By Fred Schenkelberg
When you ‘pass’ a standard based test, just what does that mean? How can you use test results in a meaningful way other than just noting the product ‘passed’? Understanding the failure mechanisms at play along with the statistics is key.
Military standards, GEIA, IEEE, ANSI, etc provide detailed test procedures for a wide variety of situations. I have yet to find a standard test procedure that details what specific materials and failure mechanisms the test is applicable. One might exist, I’ve just missed it.
Just because 7 prototypes survived 168 hours of 85%RH at 85°C and thus ‘passed’ the test, does that mean anything useful as we attempt to determine if our product is reliable or not? Of course, if one or more prototype fails and we don’t ‘pass’ the test, what does that mean? What have we learned from such testing?
Let’s explore the use of testing based on a standard. Sure, it often is required by customers and commonplace in our industries, so let’s understand what passing/failing actually suggest about our reliability. Also, when purchasing a piece of equipment and it lists 15 standards that it meets, here are a few questions to ask about that testing.
All standards are flawed, some are useful (to restated a popular phrase about modeling). Let’s sort out how to glean the useful elements and avoid the pitfalls. The use of standards facilitates communication between organizations. It is only useful if all concerned fully understand the meaning of the results.