Lower and upper tail comparisons can be useful for evaluating the amount of acceleration achieved. For example, you can limit the sampling of data from the strength distribution to the lower tail while limiting the sampling of data from the applied load distribution to the higher tail.

By probabilistically simulating interference from the partial and full distributions actually tested, an acceleration factor can be estimated by evaluating occurrence probabilities:

Where:

Ft= The probability of failure by accelerated test (partial distribution, tail comparison)

Ff= The probability of failure simulated by full distribution interference.

Fp= The probability of failure simulated by partial distribution interference.

Fs= The probability of failure (not accelerated) expected in service.

Rs= The reliability expected in service.