Search

In modern automotive control modules, mechanical failures of surface mounted electronic components such as microprocessors, crystals, capacitors, transformers, inductors, and ball grid array packages, etc., are major roadblocks to design cycle time and product reliability. This paper presents a general methodology of failure analysis and fatigue prediction of these electronic components under automotive vibration environments. Mechanical performance of these packages is studied through finite element modeling approach for given vibration environments in automotive application. The vibration simulation provides system characteristics such as modal shapes and transfer functions, and dynamic responses including displacements, accelerations, and stresses. The system level model is correlated through vibration experiments. Using the results of vibration simulation, fatigue life is predicted based on cumulative damage analysis and material durability information. Detailed model of solder/lead joints is built to correlate the system level model and obtain solder stresses. Predicted failure mechanism of the leads agrees with the experiment observation. On the test vehicle with multiple components, one of the 160-pin gull-wing lead plastic quad fiat packages was chosen as an example to illustrate the approach of failure analysis and fatigue life prediction.

The optimum maintenance time interval used for the threshold values of monitored variables under the age replacement policy can be estimated using a reliability based method which considers the value of monitored variables. This approach requires computations such as the maintenance cost equation, the proportional hazards model, the reliability function and the total time on test plot. However, the method is based on the assumption that the required values are available just before a failure.

The predictive Bayes approach to reliability can be used to evaluate the different design solutions for non-failure-tolerant systems. Using the Bayes approach, a problem-solving design for a Scram system considering constant-failure-rate components and plant-transients has been developed. An analysis of the problem-solving design reveals that all reliability problems considered by a predictive Bayes framework are sequential statistical decision problems. In addition, the results indicate that design by mean-cost minimization requires extensive numerical analysis.

Very large scale integration digital signal processors need to be fault secure. However, conventional approaches to enhancing the fault-security of such devices entails considerable hardware to accommodate the vast number of decision parameters. To overcome such cost and hardware problems, a new computer-aided behavior synthesis of fault-secure microarchictectures has been developed. The approach selects intermediate computations for additional voting, making it less hardware-dependent and costly.

Most electric motor predictive maintenance methods have drawbacks that limit their effectiveness in the mining environment. The U.S. Bureau of Mines (USBM) is developing an alternative approach to detect winding insulation breakdown in advance of complete motor failure. In order to evaluate the analysis algorithms necessary for this approach, the USBM has designed and installed a system to monitor 120 electric motors in a coal preparation plant. The computer-based experimental system continuously gathers, stores, and analyzes electrical parameters for each motor. The results are then correlated to data from conventional motor-maintenance methods and in-service failures to determine if the analysis algorithms can detect signs of insulation deterioration and impending failure. This paper explains the on-line testing approach used in this research, and describes monitoring system design and implementation. At this writing data analysis is underway, but conclusive results are not yet available.