Abstract

Topology optimization generate promising ‘heat sink fin layouts’ to satisfy the thermo-hydraulic objectives with lightweight designs. This article evaluates the performance of topology-optimized liquid-cooled heat sinks with different spanwise length-based design domains. A reduced-scale model (with 1/8th the width of the full-scale heat sink model) is selected and optimized with ‘pressure drop’ and ‘average junction temperature’ as objective functions and constraints, respectively. Fin layouts consisting of non-continuous fins are derived for different inlet flow velocities and temperature constraints using an accurate two-dimensional (2D) mathematical model, mimicking a three-dimensional (3D) heat sink geometry. Three-dimensional CFD simulation is used to compare the performance of the present fin layout with a reference topology-optimized design (with 1/16th the width of the full-scale model) and an augmented straight channel-based heat sink. To cool the heat sink to an average junction temperature of 50 °C, the present design requires 17 % and 54 % lower pumping power compared to the reference design and straight channel with significantly reduced material volume. Flow phenomena such as secondary flow-induced mixing and frequent reinitialization of boundary layers at leading edges of fins are observed during numerical investigation which are illustrated using velocity and temperature contours. The optimized fin layout's overall performance is compared with existing fin designs of heat sinks such as step fins, offset strip fins, oblique fins, and oblique and trapezoidal fins. Oblique and trapezoidal fin and oblique fin compete with topology optimized design which outperforms at lower flow rates (for velocity ranging from 0.05 m/s to 0.175 m/s). The average junction temperature using Oblique and trapezoidal fins and oblique fins are 63 °C and 61.22 °C, respectively compared to 58.5 °C for topology-optimized design with a similar pressure drop of 300 Pa.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call