of glass.

Tempered glass is glass that has been cooled rapidly in a controlled
fashion so that it ends up with its surface in compression, which is
balanced by the interior being in tension.  This phenomenon is possible
because glass gets denser the more slowly you cool it (the structure has
more time to "relax").  The fact that the surface is in compression
greatly reduces crack growth due to purely mechanical reasons, because
the stress at the crack tip must overcome the surface compression in
order to grow.  I don't know about growth due to chemical reasons, such
as being attacked by those darned polar water molecules, but presumably
it would make that more difficult too.

One of the problems with tempered glass is that the stress contained in
the glass is easily sufficient to shatter it, if it can only get out.
This means that *one* crack that gets past the surface compression zone
will be sufficient for catastrophic failure of the entire part.  This is
why car windows break up into little itty bitty pieces instead of big
shards, but usually are extremely reliable.

I heard a story once (possibly Urban Folklore) about a bunch of car
windows spontaneously shattering in Southern California once.  It seems
it became a popular topic of discussion, because people couldn't figure
it out.  So this psychic said something about the area was cursed by
some dead person or other, and got a lot of press.  Some responsible
people investigated, and it turned out that a factory got its
information wrong about how to install windows, and there was a piece of
metal rubbing on the glass that was causing the catastrophic cracking.

This has been cross-posted to sci.materials and alt.folklore.urban.  If
there are any alt.folklore.urban follow-ups, please delete sci.materials
from the subject line.

				- Rob

ps02@ns1.cc.lehigh.edu (PENELOPE E. SMITH) writes:

>>probably the reason for the demise in use of  of glass bookshelves,
>>at least in damp (UK!) houses. Big elastic load - long time - moisture.

>Question:  Suppose I want to build glass bookcases anyway.  How fast will
>those cracks grow?  Can I make my shelves heavy enough, say, to last at least
>100 years?  Does it help to smooth the stress concentrations at the supports?

Yes you can.  Whether you call the phenomenon "environmental stress cracking"
or "static fatigue" [just a little dig there!], it does depend on stress.
As one of the previous articles pointed out, crack growth rate is dependent
on stress.  So, all you have to do is ensure that the stresses are low,
which can be achieved, as you suggest, by making the shelves thick enough.

Better than that, though, I would suggest spacing your supports a bit closer.
The trick is to avoid _tensile_ stresses (and bending, of course, sets up
tensile stresses on one side);  _compressive_ stresses are not a problem,
since they tend to cause cracks to close up, not grow.  So! if you find it
aesthetically pleasing, you can happily make the uprights out of glass as
well.  And, since they are glass, no-one will feel you have over-designed
if you add a few more uprights than normal in order to reduce the bending
loads on the shelves.

As for the stress concentrations at the supports, it probably does matter,
but _not_ because such stress concentration points will _initiate_ the
cracks;  glass has quite enough surface flaws already without needing to
initiate any more.  But stress concentrations matter for the reason I gave
above  --  that it simply increases the _stress_ on any given crack in that
vicinity.

We have a really natty lab exercise for the undergrad students here,
to teach them the basic principles of fracture mechanics.
We have a simple apparatus for putting glass rods into 3-point bending
under a dead-weight load.  We then take some of the glass rods (ordinary lab
glass, 6 mm diameter) and etch them in 50% hydrofluoric acid
	* WARNING!  For anyone not familiar with HF:  It attacks glass,
	so as you can imagine, it also attacks bone, and does
	indescribably horrible things to skin and eyes.  And it's not
	easy to wash off, either.
	DON'T USE IT UNLESS YOU KNOW WHAT THE SAFETY MEASURES ARE! *
This removes the surface, including all surface flaws.  What this does to
the bend strength is really quite dramatic.  An as-received rod will typically
support about 6 kg;  etched rods will support as much as 60 kg.  The modulus
of rupture comes out well in excess of 1000 MPa!  We get the students to
calculate the flaw size from the Griffith equation (given a value for the
surface energy).  We have estimated the flaw size in the as-received
glass and it agrees well with the calculation (about 5 micron, I think).
It is interesting to note that the edge of the central beam in the loading
jig must damage the surface, yet this doesn't matter because that side
of the rod is in compression.

Students often ask about the potential engineering applications for this
"wonder glass".  We have to disillusion them, of course  --  10 minutes
of usage would introduce new surface flaws.  Unless resin-coating immediately
after etching would prevent damage .............?

Happy book-case building!

|  jeff@minmet.uq.oz.au	       |  Jeff Gates
|			       |  Dept Mining & Metallurgical	       _--_|\
|  Ph. direct: +61 7 365 3726  |		     Engineering      /      B
|  Ph. Dept: +61 7 365 3738    |  University of Queensland	      \_.--\_/
|  Fax: +61 7 365 3888	       |  Brisbane   Qld   4072   Australia	    v

WEBER@JEDI.SSD.LMSC.LOCKHEED.COM (Kevin Weber) writes:

>I once saw a rather amazing display of the effect in question.  A professor
>of mine took two 1/2 inch glass rods.  One he slammed against a table top
>which caused the rod to obviously shatter into a million pieces (yes, he
>did it in a contained area behind a shield).  The next rod he immersed into
>an acid solution (concentrated sulfuric I believe) then rinsed the rod.  He

Hydrofluoric

>slammed it agianst the table.  It did not break!  

Yes, see my earlier article re our lab demo

>Thinking back, you could conclude that the presence of moisture is not the
>sole cause of crack propogation, but also crack geometry.  The acid would

Yes, this is standard fracture mechanics.

>round the crack tips and prevent propogation.  Correct me if I'm wrong out
>there (and I'm sure you will) but this may be the one result of tempered
>glass, in addition to getting rid of crystalline defects.

I guess it's possible the heat might blunt the cracks, but the
heat treatment of glass works by another main mechanism:  it puts compressive
residual stresses into the surface, which tend to close up cracks.  To see
this, try grinding back the surface of a Duralex glass (hotel glasswear)
on SiC paper: once you expose the tensile residual stresses underneath,
you better watch out!  (WEAR SAFETY GLASSES FOR THIS ONE!)

|  jeff@minmet.uq.oz.au	      |  Jeff Gates,   Dept Mining	   |   _--_|\
|  Ph. direct: +61 7 365 3726 |        & Metallurgical Engineering |  /      B
|  Ph. Dept: +61 7 365 3738   |  University of Queensland 	   |  \_.--\_/
|  Fax: +61 7 365 3888	      |  Brisbane   Qld   4072   Australia |	    v