AbstractsEngineering

In manufacturing industry, a successful machine development requires the durability of structure components to meet fatigue life targets. The typical way to obtain fatigue design loads for conceptual design exploration is based on hand calculations or historical data to capture envelopes of expected system responses, which may not guarantee to capture actual damaging loads. In this study, a new approach is developed to extract a fatigue design load set directly from measured load data for a conceptual design exploration. The proposed framework integrates the techniques from data analytics and physics based engineering mechanics to amplify and detect fundamental damaging load patterns. Also, a practical Taguchi optimization method is proposed by using a moving window strategy to minimize the computational cost of design exploration. An industrial scale structural problem, a front linkage structure of a hydraulic excavator, is presented to demonstrate the effectiveness of the proposed methodologies.