Page 75 - The Combined Finite-Discrete Element Method
P. 75

58     PROCESSING OF CONTACT INTERACTION

              Since individual discrete elements are discretised into finite elements, each discrete
            element can be represented as the union of its finite elements:


                                                                                 (2.42)
                                   β c = β c 1  ∪ β c 2  .. . ∪ β c i  ... ∪ β c n
                                    β t = β t 1  ∪ β t 2  .. . ∪ β t j  ... ∪ β t m
            where β c and β t are contactor and target discrete elements, respectively, while m and
            n are the total number of finite elements the contactor and target discrete elements are
            discretised into. The potentials ϕ c and ϕ t are therefore a sum of potentials associated with
            individual finite elements:

                                   ϕ c = ϕ c 1  + ϕ c 2  ··· + ϕ c i  ·· · + ϕ c n
                                                                                 (2.43)
                                    ϕ t = ϕ t 1  + ϕ t 2  ··· + ϕ t i  ·· · + ϕ t m
            Integration over the overlapping area may therefore be represented by summation over
            finite elements:
                                    n  m

                               f =             [gradϕ c i  − gradϕ t j  ]dV      (2.44)
                                   i=1 j=1  β c i  ∩β t j
            By replacing integration over finite elements by equivalent integration over finite element
            boundaries (2.25), the following equation for contact force is obtained:
                                       n  m

                                  f c =            n (ϕ c i  − ϕ t j  )dS        (2.45)
                                      i=1 j=1  S β c i  ∩β t j

            where integration over finite element boundaries may be written as summation of integra-
            tion over surfaces of finite elements. In other words, the contact force between overlapping
            discrete elements is calculated by summation over the surfaces of corresponding finite
            elements that overlap.
              In this context, the following solution strategies are available:
            • Consider contact of discrete elements with disregard for finite element discretisation.
              In this case, a special data structure is needed to keep track of boundary representation.
            • Consider contact of discretised domains of discrete elements with disregard to the
              relationship between individual discrete elements. This is the simplest path to follow.
              It may lead to unnecessary integration of contact forces over the inner boundaries of
              discrete elements, slowing down execution of the algorithm.
            • Consider integration of contact forces only over boundaries (surfaces) of finite elements
              that are also boundaries of discrete elements. This approach requires the detection and
              marking of global boundaries. In return, CPU requirements are reduced.


            2.6.2  Computational aspects

            The numerical procedure for integration of contact forces makes use of the fact that
            discrete elements are discretised into tetrahedron shaped finite elements. Thus, the total
            contact force is the sum of contact forces between individual tetrahedrons.
   70   71   72   73   74   75   76   77   78   79   80