A framework in which computational engineering data is visualised and manipulated using parallel processing techniques is presented. Obviously, the parallelisation of the computational simulation process has started by focusing on the most compute-intensive stages first. In the visualisation field, the most compute-intensive process is the solution post-processing stage. Even for a reasonably small mesh (500,000-2 million tetrahedral elements), a number of post-processing techniques (for example, calculating isosurfaces, flow streamlines, contours, etc.) can require a large amount of processing power. In order to produce graphically interactive displays a number of techniques have been used to distribute the processing amongst more than one processor. However, when applying computational simulation techniques to complex aerospace configurations the sheer size of the meshes poses extra problems for the visualisation process. Today, meshes ranging from 5 to 20 million elements are commonplace, with 50-100 million elements becoming possible. The memory and processing requirements of meshes of this size means that producing interactive displays using a single graphics workstation is virtually impossible. A number of techniques are investigated and a prototype implementation is demonstrated.

purchase the full-text of this paper (price £20)

go to the next paper
return to the table of contents
return to the book description
purchase this book (price £56 +P&P)