Explicit Dynamics Simulation of Ejection Sequence and Analysis of Seat Support Structure for F-15 Aircraft

In early 2005, the F-15 rocket sled at the Goodrich Hurricane Mesa test facility experienced a failure of the ejection seat guide rails. The seat rails were the original parts delivered with the F-15A forward fuselage before the unit was decommissioned and converted into a rocket sled. In its life as a sled, the rails had experienced decades of ejections without any problem. Initially a fatigue failure was suspected and the rails were replaced. On the next sled run, the failure occurred on this replacement set of rails. At this point The Boeing Company was contacted to investigate the rail failures. The rails failed during the qualification runs for the CKU-5C rocket catapult on the ACES II ejection seat in a B-1 configuration with leg restraints. The failures occurred at the edge of the loading envelope with a heavy aircrew, an active leg restraint system, and a high ejection speed. The purpose of this analysis was to simulate the ejection of the seat from of the rails and to determine the sensitivity of the rail strength to several parameters, including aircrew size, weight, and center of gravity, aircraft speed at ejection, the type of rocket catapult used, and the effect of aircrew leg restraints. The methodology used for this analysis was an ABAQUS/Explicit finite element simulation. This transient dynamic simulation accounts for load sequencing effects that have not been possible to consider with the simplified quasistatic analysis methodology that has been used in the past. Two models were created, a system level model which includes the aircrew, seat, and leg restraint lanyards, and a local detailed solid model of the top of the rail. Stresses in the area of the failure were not measured in the rocket sled tests, but the data that was recorded during the initial tests was used to develop the applied loads to the system and also to correlate the finite element model results. Analysis indicates that the lower ejection velocity of the seat with the CKU-5C combined with the leg restraints allowed the aerodynamic load to build up to a level high enough to fail the rails for this worst case condition. Small reductions in airspeed were found to produce large reductions in rail stresses. Comparison of the seat rail stresses from the model with a later sled test was excellent.