ANDREW HEALEY / FORT WORTH
When Bell Agusta's BA609 made its first flight on 7 March, every conceivable manoeuvre had already been simulated many times in Bell's test rigs
That the first-ever flight of a civil tiltrotor went smoothly came as no surprise to those involved in the slow-burning project. Bell Agusta Aerospace's BA609 had already repeatedly been through every conceivable manoeuvre and permutation of circumstances in today's equivalent of an "iron bird" test rig.
The vehicle management system integration laboratory (VMSIL) at Bell's flight research centre in Arlington, Texas, brings together four systems test rigs - electrical, hydraulic, avionics and flight controls. Final hardware-in-loop testing of the tiltrotor's fly-by-wire flight controls in the VMSIL, completed at the end of February, was the final step before first flight of the BA609.
"The only real difference between the BA609's systems and those of, say, the Boeing 777 or the Joint Strike Fighter, is that our aircraft incorporates both helicopter and aeroplane control mechanisms," says Cliff Harrell, flight controls lead on the BA609 integrated project team. "Over the past two years we have ironed out most of the system integration problems and, in the process, saved ourselves several months of ground testing. 'Aircraft zero' is now as close as we can get to the finished article."
Aircraft zero - the VMSIL - looks nothing like the finished article, however. In place of a fuselage and wing, the BA609's systems are spread across three separate areas in the laboratory and interfaced with a flight-simulation host computer. A separate cockpit rig features dual flight controls and instruments, the pilots facing a pair of large flat-panel displays, but "flying" the VMSIL is not a simulated affair. Lifting the power lever, trimming the cyclic or altering the angle of the proprotor nacelles - via a thumb switch on the power lever - activates the same transducers, computers, displays, electrical systems and hydraulic actuators that are installed in the actual aircraft. Unseen by the pilot, the laboratory becomes filled with movements big and small.
The BA609's VMS employs triplex fly-by-wire flight and engine control systems, redundant electrical and hydraulic power-generating systems, and dual-pilot integrated glass-cockpit avionics. In the VMSIL, this equipment operates in closed-loop mode. Actuator positions are sensed and fed into high-fidelity mathematical models of the aircraft, rotors and propulsion system.
Aerodynamic loads acting on the rotor and control surfaces are computed by another mathematical model and applied to the actuator rod ends by a hydraulic rig. Aircraft variables are output from the model and fed back into the flight control computers through sensor simulations.
Hardware-in-loop simulation (HILS) was the final phase of VMSIL testing before first flight. During HILS testing, the only systems simulated in the VMSIL were the engines, rotors and air-data and inertial sensors. Within the laboratory, the largest area is dedicated to the conversion and swashplate actuator rigs, where actual interconnect drives, hydraulic power drive units and ballscrew actuators move heavy steel weights representing the nacelles and rotors. Data from the various rigs is fed to the cockpit simulator's flight displays and visual system.
Performance testing
To the extent that laboratory constraints allow, flightworthy hardware and software has been used throughout the VMSIL. In some areas it deviates from the aircraft design drawings, but these discrepancies have been analysed and allowed for, to ensure the proper performance characteristics are maintained. As a result, the simulation can compute and deliver the correct aircraft response to injected system failures. Engineers can initiate, say, a single engine shut-down followed by a dual hydraulics failure. The pilots, after dealing with the resulting emergency, can suggest handling and control feedback tweaks that can be accomplished within the VMS software and copied across to the real thing, waiting 30m (100ft) away in the hangar.
There were two main phases to the HILS test plan. In the first phase, engineers verified the functional characteristics of each system when operating in the closed-loop environment. These tests included interface control document verification and functional testing of AC and DC power, together with testing designed to verify mathematical model-to-VMS integration by comparing rig flight characteristics with specified performance requirements. VMSIL characteristics were also compared with those of the BA609 flight simulator.
For the second phase, over 200 failure scenarios were evaluated, each covering a series of system and equipment failure combinations. Engineers could look for transient exceedences on associated systems. These involved implementing around 1,100 software test cases - 30 verifying the redundancy management of the proprotor collective-pitch actuators alone. Not all were completed by first flight; some maintenance and training functions could be left until later.
Flight performance predictions are highly sensitive to the fidelity of the BA609 airframe, rotor and engine models. To this end, flight characteristics were repeatedly refined from data acquired during earlier tests of the XV-15 and V-22 tiltrotors, as well as data from BA609 windtunnel and rotor tests and test cell runs of the aircraft's Pratt & Whitney Canada PT6C-67A engines.
As a result, the tiltrotor's handling qualities were rated satisfactory, both for first flight and envelope expansion. Satisfactory is defined as "full performance criteria met with routine pilot effort and attention" and is the highest possible endorsement.
"HILS helped us meet three goals," says Harrell. "It showed us that the interfaces between each subsystem work properly. It thoroughly exercised the VMS hardware in a closed-loop environment, thus mitigating flight risk. Finally, it helped us assess and quantify failure conditions that would have been too risky to simulate in flight. These assessments will eventually be used to show compliance with specific BA609 certification requirements."
The VMSIL could not answer all of Bell Agusta's questions. Until the aircraft lifted off on 7 March, the team did not know for sure the exact take-off power required. "During the tie-down runs we were getting a few percent of engine power more than we predicted, and that could equate to an extra passenger under some range and payload conditions," Harrell says.
Source: Flight International
