Experimental investigation of thermoelectricintegration in a heat pipe electronic cooler

This thesis presents a comparative investigation of a heat-pipe–finned–fan reference cooling architecture and a hybrid architecture integrating a thermoelectric (TE) module for cooling electronic components on the same experimental platform. The experimental matrix was established by systematically varying the heat load at 10–20–30–40 W (H4–H3–H2–H1) and the fan operating levels at 0.48–0.96–1.50–2.16 W (F1–F2–F3–F4); in the hybrid scenarios, the TE module was operated at two different electrical input levels (P1: 21.6 W and P2: 32 W). Thermal performance was evaluated not only based on the heater surface temperature, but also using multi-point thermocouple measurements covering the heat pipe adiabatic/condensation sections and the inlet/outlet air temperatures across the fan, while infrared thermal imaging was employed to validate steady-state temperature distributions in a spatially resolved manner. In addition, the total electrical power consumption for cooling was determined for both architectures, and fan noise levels were measured and analyzed.For the reference architecture, increasing the fan level resulted in a gradual reduction in steady-state temperature; however, the improvement exhibited diminishing marginal returns. In the hybrid architecture, activating the TE module produced an abrupt and pronounced drop in surface temperature under all conditions, confirming the capability of active cooling to suppress peak temperatures. Nevertheless, increasing the TE input from P1 to P2 raised power consumption without providing a meaningful additional improvement in the temperature profiles. When the energy–acoustic aspects were jointly considered, the noise level is governed primarily by the fan operating level, whereas TE integration can enable achieving the same/similar temperature targets at lower fan levels without introducing an additional noise source (in the reference case H1F4, 65.72 dBA was required to maintain 59.31 °C, whereas in the hybrid case H1F1P1, 51.88 dBA was sufficient to maintain 57.26 °C). Within this framework, the study quantitatively identifies the operating points at which the reference or hybrid architecture is more rational, depending on application objectives, across the axes of thermal performance, energy consumption, and fan noise. Keywords: Heat Pipe, Thermoelectric, Electronics Cooling, Fan Noise, Peak Temperature Suppression