Seismic performance assessment of hydraulic tunnels considering oblique incoming nonstationary stochastic SV waves based on the generalized PDEM

An important aspect of a seismic performance assessment framework is to adequately consider the stochastic nature of SGMs. This requires an appropriate method for synthesizing random SGMs, a crucial step in conducting structural response analysis. Additionally, affected by topographic and geological conditions, the incident angle of SGMs is also uncertain and random, which may induce unexpected damage to structures. To date, a framework is presented to assess the seismic performance of hydraulic tunnels under stochastic obliquely incident SV waves. The framework incorporates the spectral representation method of SGMs and obliquely incident SV waves into stochastic dynamic analysis. To create the appropriate PDFs and CDFs for numerous performance indices, the generalized PDEM and equivalent extreme-value event are introduced. Subsequently, stochastic dynamic anal-ysis and seismic performance assessment of the hydraulic tunnel are performed. Finally, the disparity between the hydraulic tunnel fragility curves generated using conventional linear regression and generalized PDEM is examined. The difference between vertically incident and obliquely incident SV waves is discussed. The results indicated that (1) different performance indices exhibit great differences in the stochastic dynamic response of structures and the failure probability of seismic performance evaluation; (2) compared with the traditional linear regression method, the probability of damage of hydraulic tunnels under strong SGMs may be overestimated; and (3) nonstationary stochastic obliquely incident SV waves can cause more seismic demand than nonstationary stochastic vertically incident SV waves under low/medium earthquake intensity levels for hydraulic tunnel.