Modeling of Additive Manufacturing-like Rough Walls from Roughness-resolved LES Database

Abstract

Extended Abstract Recent development of additive manufacturing (AM) in the past decade paves the way for breakthrough designs of heat exchangers and especially for compact ones (CHX). However, the surface roughness generated with AM is larger compared to conventional manufacturing processes. In addition, the roughness is basically anisotropic. The impact of this type of roughness on pressure drop and heat transfer coefficient cannot be neglected during the design process. Numerical RANS and LES simulations have proven to be efficient tools for optimization purposes and are appropriate candidates to fulfill this need. Nonetheless, for realistic cases, required computational resources to conduct simulations with roughness are not affordable in general. Thus, the modeling of the effects of the rough elements on the flow without explicit representation of the surface details is compulsory. Common approaches for RANS/LES rough wall modeling rely on the prediction of the mean wall stress through a modified smooth law of the wall. The modification of the law of the wall mainly follows empirical correlations obtained from experimental data and roughness resolved simulations. In line with this philosophy, a first objective at our concern is to enrich the empirical correlations with data from typical AM roughness. This kind of approach is theoretically designed to predict the mean value of