Reliability based design approaches are usually adopted to explicitly consider uncertainties in rockmass and make decisions on selection of design parameters for tunnel-support problems. However, it is difficult to accurately characterize the input random variables due to fewer sample availability and costly observations. Thus, it is necessary to efficiently allocate the resources towards the characterization and reduction of epistemic uncertainty of those input variables, which have higher relative contributions towards the variability in output. These input variables are identified using global sensitivity analysis based on Sobol indices conducted for the case of supported circular tunnel in Hoek-Brown rockmass. Convergent confinement analysis method is applied to obtain the output – radius of yield zone, tunnel convergence and induced load on installed support. Results indicate that relative contribution of variation in uniaxial compressive strength of rock is highest towards the variation in the outputs followed by geological strength index of rockmass, its deformation modulus and thickness of the liner support.
Uncertainties in rockmass properties are well known among the geotechnical engineers. Conventional deterministic design approaches do not explicitly include these uncertainties, and this might lead to either a very conservative design or an unsafe design. Reliability based design (RBD) approach address this issue by modeling rockmass parameters as random variables and evaluating the probability of failure (Pf) of tunnel-support system. RBD aims to obtain design parameters which satisfy safety requirements (specified target Pf). However, determining the probability distribution and its parameters of rockmass variables is challenging due to complex geological conditions, limited availability of samples, costly observations, etc. This leads to inaccurate estimation of statistical properties of rockmass parameters which might affect the RBD analysis.
Improving the statistical characterization of rockmass parameters can be achieved by taking a greater number of measurements and addressing issues of measurement errors, transformation uncertainty. Several researchers such as Phoon and Kulhawy, 1999; Ching et al., 2018; Juang et al., 2019, etc. developed framework in this regard. Bayesian framework was also adopted for statistical characterization of geotechnical parameters (Wang et al., 2016; Zhang et al., 2017).