Simulation of a Heat Pipe Test Fixture: Laminar, Internal, Forced Convection in a Tubed Cold Plate

Abstract

Space Copper Water Heat Pipes (SCWHP) are a relatively new technology in the field of heat pipe electronics cooling. Due to the market and extraterrestrial environment, these heat pipe assemblies require extensive testing in order to verify compliance to customer requirements. Consequently, the design of the test fixtures for SCWHP, which usually support 3-4 assemblies at a time, have become relatively standardized over the past 3 years, consisting of heater blocks and a copper, tubed cold plate attached to a benchtop chiller for heat rejection. Surprisingly, the fixtures have yet to be analyzed in a Computational Fluid Dynamics or Conduction software. The objective of this project was to utilize commercial software to analyze an example test fixture as it would be used for an example thermal performance test. The goal was to study the effects of varying heat pipe power inputs and operating temperatures on the temperature gradient across the heat pipe assemblies, the required chiller setpoint, and convection parameters such as local heat transfer coefficients and Nusselt numbers. COMSOL Multiphysics was used to run 5 different power and operating temperature cases and calculate the setpoint required to obtain a desired heat pipe condenser temperature, the condenser temperature gradient across the heat pipes, the local heat transfer coefficients, and the local diametric Nusselt numbers. It was found that increasing input power required a lower chiller setpoint to maintain the same desired condenser temperature and furthermore increased the condenser temperature gradient across the assemblies from 0.2 degrees Celsius to 0.5 degrees Celsius. It was also found that the tubing bends increased heat transfer in the form of the Nusselt number by an average of approximately 38 percent due to the secondary flows induced. Finally, it was observed that increasing input power increased heat transfer coefficients by an average of roughly 7.9 Watts per square meter per Kelvin per Watt.