LAUSR

Thermal-fluid coupled calculation is an effective way to simulate the temperature rise and heat dissipation process of switchgear. But there are still problems such as huge calculation and low accuracy need to be solved. This paper discusses the optimization methods of the thermal-fluid coupled field for the switchgear from grid controlling, boundary conditions, and heat source calculation. First, the mesh adaption method based on posterior error estimation is proposed to achieve efficient mesh refinement with less redundancy. The final mesh size is only 32.3% of that obtained with the traditional global-refining method. Then, an external flow model is built to obtain the convective heat transfer coefficient of the enclosure, replacing the process of artificial choosing. The results show that the convective heat transfer coefficient at the enclosure under natural convection is 0.4 W/( $\text{m}^{2}\cdot ^{\circ }\text {C})\sim 1.4$ W/( $\text{m}^{2}\cdot ^{\circ }\text{C}$ ). Afterwards the eddy current field is used to solve the heat generation in the switchgear. The heat sources are coupled to the thermal-fluid calculation so that the influence of current non-uniformity, contact heat, and eddy loss is considered. At last, the methods are applied to the steady-state temperature rise simulation of KYN28A-12kV/630A switchgear and the results are compared with the test data. The maximum relative error between simulation and experimental results is 3.43%, which proves the validity of the mentioned methods.