Cast-in-Pair Blade Release Simulation and Comparison to Experiments With a Full Scale Rig
This paper deals with the simulation of a Low Pressure Turbine (LPT) blade release and the subsequent impact onto the casing. The blade design is a cast-in-pair novel configuration where two airfoils are attached to a single firtree. This design solution was implemented in the first stage of the Low Pressure Turbine of the TP400-D6 engine using a directionally solidified M247 alloy. Due to the novelty of the design configuration a case containment test was carried out for certification versus CS-E810 EASA airworthiness requirement. The test was full scale reproducing engine-operating temperature and red line rotational speed. The objective of the present work is the simulation of the blade release failure event including the interaction with the trailing blades and the severe impact onto the casing. The simulation is compared and validated against the results of the full-scale rig. Material behaviour under impact working condition is characterised through Hopkinson bar tests, ballistic tests and triaxial traction tests. Johnson-Cook constitutive and failure material models are generated for the blade casting, the case forge material and the shroud seal segment. This material model is a temperature and strain-rate dependant flow stress model. Containment simulation is carried out with LS-DYNA implicit/explicit solver. Simulation work includes meshing and boundary conditions effects on model fidelity. Case damage progression is linked to each of the different airfoil portions as contact occurs. The friction contribution to the case and blades interface and to the case final form is also revised. In addition, the importance of the trailing blades and the interaction with the released blade is assessed. Interaction effect comprises both the resulting damage between airfoils and the change in primary release trajectory. Finally, simulation results are compared to the full scale rig results. Case-impacted final geometry and internal energy from simulation are checked against test evidences.