Condition Monitoring of Thermal Fatigue of Power Module Based on Thermal Diffusion Time

作者：Ronghu Ke, Jianfeng Tao, Chengliang Liu · 发表于：IEEE Transactions on Power Electronics · 年份：2025 · DOI：10.1109/tpel.2025.3546940 · 被引用次数：6 · 研究领域：Engineering Applied Research、Advanced Sensor Technologies Research、Machine Fault Diagnosis Techniques

This letter proposes an aging indicator for thermal fatigue based on thermal diffusion time. Initially, two types of times are analyzed: first, thermal propagation time—the time required for heat to propagate; second, thermal diffusion time—the time required for a noticeable temperature change. The former is derived from the Cattaneo model, incorporating thermal relaxation time, while the latter is based on the Fourier model, assuming infinite heat transfer speed. Both times increase as thermal conductivity decreases. However, since the thermal propagation time is extremely short, the thermal diffusion time is better suited as an aging indicator. Building on this, an online evaluation algorithm that utilizes varying cooling conditions is designed. It also leverages the fact that, in the early stages of heat conduction, fan speed has minimal impact on junction temperature. Based on this characteristic, a data normalization method is proposed to eliminate the influence of power loss. Experimental validation confirms the effectiveness of the method.