부산대학교 도서관

Obtaining the thermal flow field distribution in transformers is essential for ensuring reliable operation. A finite volume method is a crucial approach for obtaining the internal thermal flow field in transformers through multiphysics simulation modeling. The complexity of the transformer tank-wall structure makes it challenging to accurately calculate wall surface heat transfer coefficients using empirical formulas under varying operating conditions. This study introduces a method to calculate the dynamic convective heat transfer coefficient of the transformer tank-wall, derive the wall temperature equation, and obtain the internal thermal and flow field distribution using transformer operation data. First, a finite volume method multiphysics simulation model is established for steady-state calculations of transformers, and simulated datasets of tank-wall temperatures are obtained. Subsequently, the dataset undergoes particle swarm optimization (PSO) function optimization. Finally, the internal thermal and flow fields in the transformer are determined using dynamic heat transfer coefficients and operational data. By comparing the simulated wall temperature distribution with experimental measurements, the feasibility and accuracy of the proposed method have been confirmed. The finite volume multiphysics field calculation method for the internal thermal flow field of transformers based on dynamic heat transfer coefficients can disregard the complexity of the transformer tank-wall structure and consider the varying operating conditions, compared to empirical formulas (5.238 K), it has a smaller maximum error (1.373 K), demonstrating clear practical value.