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: Sam
Date posted: Mon Jul 31 18:37:43 US/Eastern 2000
Subject: Coffin-Manson Eq.
We're using Coffin-Manson equation to calculate the acceleration factor (AF) of solder joint (failure mechanism is thermal fatigue). The AF is only related to temperature range. The question is: if the temperature transition rate affects the AF? If so, what's the equation? In addition, are there any other equations to calculate the AF for the above condition? Thanks in advance! Sam.
Subject: Coffin-Manson Relationship
Reply Posted by: Bruce Dudley
Organization: Reliability Analysis Center
Date Posted: Tue Aug 1 10:06:08 US/Eastern 2000
The Coffin-Manson relationship is the application of the inverse power law used to model fatigue failure of metal that is subjected to thermal cycling. The delta temperature, difference from high to low, is the significant parameter in the acceleration equation as it is raised by the expected acceleration factor (usually 2 or greater for solder). The rate of temperature change is not relevant to the acceleration factor, it only means that more or less cycles can be performed in a shorter or longer test time frame. You should note that the acceleration factor exponent(fatigue characteristic of a particular material) is not the same for all materials and damage mechanisms. Different values for the same or similar materials (solder with different percents of material) may occur which is why a series of tests should be performed to verify the assumed acceleration factor. Material constants are often thought of as exact and absolute values, but the reality is they are average parameters and could have wide distributions, which is why testing, testing and more testing is necessary. The final thought is, accelerated test models should be treated as estimates rather than exact solutions.