An important element of time-dependent reliability analysis of civil structures is choosing a proper model for the applied loads. Stochastic process theory has been widely used in existing studies to perform structural time-dependent reliability analysis. However, the use of many types of power spectral density functions leads to an inefficient calculation of structural reliability. This paper proposes an analytical method for structural reliability assessment, where a new power spectral density function is developed to enable the reliability analysis to be conducted with a simple and efficient formula. A non-Gaussian load process, if present, is first converted into an equivalent Gaussian process to improve the assessment accuracy. Illustrative examples are presented to demonstrate the applicability of the proposed method. Results show that a greater autocorrelation in the load process leads to a smaller failure probability. The structural reliability may be significantly overestimated if one simply treats the non-Gaussian load process as a Gaussian one. Moreover, the impact of modeling the load process as a continuous process or a discrete one on structural reliability is also investigated.
|Number of pages||10|
|Journal||Journal of Structural Engineering|
|Publication status||Published - Dec 2019|
- Load autocorrelation
- Outcrossing rate
- Stochastic process
- Time-dependent reliability