Hasan,
I agree with the flags you raised about various FEA approaches. Another issue
is that the correlation of test data has a strong dependence on element size
(this was demonstrated with hard data and careful studies by R. Darveaux at
SEM'96). This may not be a problem when you model similar joints and do
parametric studies on package and board variables alone.  However, the FEA
approach  becomes an art when you have to deal with joints of different
sizes, as in flip-chip vs. QFPs vs. BGAs. You then have to use consistent
element sizes from one model to the next and make sure that element sizes are
somewhat the same as in the FEA models that were used to develop a
correlation of test data. These problems arose on top of the singularity
issues that you mentionned. Also, because of long run-times and high
computational costs  ($25K+ a year to lease a commercial code that handles
plasticity and creep well - we don't get discounts in industry), most of us
can't afford experimenting with mesh sizes. 

The averaging techniques that you mention have helped alleviate some of the
problems. However, since by averaging results over the volume of a joint, you
have a single number (whatever is used for damage), you end up in a "1D"
situation. My view and experience has been that hysteresis loops from 1D
(plane or axisymmetric) engineering models provide sufficient information to
capture main effects and make a first order reliability prediction. (see my
home page below for more info on how this approach was implemented). First
order predictions are satisfactory for day-to-day engineering purposes (in
general). This may also be the best we can do  since we deal with a fatigue
situation with joint-to-joint, board-to-board and product variability.

I'll be looking forward to your reports on improved FEA approaches at
INTERPack'97.
With best regards,
Jean-Paul
________________________________________________________________
Jean-Paul Clech
EPSI Inc., P. O. Box 1522, Montclair, NJ 07042 - USA
tel: +1 (201)746-3796  fax: + 1 (201)655-0815
Home page: <A HREF="http://members.aol.com/Epsiinc1/index.html">http://members
.aol.com/Epsiinc1/index.html</A>
- click on the above or copy: http://members.aol.com/Epsiinc1/index.html
__________________________________________________________________







***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact   *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************