Probabilistic-based fatigue reliability assessment of carbon steel coil spring from random strain loading excitation

This paper aims to assess the fatigue reliability of random loading signals of a suspension coil spring using probabilistic approaches. Strain time histories were acquired while the car was travelling on different road conditions (i.e., in a rural area, in an industrial area, on a university campus, on a highway and on a newly constructed road). Fatigue lives were predicted from the strain histories and fitted into probability density functions. Lognormal distribution was found to be an appropriate way to represent fatigue data. Next, the reliability function and mean-cycles-to-failure (MCTF) were determined. The results indicated that fatigue reliability rapidly deteriorated under rural road conditions, which resulted in a short MCTF of 10(4) cycles. Meanwhile, the new road signals had the longest MCTF of about 10(8) cycles. Accordingly, this is due to the rural road having the most surface irregularities, which caused more severe fatigue damage to the coil spring. This study contributed to a greater in-depth understanding of the effect of loading signals on fatigue reliability. This is essential in determining the appropriate service life of the coil spring during its production to ensure vehicle safety and reduce maintenance costs.

Automated summary

This study aimed to assess the fatigue reliability of a suspension coil spring under different road conditions through probabilistic approaches. Results showed that fatigue reliability deteriorated rapidly under rural road conditions, with the shortest mean-cycles-to-failure (MCTF) of 10^4 cycles, while new road conditions had the longest MCTF of about 10^8 cycles. Surface irregularities on rural roads had the most significant impact on fatigue damage to the coil spring.