Simulation software models flight behavior

The Saab Skeldar V200 is a rotary-winged unmanned aerial vehicle (UAV) in a market dominated by fixed-wing aircraft. The Skeldar does not require runways and can hover in one position. Designed for land- and sea-based patrol, light transport, electronic warfare, and surveillance applications, the 4.0m x 1.3m x 1.2m UAV flies at speeds up to 130km/h with a range of 150km.
 

Challenge

Early in the Skeldar development process, problems arose in the prototype’s stability. The prototype had been developed with the aid of mathematical models, but the prototype behavior demonstrated that these models did not fully capture the UAV’s flight behavior. Accurate simulation of the flight behavior of a rotary-wing aircraft requires the ability to capture the interactions between the lifting forces on the rotor blades and downwash, which is the air deflected by the blade in producing lift. To the best of MSC’s and Saab’s knowledge, no simulation had ever managed to address this issue.

Validation, solution

Dr. Per Persson, technical fellow, structural dynamics for Saab, decided to use MSC Software’s Adams simulation to model Skeldar’s flight behavior. Dr. Per Weinerfelt, also a technical fellow at Saab, provided support regarding aerodynamics and inflow modelling. Persson imported a structural model of the helicopter into Adams. Researchers modeled the two rotor blades as eight flexible bodies in MSC Nastran and incorporated their modal representations into the Adams model. Dividing the rotor blades into smaller segments makes it possible for the rigid body motion of the outer part of the rotor blade to apply forces to the inner part of the blade to more accurately model the deformation of the blade during flight. Each segment contains about 25 beam elements with varying characteristics. The flight study was focused on the rotor system, so the main helicopter frame is represented simply as a rigid body.

The aerodynamic forces and moments acting on the UAV are calculated by a model that is implemented as a user-defined function (UDF) in Adams. The actuator motion of the rotor blades and helicopter frame provide input to the aerodynamic model. Aerodynamic forces are computed from the blade motion and applied at different positions along the blade. A square plate analogy computed drag force on the helicopter frame. Deformation and motion of the blades contribute to the angle of attack calculation. The lifting line method calculated lift and drag forces and the Peters-He inflow model captured the highly nonlinear effects of downwash.

The force required to create the downwash is equal in magnitude and opposite in direction to the lift force on the airfoil. Persson and Weinerfelt fed the lifting force and downwash back into the model.

A state-space representation of the actual flying control system equations controlled the simulated aircraft. The state-space system includes position error feedback, time-integrated position error, and time-derivative position feedback. Since the UAV’s motion also depends on its attitude, the control loop is fed by attitude and attitude rates. Applying rotation to the main rotor shaft drives the model. Tail rotor control prevents the UAV from spinning in reaction to the forces applied to the rotor. The motion of the rotor blades results in applied external aerodynamic loads. That caused the helicopter to move and the control system uses this motion as input to control the rotation of the main and tail rotors.

An extensive process validated the model. Static modal measurements on the blade matched up with the numerical model. Flap moments, measured in rotor rig tests, were compared to the model. The overall shape of the flapping motion was well captured by the model. The flight behavior of the model was correlated to flight tests by feeding actual flight control input from a test case into the simulation model. The model duplicated the pitch and roll response of the prototype, even though the wind during the actual flight was not known. For more controlled tests, in no wind using smaller control inputs, the response of the model compared even better to measured data.
 

Results

After validating the model, Persson applied it to the issue that had been experienced with the prototype and discovered that the simulation model accurately duplicated the problematic behavior. The simulation model provided far more detailed information than could be obtained by instrumenting the prototype, such as the aerodynamic forces acting on each section of the blades. The model also made it possible to evaluate the performance of the UAV under a much wider range of conditions than could ever be evaluated with the prototype due to the time, cost, and risks involved in actual test flights. The simulation results helped Persson and other Saab engineers diagnose the cause of the stability problem and develop a solution. After the model was updated to evaluate this proposed fix, the problem disappeared in the simulation. At this point, the same change was made to the prototype, and test flights showed that the problem had indeed been solved.

“The Adams simulation saved us at least six months that would have otherwise been required to solve the problem by modifying and testing the prototype,” Persson concludes.
 

MSC Software
www.mscsoftware.com

Saab Aeronautics
www.saabgroup.com