A GPU-Accelerated Compressible RANS Solver for Fluid-Structure Interaction Simulations in Turbomachinery

Computational Fluid Dynamics (CFD) is a fundamental tool fo the aerodynamic development in industrial applications. In th usual approach structural deformation due to aerodynamic an thermal loads is often neglected. However, in some cases, wher power efficiency is the ultimate goal, an accurate prediction o the structure-flow interaction is essential. This is particularl true for trim and flutter analysis of aircrafts, helicopter and turbomachiner blades. Particularly, turbomachinery trim and flutte predictions still represent a challenge due to phenomena lik rotor-stator interaction, separations and shock waves. The usua time-linearised, frequency-domain strategies can be inadequat when this kind of strong non-linear phenomena occur in the flow making necessary full non-linear time-domain simulations or th harmonic balance technique Beside flutter, another important aspect, not yet adequatel investigated, is the trim analysis, which is fundamental for an accurat steady simulation that aims to consider static blade elasticit for the performance evaluation of turbomachines Moreover, alongside the obvious contribution given by centrifuga loads to the blade deformation, a not less importan source of blade displacement is the thermal effect due to the hea exchanged between the solid and the fluid domains. In particular for some geometries and operating conditions, thermal effect can be more important than centrifugal effects for the blad deformations Considering multiple sources of blade deformation (elastic centrifugal and thermal) in a what is often called "multiphysics approach is nowadays more and more important, i the goal of the analysis is geometry optimization. To achiev this, next to result's accuracy also computational efficiency i required, when hundreds of aeroelastic simulations have to b performed in a typical optimization loop. Modern GPUS can b exploited to pursue this goal thanks to their high peak computationa power available at relatively low costs and low powe consumption with respect to the usual CPUs In this paper a pioneer work describing the impact of stati deformation due to blade elasticity, thermal and centrifugal effect on the performances and power efficiency will be provided Alongside with accurate results, computational efficienc is taken into account. The purpose of this article is to sho the architecture of a GPU-Accelerated Fluid-Structure Interactio (FSI) solver for compressible viscous flows. The propose approach is validated with a typical industrial case, i.e. a turbocharge transonic centrifugal-compressor provided by ABB The effects of trimmed solutions on the most important integra quantities (i.e. mass flow, characteristic curves, mass-Average outflow profiles) are investigated and a comparison with pur aerodynamic results is provided. Due to the high blade stiffnes and thus the very small displacements obtained with the trim solutions for the particular case presented in the paper the aeroelasti solutions basically provide nearly the same results as th pure aerodynamic solutions