Fatigue failure is one of the major failure modes for internal combustion engines, especially with reduction in engine size and increase in combustion pressure and operating temperature. Dynamometer tests are devised to ensure engine durability for high and low cycle fatigue. With the advent of CAE technology, the dynamometer test behavior can be simulated using CAE analysis and engine durability can be assessed. The data generated in CAE analyses can be used to predict failure of the engines or future engine design modifications.
The present paper has two parts - first is running finite element analysis (FEA) to get stress, strain data and running high cycle fatigue analysis to get safety factors and second is creating a predictive tool to assess failures using data from the first part as inputs. Using advancements in the field of machine learning, the paper presents use of support vector machine (SVM) algorithm to predict failure of the engine based on inputs.
The paper discusses procedure to build the data driven model and its results. Such data driven model can be used as a tool to predict failure of engine arising from future design changes or as a variation tool to predict failure arising from changes in inputs.