The criteria are explored which are based on one-dimensional problems and are related to the maximum permissible dimenstions of elements in relation to the wave length of the propagating waves. Attention is focused on how much coarser the element mesh can be made for regions away from the free surface. Work is quoted that reconfirm that a h/lambda ratio of 1/12 perpendicular to the direction of propagation does provide satisfactory results and indicates that even a local refinement of the mesh leads to an improved accuracy. A symmetry test is described which is a useful technique for evaluating the adequacy of a graded mesh. The results of research lead to the conclusion that although the criterion of h/lambda less than or equal to 1/12 is necessary near the surface for finite elements to produce highly accurate predictions of surface wave amplitudes, little deterioration in accuracy is detected if the mesh is graded to h/lambda less than or equal to 1/6 at the fixed base.

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00131543
  • Record Type: Publication
  • Report/Paper Numbers: ASCE #11684 Proceeding
  • Files: TRIS
  • Created Date: May 14 1976 12:00AM