Hello Werner,
At 11:26 AM 11/11/99 EST, Werner Engelmaier wrote:
>Hi Ingemar,
>I certainly would agree that 'Design for Reliability (DfR)' is necessary, in
>this case and all others, to assure adequate long-term reliability of the
>solder joints. However, for chip caps "FEA and solder joint CAD simulation
>tools" are unnecessary overkill. For such a simple arrangement closed-form
>equations will do nicely, see IPC-D-279. The more complex the modeling, the
>more assumptions, often not very well founded ones, have to be made. FEA and
>CAD tools are necessary for complex structures, such as BGAs, but not for
CCs.
I don't have a copy of D-279 - are its "simple arrangement closed-form
equations" simple shear models for cyclic strain combined with a Coffin-Manson
fatigue life model? If not then please ignore the following!
The tests I mentioned earlier in the 1812 capacitor thread (repeated below)
were the first hard evidence I had that such models might not be adequate,
even
for relatively simple situations - the 1812 DNP is 50% more than the 1206 and,
using typical Coffin-Manson coefficients, an approx 50% lower life would
therefore be expected. The 1812 results were however very much worse that.
I also presented a study at the 1989 ISHM Europe Conf. which showed that such
models gave very poor correlation with a wide range of published experimental
results. I inevitably had to make a lot of assumptions regarding the
experimental data, as few people publish the full story, & your own work (and
others') has enormously improved the predictive power of this type of model
under different cycling conditions, but can they explain my experimental
results?
David Whalley
>I ran some tests over 10 years ago where we did -55 to +125 cycling tests
>of (mostly) 1206 and 1812 chip components wave-soldered onto FR4. The 1206s
>survived 2000 cycles with little evidence of degredation, but the 1812s had
>severe cracks in the solder fillets after only 1000 cycles. The test regime
>may or may not have much to do with your own product use environment, but
>the results do show that things rapidly get worse for larger components.
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
|
