Heat transfer prediction in rough cooling channels for liquid rocket engines

Additive layer manufacturing techniques (ALM) are becoming of increasing interest in the space industry due to their advantages in terms of production costs and times. A peculiar feature resulting from ALM is the large surface roughness, which increases skin friction and heat transfer. This latter aspect can be particularly interesting in the design of regenerative cooling channels as it could increase cooling efficiency. In the field of RANS simulations, the most employed approach to account for wall roughness is based on the equivalent sand grain assumption. Despite this assumption allows to satisfactorily model the friction coefficient, it needs suitable improvements to correctly predict heat transfer at high roughness. To this goal a thermal correction is introduced in the Spalart-Allmaras turbulence model. The implementation of the proposed correction allows to predict realistic values of the convective heat transfer coefficient, in agreement with the experimental results reported in the literature.