Related Topics:

Test Design Assistant

Difference Detection Matrix

The Difference Detection Matrix calculates how much test time is required before it is possible to detect a statistically significant difference in the mean life or BX% life (e.g., B10 life) of two product designs by analyzing the data from a reliability life test. You can use this tool to evaluate different test plans in order to choose the one that will be most efficient to compare the reliability of the two designs. To use the matrix, you must provide various inputs—including the sample sizes that will be used to test each design and the life distributions that describe each design’s life behavior.

The ReliaWiki resource portal has more information on the Difference Detection Matrix at: http://www.ReliaWiki.org/index.php/Reliability Test Design.

Reading the Detection Matrix

In this section, we will look at an example detection matrix and discuss how to read it. The next section explains how to set up the matrix. We will assume that a sample size of 10 will be used to test both product designs.

Because the column title in the above matrix is “Design 1 Mean Life,” we know that the metric being compared is mean life (rather than BX% life). The column headers represent possible mean life values for Design 1, and the row headers represent possible mean life values for Design 2. We’ll begin by examining the cell in the second column and the first row of the matrix, which corresponds to a Design 1 mean life of 1,000 hours and a Design 2 mean life of 500 hours.

The number inside this cell is 4, and the legend below the matrix tells us that a 4 means a difference between the two designs can be detected, but only after at least 1,500 hours of testing. In other words, one thing the matrix tells us is that if Design 1 has a mean life of 1,000 hours and Design 2 has a mean life of 500 hours, then the tests we use to compare these designs will need a duration of at least 1,500 hours in order for the analysis of collected data to show that there is, at the specified confidence level, a statistically significant difference in mean life between the two designs.

Tip: You can click a cell in the matrix to see the difference in the compared metric that was calculated for both designs based on the associated test time. For example, if a cell in the matrix shows that a difference in mean life can be detected with 2,000 hours of testing, clicking the cell will show the mean life calculations (including two-sided confidence intervals) that would be obtained for each design if you analyzed the test data after 2,000 hours of testing. (If only one of the designs would produce enough failures to fit to a distribution during the given test time, no intervals will be shown.)

Setting Up the Detection Matrix

Follow the steps outlined below to set up the detection matrix. All the settings are located on the control panel.

Your selected color will be used to indicate when the shortest test time provided in the Test Time Matrix Setup area is sufficient to detect a difference in the specified life characteristic. Darker versions of that color will be used to indicate when longer test times are required.

 

© 1992-2015. ReliaSoft Corporation. ALL RIGHTS RESERVED.