Abstract
This study aims to investigate the characteristic of the random strain load signal in the time domain to predict durability of the lower arm. Cyclic loading on the suspension arm of vehicles leads to fatigue damage. The suspension arm is subjected to random variable amplitude loading due to different road conditions. Strain load signal data from a previous experiment are used for the analysis, and statistical parameters from the signal, such as mean, standard deviation, root-mean-square, kurtosis, skewness are extracted in time domains, while the power spectrum density in frequency domains. The fatigue life of three different roads was predicted using Coffin–Manson (CM), Morrow and Smith–Watson–Topper (SWT) strain-life models. Results showed the strain life of the lower arm ranges from 4.47 × 10
6
cycles/block to 8.8 × 10
9
cycles/block, 2.58 × 10
6
cycles/block to 3.18 × 10
10
cycles/block, and 2.21 × 10
6
cycles/block to 5.06 × 10
10
cycles/block for Coffin–Manson, Morrow and Smith–Watson–Topper model, respectively. Hence, the fatigue life of lower arm can be modeled using the mean stress effects due to compression and tension for various road condition based on the statistical properties of the strain signal i.e., kurtosis and root-mean-square value.