Premium
INFINITE DOMAIN CORRECTION FOR ANTI‐PLANE SHEAR WAVES IN A TWO‐DIMENSIONAL BOUNDARY ELEMENT ANALYSIS
Author(s) -
HEYMSFIELD ERNEST
Publication year - 1997
Publication title -
international journal for numerical methods in engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.421
H-Index - 168
eISSN - 1097-0207
pISSN - 0029-5981
DOI - 10.1002/(sici)1097-0207(19970315)40:5<953::aid-nme100>3.0.co;2-y
Subject(s) - discretization , mathematical analysis , truncation (statistics) , half space , mathematics , space (punctuation) , boundary (topology) , boundary element method , geometry , shear waves , finite element method , domain (mathematical analysis) , shear (geology) , physics , geology , computer science , petrology , statistics , thermodynamics , operating system
A method is described in this article to correct for the error that arises with the discretization of domains that include boundaries that extend to infinity. Typically, when these types of domains are discretized, part of the boundary is excluded from the calculation resulting in a truncated region. Of particular interest are seismic wave problems through zoned media since these types of problems typically have discretization schemes which result in truncated regions. In this article, the correction for incident SH waves is addressed. The correction for layer truncation has been investigated previously by Hadley et al. for SH waves in an elastic material and here it is extended to damped materials. For the truncation of the half‐space, an alternative approach to the enclosing element method proposed by Ahmad and Banerjee is described here. The solution for the half‐space correction described in this article is primarily analytic and therefore is a simple implementation. To prove the validity of the proposed technique, the method is checked by calculating the soil amplification of a unit SH wave through a soil layer on a rock half‐space. Since an analytic solution exists for this particular problem, the problem serves as a good basis to compare results with and without the truncation corrections. Numerical solutions for this problem are included considering (1) neither layer nor half‐space correction, (2) layer correction, but excluding the half‐space correction, and (3) both the layer and half‐space corrections. © 1997 by John Wiley & Sons, Ltd.