The Fly-By-Light Advanced Systems Hardware (FLASH) program developed Fly-By-Light (FBL) and Power-By-Wire (PBW) technologies for military and commercial aircraft. FLASH is a Defense Advanced Research Project Agency (DARPA) Technology Reinvestment Program (TRP) consisting of three tasks. Task 3 of the TRP, Fly-By-Light Actuator Development, developed an advanced smart, rotary thin wing and two 20 horsepower (52 corner horsepower) electrohydrostatic Actuators (EHA). This paper summarizes the results of McDonnell Douglas Aerospace (MDA) modeling, analysis, testing, and system demonstrations performed under Task 3 of the DARPA agreement for the EHAs developed.
EHA systems for the FLASH program were developed by two suppliers to meet the high power F-15 stabilator actuator performance requirements. One EHA included an AS-1773A optical data bus, and an optical position transducer for position feedback. These items created an EHA system utilizing complete closed loop FBL control.
Complementary modeling and analysis was performed which consisted of detailed dynamic performance and thermal models. The dynamic performance model provided data to design and initially train a neural network used to demonstrate on-board fault diagnostics. The software application employed to develop the dynamic performance model included an autocoder feature that generated C++ code for outer position loop closure. The thermal model was integrated with an MDA airframe thermal analysis program using representative flight data to predict actuator temperatures during a typical tactical aircraft mission.
A comprehensive test program was conducted to fully characterize EHA performance and validate the dynamic performance and thermal models. Data analysis was performed so that issues such as time dependent, nonlinear stiffness effects on flutter margin, acceleration capability vs. requirements, and thermal characteristics were understood.
EHA hardware demonstrations were conducted proving the viability of electrically powered actuators, Fly-By-Light control, and neural net based diagnostics.