"Andy Hall" wrote in message
...
On Wed, 25 Feb 2004 11:39:34 GMT, Tony Bryer
wrote:

In article , Andy Hall
wrote:
Are you sure about that? Granite is not going to flex at all, so
fixing it with an adhesive that has no give to something that might
but shouldn't doesn't seem intuitively right to me.

If it doesn't flex at all then the steel is redundant as it must
deflect by some (minuscule) amount if it is carrying a proportion of
the load. If the granite just bears on the steel you'll have a load
sharing situation. If you fix the steel to the granite with epoxy or
whatever and put weight on the granite it will be in compression and
the steel in tension like a RC beam.

Makes sense. Since the steel, I would have thought would flex rather
more than the granite, I would expect that the granite would crack.
Unless I've missed something?



yes, I think you've missed something in this.

The granite will break not because of the compression it suffers at the top
but due to it's weak tensile strength on the underside.

If you bond steel section firmly to the underside it gives the granite the
required tensile strength and therefore prevents the breakage.


The alternative is to provide a firm inflexible support for the granite and
not bond it securely. You limit the amount that the granite can deflect and
hence the maximum tension that it will suffer underneath. This would mean
that any flexible bedding or sealant used between steel and granite must be
as thin as possible, because this would allow slight movement.


One thought that does strike me relates to the expansion of the steel with
heat and whether the firmly bonded solution would be problematic in this
instance. Might the heat kicked out from the WM and TD be sufficient to
cause problems with this? (thinking aloud here, the granite ought to
provide a good thermal sink thus preventing the steel from heating up and
therefore expanding too much).


--
Richard Sampson

email me at
richard at olifant d-ot co do-t uk