Advanced

Predictive design analysis utilizing ansys in an internet environment

Eriksson, Martin LU and Burman, Åke LU (2000) ANSYS 2000 Technical Conference
Abstract
During recent years the increasing power of computers and communications has led to the development and use of more advanced analysis techniques. Design of Experiment (DOE) has been used together with Finite Element (FE) analysis software such as ANSYS, see e.g. Johnson and Heald (1998) and Eriksson et al. (1998). Predictive Design Analysis (PDA) is an mechanical engineering design, or design for short, approach in which the final behavior of a product is predicted, at least to some extent, by combining information from different analysis techniques, see Eriksson and Burman (1999). This work presents an implementation of PDA consisting of DOE together with ANSYS.

The implementation is developed as a World Wide Web (WWW)... (More)
During recent years the increasing power of computers and communications has led to the development and use of more advanced analysis techniques. Design of Experiment (DOE) has been used together with Finite Element (FE) analysis software such as ANSYS, see e.g. Johnson and Heald (1998) and Eriksson et al. (1998). Predictive Design Analysis (PDA) is an mechanical engineering design, or design for short, approach in which the final behavior of a product is predicted, at least to some extent, by combining information from different analysis techniques, see Eriksson and Burman (1999). This work presents an implementation of PDA consisting of DOE together with ANSYS.

The implementation is developed as a World Wide Web (WWW) Client/Server application. The user interface is constructed as a standard HTML page that can be viewed in a standard web. Java scripts and Java Applets are used to verify the data given by the designers or analysts before it is sent to a Common Gateway Interface (CGI) application on the server for evaluation. The result delivered from the CGI application is VRML files of selected responses along with GIF images and data based on the statistical evaluation of the studied responses.

The issue of decreasing the computation time is addressed in this work by running the FE analyses in parallel. The fact that all analyses based on DOE are performed independently of each other makes them well suited for distributed processing, for example on a cluster of computers connected through a network. Studies of the implementation regarding the timesaving of distributed computations are performed and documented. This application serves as a powerful tool for the designer or analyst at all levels of abstraction in the design process, when examining the influence of different design variables on the product studied. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to conference
publication status
published
subject
keywords
Finite Element Analysis, Client/Server, Design Of Experiment, Statistics, Distributed Computing, DOE, WWW, FEA, Parallel processing, Internet/intranet
pages
10 pages
conference name
ANSYS 2000 Technical Conference
language
English
LU publication?
yes
id
d11048ad-6142-409c-8289-4c2b8f9bbd5c (old id 1024347)
date added to LUP
2008-02-04 12:43:10
date last changed
2016-04-16 10:59:16
@misc{d11048ad-6142-409c-8289-4c2b8f9bbd5c,
  abstract     = {During recent years the increasing power of computers and communications has led to the development and use of more advanced analysis techniques. Design of Experiment (DOE) has been used together with Finite Element (FE) analysis software such as ANSYS, see e.g. Johnson and Heald (1998) and Eriksson et al. (1998). Predictive Design Analysis (PDA) is an mechanical engineering design, or design for short, approach in which the final behavior of a product is predicted, at least to some extent, by combining information from different analysis techniques, see Eriksson and Burman (1999). This work presents an implementation of PDA consisting of DOE together with ANSYS. <br/><br>
The implementation is developed as a World Wide Web (WWW) Client/Server application. The user interface is constructed as a standard HTML page that can be viewed in a standard web. Java scripts and Java Applets are used to verify the data given by the designers or analysts before it is sent to a Common Gateway Interface (CGI) application on the server for evaluation. The result delivered from the CGI application is VRML files of selected responses along with GIF images and data based on the statistical evaluation of the studied responses. <br/><br>
The issue of decreasing the computation time is addressed in this work by running the FE analyses in parallel. The fact that all analyses based on DOE are performed independently of each other makes them well suited for distributed processing, for example on a cluster of computers connected through a network. Studies of the implementation regarding the timesaving of distributed computations are performed and documented. This application serves as a powerful tool for the designer or analyst at all levels of abstraction in the design process, when examining the influence of different design variables on the product studied.},
  author       = {Eriksson, Martin and Burman, Åke},
  keyword      = {Finite Element Analysis,Client/Server,Design Of Experiment,Statistics,Distributed Computing,DOE,WWW,FEA,Parallel processing,Internet/intranet},
  language     = {eng},
  pages        = {10},
  title        = {Predictive design analysis utilizing ansys in an internet environment},
  year         = {2000},
}