Health Monitoring Method for Load Assessment and Cooling Performance Degradation of Printed Circuit Boards

Abstract

Health monitoring technologies, which can evaluate the performance degradation, load history and degree of fatigue, have the potential to improve the maintenance, the reliability design method and the availability in improper use conditions of electronic equipment. In this paper, we propose a method to assess the cooling performance degradation and load history of printed circuit boards in electronic equipment by use of a hierarchical Bayes model based on CAE (Computer Aided Engineering) results of thermal stress simulation and experiment data from actual measurements. We applied this method to a notebook PC that can monitor the device load factor and revolution number of cooling fan. It is shown that this method can estimate the temperature and deformation distribution of the printed circuit board from monitoring variables through latent variables such as thermal dissipation of the device and thermal boundary condition by use of the hierarchical Bayes model. And it is confirmed that the statistical load assessment concerning thermal cyclic load and the maximum load distribution can be conducted using the estimated temperature and deformation data. We verified that the cooling performance degradation can be assessed, if the temperature difference per unit thermal value between two suitable points on the printed circuit board can be obtained. Furthermore, we attempted the estimation method based on the hierarchical Bayes model to dynamic load assessment such as cyclic drop impact for PCB. The assessment method for the strain range distribution of the solder joints on the PCB was conducted to lead to the conservative results for reliability design. It is concluded that the proposed method can be effective to assess the field load history and cooling performance degradation.

Prognostics and health monitoring of electronic system: A review

“…It is using parameters from the ECU such as ECU reset and initialisation statistics, ECU error counts, function activation statistics, network status and performance statistics. Also, FMMEA can indicate software parameters to be monitored such as CPU usage, CPU throttle [21], CPU loading factor [28]. Other examples can be fault codes, scan error, memory usage capacity or queue lengths.…”

“…Hirohata et al [28], monitors the cooling performance degradation and load history of a Printed Circuit Board (PCB) in digital equipment. A hierarchical Bayes model based on Computer-aided Engineering results of thermal stress simulation and experiment data from actual measurements is used.…”