<<< Back to Index
This Month's Article >>>

How do I use Weibull++ 8's Difference Detection Matrix to Compare the Reliability Between Product Designs?



Engineers often need to design tests to compare the reliability between two or more product designs. To do this, they must determine the sample size and test duration needed to show that one design is more reliable than the other. For simple cases of comparing two product designs, Weibull++ 8 includes a tool called the Difference Detection Matrix, which shows what test durations are required given different assumptions about the reliability of the compared products.

To access the Difference Detection Matrix, choose Insert > Tools > Test Design. In the window that appears, select Difference Detection Matrix.

To use the tool, you must first provide all the required inputs on the control panel, including sample size and life distribution for each product, the reliability metric to be compared (i.e., mean life or BX% life) and the maximum test time available. As an example, suppose two products will be tested using sample sizes of 10. Furthermore, suppose mean life is the metric of comparison and the goal is to demonstrate, with 90% confidence, that one design is more reliable than the other. A simple matrix based on these inputs is shown next.

Difference Detection Matrix

You can read the matrix to determine whether it is possible to detect a difference between the two designs and, if so, how many hours of testing would be required with the specified sample size. For example, if Design 1 has a mean life of 1,500 hours (third column) and Design 2 has a mean life of 2,500 hours (fifth row) (last two rows), then 10 units of each design must be tested for 3,000 hours (as indicated by the index 3) to demonstrate, with 90% confidence, that Design 2 is more reliable.