Investigation and Evaluation of the Volvo Aero Tool Volvane for Outlet Guide Vane Design

Abstract: One of the main functions of the turbine rear frame, located in the aft part of the engine, is to remove theangular component of the outgoing flow from the low pressure turbine. The angular component in the outgoingflow is due to load on the low pressure turbine. The most interesting part of the turbine rear frame from anaerodynamic point of view is the gas channel consisting of outlet guide vanes, a hub, and a shroud. Volvane isa powerful in-house design tool used to create guide vanes and includes a 2D panel which is a potential flowsolver and a G3D Euler solver. Minimizing the number of design parameters is a main issue and may result in asignificant improvement in the outlet guide vane design process with Volvane. Initially a reference design of anoutlet guide vane is created in Volvane. The design is simulated in Volvane and with different RANS-solversand computational grids. After analysis of the ow results, a RANS-solver is selected for further use in the investigation. In Part 1, the design tool validation is the target to reduce the number of design parameters.Three different design of experiments (DoEs) are created. Design parameters and boundary conditions werevaried to investigate how the flow behaves around the outlet guide vanes for different geometries and operatingconditions. The same changes were simulated in RANS-solvers and the results compared with those of theVolvane program. Analysis in Part 1 resulted in updates of new V olvane versions. The V olvane G3D Eulersolver has more difficulties in predicting changes in lean compared to changes in sweep and is more able topredict an increased than a decreased sweep angle. If the boundary condition of the swirl is increased, theV olvane G3D Euler solver has difficulties in predicting the diffusion factor. Simulations with wall functionmesh in the selected RANS-solver showed a 2% outlet swirl angle difference from the low Reynolds mesh.This aspect should be taken into account when analyzing results from flow simulations. The loss estimationsfrom the V olvane G3D Euler solver are approximately 35% of the loss estimations from the RANS-solver.Generally the V olvane G3D Euler solver and the RANS-solver are in good agreement.In Part 2, the design process rationalization, response surface methodology (RSM) is created to find correlationsbetween design and output parameters. A statistical software is the main tool used. The correlations measurethe relations between different design and output parameters. Two additional DoEs are constructed. Theanalysis suggests the maximum thickness location of the vane can be shifted without affecting the axial positionof maximum velocity. The correlations are unaffected when the parameter bounds are increased.