Generally poor geological and geotechnical data are obtained from initial investigation carried out before the planning of underground construction. Traditional design methods, based on empirical or theoretical approaches, did not reflect the real process of support interface with rock mass. A new strategy is recommended to improve practice for design of underground construction based on automated design system TUNNEL and simple measurement of contour displacements during construction in the tunnel or in the tunnel prototype.
Most of the well known methods for design of support of underground construction are experimental methods. They are based on the results of field observations and measurements. Results of these methods can be accurate only for a narrow range of conditions where measurements were conducted. However engineers usually use empirical methods as universal, forgetting about limitations and poor statistics of the measurements. That is why the results, especially calculated loads on support, can vary from the real values. For example, Uo-method (Basinski & Ivanov 1972), based on empirical generalization of numerous shaft measurements of contour displacements Uo for condition yielding support, gives a reliable results for evaluation of stability and displacements for different geotechnical conditions including great depths. However the empirical approach / hypothesis for calculation of the load on rigid support, used in this method, is not justifiable. According to the empirical approach, loads on rigid support are equal for different depths, if calculated displacements for these depths Uo are equal. This hypothesis is not confirmed by theoretical elastic-plastic solutions and shaft measurements (Amusin 1977, Ardashev et al. 1985, Basinski & Ivanov 1972). Modern theoretical methods can take into account most of the peculiarities of rock mass manifestation in tunnel and interaction support with rock mass.