Nusselt correlation for ellipsoidal particles

Abstract

Direct numerical simulations of a fixed, prolate ellipsoid of constant temperature embedded in uniform flow conditions are performed. The investigated parameter space covers the Reynolds number Re, the inclination angle ϕ, the particle aspect ratio β, and the temperature ratio TR. The parameters are varied in the range 1≤Re≤100, 0°≤ϕ≤90°, 1.5≤β≤8, and 0.35≤TR≤1.65 resulting in more than 6600 computations. For selected configurations, flow and temperature fields are presented and discussed to determine the physical impact of the parameter space on the heat transfer. The analysis shows that the heat transfer is mainly governed by the Reynolds number and the imposed temperature ratio, while the inclination angle shows an increased impact at higher particle aspect ratios. An artificial intelligence algorithm is used to derive a correlation function for the Nusselt number Nu. The statistical contribution of each parameter to the correlation function is evaluated and discussed. The results show that the fluid-particle dynamics is accurately modeled by the novel correlation function not only within the trained but also an extended parameter range.