LAUSR

Abstract Additive manufacturing is a promising candidate to eliminate design constraints in thermal-fluid engineering problems as it enables optimized free-form designs to be produced. The brute force optimization needs too many experiments and/or analyses to obtain an optimized design. However, the same engineering design can be achieved by fewer experiments by using design of experiment (DOE) methodology, which minimizes the work load. In this study, local and mean heat transfer coefficients for a grooved pipe model are optimized and the system constraints are determined in compatible with the additive manufacturing method. DOE based response surface methodology (RSM) and an in-house developed optimization code are used to obtain the optimal local and mean heat transfer coefficients. The analyses are performed for both laminar and turbulent flow cases for understanding the backward and forward effects of grove dimensions better. The comparison of the recirculation regions in the groove geometries for optimal cases of laminar and turbulent flows reveals that the grooves in the pipe do not have to be uniform along the streamwise direction to obtain optimal heat transfer for local needs. This optimization approach provides both local and global controllability of heat transfer and the heat transfer values can be increased up to 20% in this thermal-fluid engineering problem. In addition, pressure drop and heat transfer contribution ratio of each grooved section are presented in detail.