Probabilistic Risk Analysis of Aircraft Self-Collisions: A Case Study

Abstract

Airborne self-collisions occur primarily in military aircraft because of external stores and are frequently experienced by personnel operating these aircraft. In most cases, objects causing self-collisions are irregularly shaped and unstable. Consequently, the trajectories of these objects are uncertain. A framework for the probabilistic risk analysis of aircraft self-collisions is proposed in this study. Based on the probabilistic trajectory prediction model, methods for estimating the probability of collision (POC) and the corresponding risks were developed. Subsequently, a self-collision event involving an ejected gun shell was analyzed as a case study. A model considering random shell rotation, which continuously changes the drag characteristics and trajectories, was developed. Other uncertain factors associated with the aircraft and shell cases were considered. The most influential factors were selected based on the sensitivity analysis and were then used to calibrate the likelihood of the event using historical data. A Monte Carlo simulation, in conjunction with the probabilistic ballistic model, was performed to evaluate the POC. The POC was used to reflect the risk of engine failure up to the operational limit. The calculated risk indices were objective functions used for the design or operation optimization. Various risk measures were evaluated to reduce the incidence of failure and extend the aircraft’s flight envelope.