Reliability Tests and Thermal Modelling for Switching Device in Power System

Due to the increasing importance of power electronics in control of devices particularly in electrical vehicles the reliability analysis becomes important. The reliability R�(t) of a component is the probability that this component will perform its intended function after a time 't' in a given operating condition. Nowadays component reliability is not very important by considering only the power losses. For predicting reliability of power electronics components temperature and temperature cycle are to be determined.. The goal of this paper was to accurately simulate power loss and thermal behavior of a IGBT based Inverter using vendor data and a validated electrical model of a IGBT based VSI. Successful creation and implementation of these models can reduce the cost of design and production, increase reliability, quantify the accuracy of the estimated thermal impedance of an IGBT based VSI module, predict the maximum switching frequency without violating thermal limits ,predict the time to shut down on a Loss of Coolant Casualty (LOCC), and quantify the characteristics of the heat-sink needed to dissipate the heat under worst case conditions. I. INTRODUCTION The goal of this thesis was to accurately simulate power loss and thermal behavior of a IGBT based Inverter using vendor data and a validated electrical model of a IGBT based VSI. The model overview is shown below in Figure 1. Successful creation and implementation of these models can reduce the cost of design and production, increase reliability, quantify the accuracy of the estimated thermal impedance of an IGBT based VSI module, predict the maximum switching frequency without violating thermal limits, predict the time to shut down on a Loss of Coolant Casualty (LOCC), and quantify the characteristics of the heat-sink needed to dissipate the heat under worst case conditions. A thermal model was defined, created in Simulink, and calibrated to the heatsink. The input required for the thermal model was the average power output of the semiconductor devices on the VSI module. A power losses model of the was created in Simulink to obtain the average power outputs of the semiconductor devices for the thermal model. The thermal model created in this paper represents the thermal model that the vendor states approximates the thermal response of the system. The solution to the thermal model has many assumptions worth mentioning. First, it ignores the thermal losses between the inverter case and the heat sink. Normally, a thermal model would be generated to account for the thermal losses due to the thermal paste that is applied between the inverter case and the heat sink. The vendor in the application notes indicates that the thermal losses due to the thermal paste boundary are insignificant compared to the other thermal losses. Second, the thermal coupling between the IGBT and the Diode due to their close proximity is ignored because the vendor states that the thermal coupling is minimal in the application notes.

[1]  Yasushi Yamada,et al.  Novel Electro-Thermal Coupling Simulation Technique for Dynamic Analysis of HV (Hybrid Vehicle) Inverter , 2006 .

[2]  R. W. De Doncker,et al.  Reliability Prediction for Inverters in Hybrid Electrical Vehicles , 2007 .

[3]  S. Munk-Nielsen,et al.  Simulation with ideal switch models combined with measured loss data provides a good estimate of power loss , 2000, Conference Record of the 2000 IEEE Industry Applications Conference. Thirty-Fifth IAS Annual Meeting and World Conference on Industrial Applications of Electrical Energy (Cat. No.00CH37129).

[4]  A. Goel,et al.  Electronic system reliability: collating prediction models , 2006, IEEE Transactions on Device and Materials Reliability.

[5]  Anil Kr. Chaudhary,et al.  RELIABILITY TESTS AND THERMAL MODELLING FOR INVERTER IN HYBRID ELECTRICAL VEHICLES , 2012 .

[6]  A. Morozumi,et al.  Reliability of power cycling for IGBT power semiconductor modules , 2001, Conference Record of the 2001 IEEE Industry Applications Conference. 36th IAS Annual Meeting (Cat. No.01CH37248).