Roger,

A very good place to start is the proceedings of Surface Mount International
for 1994 and 1995.  There are several articles that present data on this very
topic.  It is interesting to review the 1995 articles after reading the 1994
stuff.  Some of the concerns regarding solder joint cracking that existed in
1994 appear to have been shown to be non-problems? by 1995.  Look particularly
for articles by Compac Computer Corporation.

You might also consider articles that describe the use of an epoxy underfill
to reduce thermal-expansion induced joint cracking.  Hopefully some folks from
Motorola and IBM will respond to your inquiry as well.  I have seen data that
predicts lifetimes of BGA solder joints in typical commercial product 
environments.

For what its worth:

We have been conducting joint lifetime tests with very small, 10-mil diameter
"solder domes", over the last couple of years.  Epoxy underfill clearly reduces
solder joint failure.  And just like Motorola and IBM have advertised...the
solder assembly reliability of this packaging scheme is very good.  Even the
mini-BGAs we have been building (10-mil diameter x ~ 4-mil high domes on 20-mil
pitch) are quite easy to solder reliably.

For more information regarding the work Sandia Labs is doing, contact 
Keith Treece at email: [log in to unmask]
                phone: 505-844-9684

Gary P.
---
                                  Gary D. Peterson
    _/_/_/   _/    _/  _/        SANDIA NATIONAL LABORATORIES     _/_/_/
   _/       _/_/  _/  _/        P.O. Box 5800, M/S 0503            _/_/
  _/_/_/   _/ _/ _/  _/        Albuquerque, NM 87185-0503     _/_/_/_/_/_/
     _/   _/  _/_/  _/        Phone: (505)844-6980           _/  _/_/  _/
_/_/_/   _/    _/  _/_/_/_/  FAX: (505)844-2925             _/  _/_/  _/
                            E-Mail: [log in to unmask]        _/_/_/