Coupled EM–CFD analysis of an electrical three-phase low voltage line reactor equipped with liquid- and air-based cooling systems

2021 
Abstract Power electric devices, such as line inductors or transformers, are limited by current, voltage, and operating temperature. Current is the main culprit behind the heating of all electric devices. Excess heating leads to faster degradation of the insulation and lowers the breakdown voltage. These factors determine the lifetime and reliability of these devices, which can be extended by lowering the operating temperature via cooling. This paper aims to present the numerical investigation of the coupled EM-CFD model of a line reactor (LR) with two cooling systems: air-based natural convection cooling system (AN), and air-natural with water-forced cooling system (ANWF). The design of the inductor cooling system was presented and described. The numerical investigation of both cooling systems is done considering power loss generation, cooling efficiency, and temperature distribution. In addition, the EM-CFD models of examined devices were validated by experiments considering thermal measurements using calibrated thermocouples, thermal image camera, pressure and flow rate measurements, where the temperature was measured in at least 35 points with probes. The tests were performed for three operating currents, resulting in 100%, 75%, and 50% of total power losses in the inductor. The numerical models reach an accuracy in temperature difference with respect to experiments within 5 K and 3.5 K for AN and ANWF cooling, respectively. The presented research shows that the ANWF system is superior and provides significant temperature reduction up to 68 K as well as the maximum temperature of the windings was reduced up to approx. 29.0K at 390 A RMS by use of ANWF. Furthermore, the hot-spot temperature in AN cooling system reached 135 °C, and up to 110 °C for the ANWF system, which can be further decreased down to 60.0 °C by lowering contact resistances.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    33
    References
    0
    Citations
    NaN
    KQI
    []