The Research On The Temperature Field Of Transformer Winding

Under the case of a single transformer capacity and the growing failure problems caused by the temperature rise more and more, the research of transformer temperature field is increasingly important. It is urgent to know where transformer winding hot spot occurs when at normal running, At present, the traditional calculation of the temperature rise of transformer winding design has been difficult to meet the performance requirements of this,the contents of this paper is to propose in this background.Through the establishment of appropriate equivalent transformer model, using finite volume method in Fluent can approximate the actual operation temperature distribution of the transformer winding,it provides a temperature calculation,analysis method for transformer design engineers, it also can prevent the temperature rise failure problems caused by the unreasonable design process, thus avoiding some of the economic loss and it has important practical significance.In this paper, work done by the following:1) through the familiar and master the electromagnetic loss calculation of special software MF2D, in the structure and electrical analysis of dry-type transformers calculate the DC resistance loss and eddy current losses in the normal operation to calculate the winding volume loss density in different districts;2) Based on detailed analysis of the dry-type transformers, retain its influence to the winding temperature, as far as possible to maintain the original structure of some of the characteristics, establish two-dimensional model by the commercial software Fluent, the model structure equivalent used to calculate dry-type transformer winding temperature field, it provides a reliable idea of simplification;3) Set the actual run-time parameter electrical and structural characteristics of model, combined with the calculated core and winding loss density, calculate the winding temperature field in normal operation with fluent, the results were verified reliable by the experimental data, the error is in allowable range.