The environment at the component level affects the inherent failure rate of the component.

The environment at the assembly level can expose defects introduced somewhere in the PCA manufacturing process. As stated in MIL-HDBK-344A (Environmental Screening Strength)

“Manufacturing techniques for modern electronic hardware consists of hundreds of individual operations and processes through which defects can be introduced into the product”.

Note that the way in which the ‘environment’ is included in the component models is completely different — and independent from – the way in which it is included at the assembly or system level. Both are necessary as they relate to different sources of defects.

PI_E in the system model is derived from MIL-HDBK-344A. Some assumptions must be made (e.g., the detection and precipitation efficiencies are assumed to have a combined effect of 0.855 — as is typical in MIL-HDBK-344A). Ideally, if this were a testing scenario, the ‘number of cycles’ should also be used. However, in an actual product application, the number of cycles experienced varies with time. This would cause PI E to be a non-linear function of time (cycles). Obviously, that is not workable in a prediction of a constant failure rate. Therefore, the 217Plus implementation of PI_E assumes 10 cycles in order to provide a reasonable, time invariant factor that reflects the experience data that went into the development of the 217Plus models.

Paul Wagner
Senior Engineer
ph: 315 351-4227