Thank you very much for commenting, I am happy that you found the article interesting. My intention was indeed to raise awareness on the possible change of the TIM parameters in an application. If the stability of temperature sensitive parameters is a design goal, a highly stable TIM material should be selected, or perhaps the low thermal resistance should be sacrificed by not using TIM at all.
I was not aware that there was such a significant change in TIM performance over time. I had erroneously assumed that this is relatively constant (which it is not). As the article states, this effect can have a significant impact upon reliablity as the number of cycles increases and is an important consideration for product design for reliability.
This is a useful and informative writeup. I had not considered all of those possible thermal interface materials, so that was educational as well.
What was possibly beyond the scope of this article is a discussion of the mechanism of change in thermal conductivity of the interface. That information would lead ti a good deal od insite on the whole topic, and move us toward the realm of "expert."
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.