Page 200 - Rock Mechanics For Underground Mining
P. 200

METHODS OF STRESS ANALYSIS

                                          The discretised form of equation 6.27 can be written

                                                           n %

                                                                      x         y
                                                                q x (S)T + q y (S)T xi  dS = t xi     (6.29)
                                                                      xi
                                                          j=1  s e
                                                           n %

                                                                      x         y
                                                                q x (S)T + q y (S)T                   (6.30)
                                                                      yi       yi  dS = t yi
                                                          j=1  s e
                                        where n is the number of boundary elements, and each surface integral is evaluated
                                                     j
                                        over the range S e of each boundary element j. Considering a particular element, one
                                        of the surface integrals can be expressed by
                                                %
                                                         x         y
                                                   q x (S)T + q y (S)T xi  dS =
                                                         xi
                                                 s e
                                                        %  1                       %  1
                                                                  x  dS                     y   dS
                                                     q x    N   ( )T ( )  d  +  q y   N   ( )T ( )  d   (6.31)
                                                                 xi                         xi
                                                         −1          d              −1          d
                                        Theintegralsoftheinterpolationfunction(N)–kernel(T )productsdefinedinequation
                                        6.31 can be evaluated readily by standard Gaussian quadrature methods. When all
                                        components of equations 6.29 and 6.30 have been calculated using the procedure
                                        defined in equation 6.31, it is found that for the m boundary nodes
                                                              m


                                                                 q xj T  x∗  + q yj T  y∗
                                                                     x       x  = t xi
                                                              j=1
                                                                                                      (6.32)
                                                              m


                                                                 q xj T  x∗  + q yj T  y∗
                                                                     x       y  = t yi
                                                              j=1
                                        where T ,etc.,aretheresultsofthevariousinterpolationfunction–kernelintegrations
                                               x∗
                                              x
                                        and, for the end nodes of each element, a summation with the appropriate integral for
                                        the adjacent element. When equations similar to 6.32 have been established for each
                                        of the m boundary nodes, they may be recast in the form
                                                                      ∗
                                                                    [T ][q] = [t]                     (6.33)
                                        Equation 6.33 represents a set of 2m simultaneous equations in 2m unknowns, which
                                        are the nodal values of fictitious boundary load intensity.
                                          Once equation 6.33 has been solved for the vector [q] of nodal load intensities,
                                        all other problem unknowns can be calculated readily. For example, nodal displace-
                                        ments, or displacements at an internal point i in the medium, can be determined
                                        from

                                                                 %
                                                                          x        y
                                                            u xi =  q x (S)U + q y (S)U xi  dS
                                                                          xi
                                                                  s
                                                                                                      (6.34)
                                                                 %
                                                                          x        y
                                                            u yi =  q x (S)U + q y (S)U yi  dS
                                                                          yi
                                                                  s
                                        182
   195   196   197   198   199   200   201   202   203   204   205