Abstract
Additive manufacturing (AM) enables designs that are not manufacturable with conventional techniques and that offer reduced weight and better thermal performance. This is of particular relevance for gas turbine cooling as well, as proved by the first applications of combustors and turbine blades made entirely by AM.
Despite that, the effects of the increased surface roughness on friction and heat transfer are not fully understood yet and this hinders the design process demanding further characterization by means of experimental tests. Also the exploitation of CFD simulations is complicated by the current modelling approaches, still based on the equivalent sand-grain roughness (ks) that results in a velocity shift in the boundary layer. The estimation of such parameter is not straightforward and the approach leads to an unphysical overestimation of the heat transfer even using ks values calibrated with experiments.
This work represents the continuation of the benchmarking and calibration of a CFD methodology based on friction and thermal corrections proposed by Aupoix from ONERA. The results presented in the previous two papers focused on three literature test cases representative of straight and wavy additively-manufactured minichannels with circular and rectangular cross-section. In this paper a new dataset published by Penn State University is considered, investigating the effect of diameter and printing direction on the fluid-dynamical performance of circular straight minichannels. This work confirms the validity as a first approximation of a linear relation between ks/Dh and the relative arithmetic roughness Ra/D, provided that the build orientation is sufficient to avoid excessive channel deformation (i.e. greater or equal to 45°). On the contrary, a similar calibration for the heat transfer was unsatisfactory, suggesting a role of channel shape or other parameters not understood at the moment.
