New results from reduced-order cfd-based multi-rating of aviation heat exchangers

In AIAA Paper 2019-1560, the present author presented an approach toward a CFD-based reduced-order modeling of advanced aviation heat exchangers with the objective of generating the j and f factors required for physics-based, but quasi-empirical analysis of new heat exchanger designs, such as additively-manufactured components, for which no correlations are available. The goal envisioned for Reduced-Order Computational Fluid Dynamics and Heat Transfer (ROM-CFD) is obtaining fast turnaround in simulation time so that CFD could form a component of a system-level synthesis/design tool. In the cited reference, an illustration is given for the Nusselt number distributions for a classical natural convection cooling problem, where we show that a few finite elements (less than 10) can provide acceptable accuracy relative to models using thousands of elements. Moreover, the ROM-CFD accuracy alluded to is significantly higher than those normally encountered in realistic designs which are invariably based on empirical data. In the present paper, the ROM-CFD procedure is applied to the mixed convection problem associated with non-isothermal flow over a bank of tubes for a range of flow strength.