Modelling Particulate Deposition in Gas Turbines Part I: a Non-Spherical Bounce Stick Model

Abstract The secondary air system and cooling passages of gas turbine components are prone to blockage from sand and dust. Prediction of deposition requires accurate models of particle transport and thermo-mechanical interaction with walls. Bounce stick models predict whether a particle will bounce or stick upon impact and calculate rebound trajectories if applicable. This paper proposes a bounce stick model that uses analytical solutions of adhesion, plastic deformation and viscoelasticity to time-resolve collision physics. The Discrete-Element Methods (DEM) model shows good agreement when compared to experimental studies and FEA simulations, requiring minimal parametric fitting. Further comparison is made to the best resolved and industry standard models available. In addition to coefficients of restitution, other variables crucial to modelling rebound, for example angular velocity, are predicted. The time-stepping explicit approach allows coupling between internal processes during contact and shows that viscoelasticity plays a significant role in adhesion. Modelling time-dependent internal variables like wall-normal force allow the modelling of non-spherical particles, the physics of which has been shown by previous work to differ significantly from that of spheres. These effects have not been captured well, if at all, by other models. The spherical version of this model has previously been shown to produce good predictions of deposition distribution in CFD. The developments presented here allow the inclusion of more complex collision physics which will improve numerical predictions of deposition. The accompanying second part to this paper demonstrates this with the addition of surface roughness modelling.