Algorithm for cumulative damage probability calculations in S-on-1 laser damage testing

摘要

In the determination of the laser-induced damage threshold of optics, the correct estimation of damage probabilities is essential. In this publication, a simple procedure based on physical considerations is proposed to optimize the calculation of the damage probabilities by using a cumulative algorithm. The predicted status of test sites at higher and lower fluences than actually tested provides the basis for a new data reduction. It is shown that the proposed algorithm increases the statistically relevant amount of data per fluence interval dQ_i by using virtual test sites. Thus, the uncertainty σ_i in the calculation of the damage probabilities is reduced significantly and the subsequent linear regression of the damage probabilities will have a reduced least squared error. Non-linear regression of the damage probability according to defect-induced damage models as published numerous times in recent years are also performed to utilize a better confidence level which will be shown exemplarily.