Page 256 - Numerical Analysis and Modelling in Geomechanics
P. 256

F.PERGALANI, V.PETRINI, A.PUGLIESE AND T.SANÒ 237
            direction j at point r′ and φ  (r′) is the force density on the boundary in direction
                                  j
            j.  The  closed-form  solution  of  the  Green  function  has  been  derived  for  two-
            dimensional elastodynamic problems of the whole space (Kummer et al., 1987).
            Equation  (8.5)  shows  that  the  displacement  at  any  internal  point  can  be
            determined as a sum of the effects of forces applied at the boundaries. Such an
            equation  stems  from  the  Somigliana  identity,  which  is  the  base  of  the  direct
            approach  of  the  BEM.  Kupradze  (1963)  showed  that  the  displacement  field  is
            continuous across the boundary, S, if ф  is continuous along S.
                                           ij
              Stresses and tractions can be calculated by direct application of Hooke’s law
            except  at  the  boundary  singularities,  i.e.  when  r→r′  on  the  boundary.  In  this
            situation the Green function, G  has a logarithm-type integrable singularity, but
                                     ij,
            its  derivative  can  be  computed  only  if  the  singularity  is  extracted  (Kupradze,
            1963). The integral can be calculated by a limiting process based on equilibrium
            considerations around the singularity, in the following form:

                                                                         (8.6)


            where t  is the i-th component of traction at the boundary and C is equal to zero
                  i
            outside the boundary and equal to ± 0.5 for a smooth boundary. The signs + and
            −  are  valid  respectively  for  the  interior  and  exterior  domain.  T ij  (r,  r′)  is  the
            traction Green function and represents the traction in direction i at point r on the
            boundary due to the application of a unit force in the direction j applied at r′.
              The  equations  (8.5)  and  (8.6)  are  the  basic  formulations  for  solving  the
            problem  of  wave  propagation.  Consider  the  space  in  Figure  8.10,  in  the
            homogeneous  infinite  domain  external  to  the  valleys  on  the  right,  under
            incidence  of  elastic  waves.  It  is  usual  to  distinguish  the  resultant  motion  as
            composed  of  the  free  field  motion,  i.e.  the  incident  waves,  and  the  “scattered”
            waves, i.e. those reflected, diffracted and refracted from the boundary. In fact the
            ground  motion  in  this  irregular  configuration  physically  comes  from  the
            interference of incoming waves with those generated by the boundaries. On the
            contrary, in the limited areas of the valleys, on the right-hand side of Figure 8.10
            and  also  called  inclusions,  the  motion  is  due  only  to  diffracted  waves.  The
            resultant motion is expressed by:

                                                                         (8.7)

                                     s
            where u° is the incoming and u  the scattered motion. It is assumed that both the
            waves  also  exist  for  z>0,  that  is  in  the  air,  fulfilling  the  same  analytical
            expression valid for z≥ 0.
              Similarly for stresses and tractions the resultant value is:

                                                                         (8.8)
   251   252   253   254   255   256   257   258   259   260   261