Time Series Behaviour of Lower Arm Suspension Fatigue Data Using Classical Decomposition Method

The study of time series behaviour refers to the analysis of certain unique attributes that exist in the time series data. The presence of these attributes in the data series may influence the decision making process. These attributes are generally grouped into four main component types which are trend, cyclical, seasonal and irregular components. In this study, fatigue signal data with three different road factors from a lower arm suspension for a mid-sized car were used as the case study. ldquoClassical decompositionrdquo time series method was used to segregate and to analyse the existence components in a systematic manner. Although fatigue data is a time series signal, not all components were considered. This is due to the nature of fatigue behaviour itself which is different from a normal time series data. From the study, it was found that only trend, cyclical and irregular component existed in the fatigue data signal. The study also revealed the additive effect that existed between these three types components as the absolute sizes of the seasonal variation are independent of each other.