In the past, Sikorsky had to perform expensive bird strike test programs to certify helicopter components. These tests are expensive because they require building prototype components and testing them on a very elaborate rig. For example, the rig used to test the ability of a tail rotor to withstand bird strikes uses a hydrostatic drive system transmission to provide accurate speed control in driving the rotor and a gas cannon to fire gelatin projectiles.
To optimize component design and reduce their physical testing process, Sikorsky started using MSC’s explicit finite element analysis software, Dytran.
First, they created a finite element model consisting of beam elements for the flexbeam, torque tube and inboard rotor blade and with plate and hex elements for the outboard rotor blade at the bird impact point. The composite flexbeam makes it possible to change the pitch of the flap and lag direction without the need for moving parts. The torque tube is used to provide torsional rigidity in the pitch axis.
The structural model was constructed with Lagrangian finite elements. Two types of bird models were used in the analysis; a Lagrangian bird consisting of hex elements with the density and bulk modulus of water and an Eulerian fluid flowing through a finite difference mesh. The key difference between them is that the Largrangian description follows one particular particle while the Eulerian description of motion observes a parcel of material though space and time.
For the Lagrangian bird technique the bird elements are given an initial velocity that causes them to impact the structure FE model. Master-slave contact interfaces apply bird element loads to the structure. Using the Eulerian technique, the bird elements apply loads to the structure elements using Dytran’s general coupling option. The Eulerian bird model gave better correlation to the test data than the water bird model.
“The correlation between test and analysis was close enough to enable the successful design of rotors for bird strike. The calculated moments on the flexbeam ranged from right on to twice the test values while the calculated moments for the blades were to 2/3 of the test values. This level of accuracy is acceptable for component design since the rotor blade is normally much stronger than required for the bird strike load while the flexbeam is closer to being critical. Overpredicting the flexbeam loads will only make it stronger than necessary and underpredicting the blade movements can be adjusted for by maintaining a minimum margin of safety of 2.0 for the blade design” said Alan Dobyns, Senior Engineer at Sikorsky.