Monte Carlo analysis of ISO and raster scan laser damage protocols

Abstract

The typical measurement error reported for laser damage tests is the fluence uncertainty due to inaccuracies in measuring the laser beam energy and its diameter. However, the inherent uncertainty of the testing protocol should also be included in the reported laser damage threshold error bars. Underestimating measurement errors can lead to false conclusions about the impact of process changes on laser damage resistance. In this study, four different laser damage precursor fluence distributions were created from randomly generated numbers and then evaluated using the ISO and raster scan laser damage test protocols to determine a laser damage threshold. Measurement errors are determined for flat top test beams for multiple cases. To add real world relevance, the impact of Gaussian test beams with beam pointing instability was modeled for the lowest accuracy laser damage precursor distribution. The impact of damage test area compared to optic dimension is also examined. The measurement error for the raster scan test ranged from 8% to 24% depending on the test beam spatial profile (flat top or Gaussian) and beam pointing stability. ISO measurement errors ranged from 4% to 250% for a simulated 10 J/cm2 test and was much more sensitive to the laser damage precursor distribution as well as the spatial profile and pointing of the test beam. Both testing protocols poorly predicted the laser damage resistance of large areas with Gaussian precursor laser damage distributions.