Perpustakaan Digital ITB

This thesis employs high-resolution QAR data to forecast aerodynamic stall and in-flight loss of control probability during the final approach phase of Boeing 747-300s. It presents a modeling framework as it calculates the aircraft's total mechanical energy based on the relationships between physical force and motion, comparing measured and simulated energy trajectories throughout a 95-s interval preceding touchdown. Subset simulation utilizing Markov Chain Monte Carlo (MCMC) sampling is implemented to evaluate the likelihood of the failure event, defined by the aircraft's energy dropping below a threshold determined by impaired conditions. The baseline overall failure probability is assessed to be around 3.57 × 10??; while the dynamic, per-second evolution is estimated to be an exponential decay (of base 10) towards zero. Two sensitivity analyses are conducted to evaluate the influences of parameters, specifically angle of attack, crosswind speed, and N1 fan speed, which are the primary factors contributing to the failure event. A mitigation plan is formulated by increasing approach speed by 10 knots while decreasing pitch and angle of attack, resulting in a 45.56% reduction in the probability of LOC-I.