Probabilistic Evaluation of Automotive Cold Cranking Performance

A probabilistic model proposed by Wang (1990) is summarized and extended to quantify the approximate cranking probability at a specified confidence level. The criterion for acceptance is the ability of a randomly selected starter/battery set which delivers acceptable cold cranking performance on a specified engine application above a minimum targeted speed required to start. The model serves as a decision-making tool for engineers to (1) verify existing cranking system performance, (2) assess new combinations of existing cranking components, (3) evaluate the performance of newly developed cranking components, and (4) provide functional requirements of starter/battery sets for component sizing purposes. The model can be applied to perform trade-off analyses of cranking system performance with test results or computer simulation conclusions. An example is discussed and engineering usage is illustrated. A sensitivity analysis is performed under the uncertainty of engine variability due to small testing sample size. Future research direction is also recommended.