What do I need to know before analyzing data from an accelerated life test with more than one type of stress in ALTA?
Before analyzing data with more than one stress type (e.g. temperature and humidity), there must be an “overlap” of the stress levels in order to be able to fit a multiple stress life-stress relationship. This is because a relationship between the stress levels and the failure behavior of the product needs to be established. For example, consider a situation where Stress 1 has levels A and B and Stress 2 has levels X and Y. If the failure information is only collected where the stress levels are at A and X and at B and Y, a mathematical relationship between the two stress types and reliability cannot be obtained. The same failure data will have to be collected at a stress combination of A and Y or B and X in order to obtain a solution for the data set. If you attempt to analyze a non-cross-referenced data set in ALTA with a multiple stress model (such as the temperature-humidity and temperature-nonthermal models in ALTA 7 and the generalized Eyring, proportional hazards, general log-li