Experimental investigation on active temperature control performance of a loop heat pipe integrated with thermoelectric coolers

This article presents an experimental investigation into the active temperature control characteristics of a stainless steel-ammonia dual compensation chamber (CC) loop heat pipe integrated with thermoelectric coolers (TECs). The study systematically examines the effect of different parameters such as heat loads (85 W ~ 250 W), heat sink temperatures (14 °C and 18 °C), relative orientations of evaporator and condenser (horizontal and anti-gravity), and the temperature control method (Method 1 ~ Method 3) on temperature control characteristics. Method 1 and Method 2 represent cooling or heating the compensation chamber without the bayonet (namely CC1) and with the bayonet (namely CC2) by a TEC, respectively. Method 3 involves simultaneously cooling or heating both CC1 and CC2 using TECs. Experimental results reveal that: (i) a heat load range exists where the operating temperature can be regulated to the target value, termed the ‘potential zone of active control’. This zone is larger at a heat sink temperature of 14 °C compared to 18 °C and is also more extensive in the horizontal orientation than in the anti-gravity orientation; (ii) the operating temperature profile closely follows that of CC2 but differs from that of CC1 across all three temperature control methods. The temperature control accuracy is approximately ±0.2
°C after achieving active control, with a maximum input power below 3.8 W; (iii) heating CC1 using Method 1, even with high TEC input power above 2.5 W, may result in temperature control failure, including the inability to reach the target temperature and
uncontrolled temperature rise; (iv) among the three methods, Method 2 demonstrates the most effective active temperature control, achieving the target temperature with the shortest response time, which is no more than 168s. The response time of Method 2 is
at least 79.8% shorter than that of Method 1, and at least 39.7% shorter than that of Method 3.