Approximation assisted optimization of headers for new generation of air-cooled heat exchangers

Abstract

An online multiobjective approximation assisted optimization approach is used to design optimum headers for compact air-cooled heat exchangers. A new CFD model is developed to predict single-phase fluid flow in headers with multiple parallel ports. This CFD model applies the porous jump interior condition in order to reduce the computational domain. In addition, Non Uniform Rational B-Splines (NURBS) are used to define and manipulate the header outer shape with the purpose of reducing the mass flow rate maldistribution. Design optimization is performed using a multiobjective genetic algorithm while the computational cost due to CFD analysis is reduced significantly by applying an online approximation technique. Optimization is performed to reduce both the mass flow rate maldistribution in different ports and the header frontal area with respect to the total heat exchanger frontal area. The optimization results predicted from metamodels are verified using CFD runs with high accuracy of prediction.