Mathematical models are the cheapest tools for assessment of tunnels. The constitutive law has to be determined from either in situ measurement (too expensive), or from experimental physical models. A special coupling of physical and mathematical modeling is presented in this paper. Based on results from physical models, input material properties in mathematical model were specified. Stability of tunnel face was solved in dependence of loading of the terrain.
The construction of any underground work causes both qualitative and quantitative changes in rock muss surrounding the underground structure. First. the changes of original stress state can be expected. The recovery of the equilibrium of the stress state is accompanied by deformations in the rock mass. The deformations- stress responses of the rock mass can be determined straight in situ and/or studied and predicted by either numerical or experimental modeling. Model methods in geotechnics enables one to investigate mechanisms of geomechanical phenomena, to found stress changes under various processes of construction of under-ground structures. The time-dependent conditions can also be simulated. In both mathematical and experimental models the rock mass is substituted by ones, which have to fulfill all determinative properties of real natural conditions. Success of these methods is strictly given by accuracy of input parameters. Numerical methods are burdened with an error, which is an aftermath of unsurely determined input data, mainly of improper constitutive law. The advantage of the numerical solutions consists of flexibility [input parameter variations and high speed of results are attained]. The advantage of physical (experimental) models consists of real and complex simulation of deformation process in the rock mass. The formulation of boundary conditions in experiments are based on physical and geometrical similar- ity,(Kozesmik, 1983; Skorepova,1991; Stilborg et al, 1979). The most objective results can be reached by connecting both methods. Consequently, such a coupled modeling seems to be the most appropriate procedure in estimating the rock mass behavior. The results from experiments can be used as the input data for mathematical modeling. In order to study stability of tunnel face, it means determination of its deformation and failure in dependence of stress change due to loading of terrain, experiment on models from physically equivalent materials was created. The tunnel, reinforced by concrete lining was also simulated in the physical model. The numerical model follows the FEM, extended by the influence of either eigenstrains or eigenstresses (Prochazka & Sejnoha, I 995; 1997). The eigenparameters stand for plastic strains or relaxation stresses, and they describe the constitutive law needed for the numerical analysis. Their distribution is determined from comparison with the experiments in a mathematical formulation. The process how to determine these parameters is presented in this paper. The eigenparameters can depend of a position. All these requirements and processes may be involved in accordance with the experiment and measurement in current time. The numerical models using the experiment results may show that something is out of validity of law being generally accepted.