Finite Element Processing

Finite element analysis consists of several steps from creating geometric coordinate data to visualizing the analysis results. The core of the finite element analysis is to form the system equations using the data pre p a red as described in the previous chapters, and to solve them. This procedure is called "Processing of finite element analysis," and termed here as "Einite element processing." Among all the procedures in finite element analysis, this is the one which requires minimum user interaction. Once the processing begins, it goes on continuously by itself to the end without requiring user intervention. However, the processing is the step which involves the most intensive computation. This usually demands huge computational resources in terms of computing time and memory space. The computing time for processing varies largely depending on the size of the problem: from less than a second to several hours or more. The computing time as well as the memory space required for finite element processing increases drastically as the size of the problem becomes larger, in going from 2-D to 3-D, linear to nonlinear, static to dynamic. Computational efficiency in terms of computing time and memory space is an important issue in the finite element processing. There are a number of factors affecting computational efficiency. Either node numbering or element numbering depending on the solution scheme is one of the major factors determining the efficiency. Optimizing the node numbering or the element numbering is an essential step which should be taken prior to the processing. Accuracy as well as computational efficiency is affected also by integration schemes and element types employed in the processing. Therefore, they should be manipulated by the users for the best accuracy and efficiency. This chapter describes the usage of functions related with finite element processing and its computational efficiency. These functions are arranged as items of menu as shown in the figure below.