基于集总参数法的IGBT模块温度预测模型

Temperature Prediction Model of IGBT Modules Based on Lumped Parameters Method

从IGBT模块的内部结构和故障机理分析,得到影响IGBT模块可靠性的主要因素是温度的结论,而IGBT模块各层的温度是很难用实验的方法测取的。为了解决这一问题,在分析IGBT模块内部导热机理的基础上,利用瞬时非稳态导热的集总参数法建立热网络模型,并给出热损耗值、等效热阻、等效热容的提取方法。通过与制造商提供的IGBT模块结温实验数据、实测的底板温度和有限元模型相比较,热网络模型温度预测误差小于5%。

Temperature of insulated gate bipolar transistor （IGBT） modules is a main influence factor of reliability, which is very difficult to measure. In order to solve this problem, interior heat conduction mechanism is analyzed. Then thermal network model of IGBT modules is constructed by employing lumped parameters method of instantaneous unsteady heat conduction. Parameters extraction methods of equivalent thermal resistance, equivalent thermal capacitance and heat losses are preferred. The paper gives experimental and simulated temperature curves, which datum from thermal network model, technical documents, base-plate measurement and finite element model. Result show that thermal network model is validated.