Local uniform mesh refinement on loosely-coupled parallel processors (Q1113648)

Local uniform mesh refinement (LUMR) is a powerful technique for the solution of partial differential equations where uniform refinement grids are slaced on a coarse `parent grid'. This paper discusses the implementation of LUMR on a system of loosely coupled parallel processors. The processors are assumed to be connected either by a common bus of local area network of in a linear array and the cost of communication is estimated. Finally the theoretical results are compared with experiments with the revolving cone problem on a 14 node Apollo ring.