Page 84 - Numerical Analysis and Modelling in Geomechanics
P. 84

WAVE-SEABED-STRUCTURE INTERACTION 65


                                                                        (3.9b)


            in the x- and z-directions, respectively.
              Appropriate boundary conditions are required to solve the governing equations
            (3.1)  and  (3.9).  For  a  porous  flow  in  a  seabed,  the  boundary  conditions  at  the
            impermeable  rigid  bottom  require  that  the  dynamic  fluctuations  of  all  the
            physical quantities vanish, i.e.,


                                                                        (3.10)

            For the lateral boundaries of the computation domain, since the existence of the
            structure  only  affects  the  wave-induced  soil  response  near  the  structure,  the
            disturbed  components  due  to  the  existence  of  a  structure  will  vanish  far  away
            from the structure. Thus, the soil response at these points should be that induced
            by  waves  without  any  structures  (see  Yamamoto  et  al.,  1978;  Jeng,  1997b).
            These  lateral  boundary  conditions  should  be  determined  before  including  the
            structure  into  the  whole  model.  The  details  of  the  numerical  procedure  will  be
            described in the section headed ‘Numerical procedure’.
              Besides  the  bottom  boundary  conditions,  the  seabed  surface  conditions  are
            also  required  for  the  wave-seabed-structure  interaction.  Since  these
            boundary conditions will vary with the type of structure, they will be described
            in each example.


                     General finite element model for wave-seabed-structure
                                 interaction (GFEM-WSSI)
            Since  the  wave-induced  oscillatory  soil  response  fluctuates  periodically  in  the
            temporal domains under harmonic wave loading, the wave-induced soil response
            can be assumed to take the form



                                                                        (3.11)



            where  subscripts  “r”  and  “c”  represent  the  real  and  imaginary  parts  of  the  soil
            response, respectively.
              Substituting  (3.11)  into  (3.1)  and  (3.9),  then  directly  applying  the  Galerkin
            method  (Zienkiewicz  and  Taylor,  1989)  to  these  equations,  the  finite  element
            analytical formulations can be expressed in matrix form as
   79   80   81   82   83   84   85   86   87   88   89