Estimation of multimode pump ensemble reliability using Monte Carlo simulation to account for derating and variable stress profiles

Abstract

High power optical amplifiers (HPOAs) are utilized in many high reliability applications. The need for greater bandwidth in flight and space missions is driving the development of free space optical transmission systems which require high reliability HPOAs over mission lifetimes which may exceed 10 years. One challenge in developing such models is that typically when designing HPOAs, redundancy and derating approaches are utilized to augment reliability and avoid single points of failure. In these designs, redundant components experience a stress profile which depends on statistical failure probabilities of its sister components. For example, in the case of HPOA gain stages, typically several multimode pumps are combined to pump a gain fiber. These multimode pumps are typically run derated when the mission begins, but in the event of a component failure, the power to each remaining operational pump is increased to maintain constant output. As such it is more important to characterize when the entire ensemble of pumps fails. Accurate models for ensemble reliability require an approach which accounts for time and stress profile dependent failure rates which are hard to access experimentally (due to high component reliability) or analytically. In order to address this problem, we have utilized a Monte Carlo simulation approach which can rapidly simulate ensemble failures for any given stress profile. By running several simulations, we are able to build up the failure function for the ensemble, thus providing a more reliable model for failure rate for the ensemble. This failure model can be used to build a more accurate picture for HPOA reliability.