SRC Forum - Message Replies
Forum: Reliability & Maintainability Questions and Answers
Topic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Posted by: Tom Donnelly
Date posted: Sun Jan 3 1:59:31 US/Eastern 1999
Subject: Screening solder workmanship
I am currently working with space related hardware. The issues I am having are with solder joint reliability and possible sources for specific failure rate information as it relates to the thermal stress testing of PWB's.
What is the difference in solder joint failure rates when the electronic equipment is exposed to -45F versus -30F?
and what failure mechanism's did I miss by only exposing the equipment to -30F.
I have the same question for taking the equipment to +60F instead of +85F. What is the differnce in failure rates for the solder connections and what failure mechanism's did I miss.
From my many years in Reliability I believe that the difference is insignificant, but my customer's would like me to quantify it....I appreciate any thoughts you could contribute to solving this problem...... T.Donnelly
Subject: Thermal stress testing of PWBs
Reply Posted by: Larry Mosher
Organization: Abex NWL Division of Parker
Date Posted: Tue Jan 5 14:56:16 US/Eastern 1999
Several years ago, I evaluated a software package called "PRAC" from International Analytics Co., P.O. Box 2051, Huntsville AL 35804-2051. Phone 205/539-8118. This package modeled both thermal stresses and vibration stresses and predicted solder joint life. It struck me as very easy to use and relatively cheap ($1500 in 95).
Subject: Solder Screening
Reply Posted by: Bruce Dudley
Organization: Reliability Analysis Center
Date Posted: Wed Jan 6 10:46:14 US/Eastern 1999
Expanded temperature differences, both high and low, can cause reduced reliability values especially when temperature cycling is applied. A fatigue failure equation to estimate the reliability of a surface mount solder joint is contained in the MIL-HDBK-217 " Reliability Prediction of Electronic Equipment". This equation accounts for delta temperature differences, cycling rates and circuit board substrate thermal coefficient of expansion changes. The greater the delta temperature or cycling rate, the lower the reliability expected. If screening is the process that you want to examine, then MIL-HDBK-344 "Environmental Stress Screening of Electronic Equipments" has techniques for thermal and vibration testing conditions at the assembly level. For more information on solder reliability, a book by John Lau "Solder Joint Reliability Theory and Application" should be read.
Subject: Solder Failures and Fatigue
Reply Posted by: Len Stascavage
Naval Surface Warfare Center
Date Posted: Mon Jan 11 14:06:02 US/Eastern 1999
I don't know if this will help you, but there is an interesting article, "Temperature Cycling and Thermal Shock Failure Rate Modeling", Page 10 in the 1997 Reliability Physics Symposium Proceedings. If the article doesn't help, you might fing the references of use.
Subject: Screening solder workmanship
Reply Posted by: Bill Lieb
)Naval Surface Warfare Center
Date Posted: Fri Feb 5 14:43:56 US/Eastern 1999
No, I can not quantify the difference in screening results if the temperatures had been slightly more extreme. But I can say this. First, you are correct in the assumption that the difference would be minimal. But what is more important is that you are not stressing the hardware nearly enough to precipitate workmanship defects in soldered connections. Your temp range is way too small. I can't speak to your temp rate of change since you didn't mention it, but the temp rate of change is AT LEAST as important as the temp range. I need to work in degrees C: your numbers -30F to +60F equate to about -34C to +16C. So your temp range was about 50C. This is too benign to surface defects in soldered electronic assemblies. We in the Navy like a temp delta of around 120C (somewhat less or more would also be OK) with the extremes normally something like -40C and +80C and the temp rate of change something like 10C/minute. Rate of change less than 5C/minute is worthless and greater than 15C/minute can be bad for the hardware. And all of those temps are to be the hardware temps, not just the chamber air temp. Of course, once you correlate the chamber air temp to the hardware temp, you can monitor the chamber air temp (much easier than measuring the hardware temp) "knowing" that air temp of X equates to hardware temp of Y. Hope this helps.