Share Email Print

Proceedings Paper

Approach to modeling the spray-forming process with artificial neural networks
Author(s): M. Allen Matteson; R. D. Payne; Craig Madden; A. L. Moran
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

In this study artificial neural networks were used to model the spray forming process. Networks were developed and trained using process parameter and product quality data collected from a series of five spray forming runs. Process parameters of time into run, melt temperature, and gas to metal ratio were used as inputs and the networks were trained to predict the corresponding values of exhaust gas temperature, preform surface roughness, and porosity in the product. These networks were then tested with actual and hypothetical data. The results of the study showed that the networks can determine relationships between process parameters and the end product quality. It was also shown that the networks can be used to predict the effect on product quality from changes in process parameters. Additional work is in progress to create a larger data set for training over a broader region of the operating envelope. The result of this ongoing work will provide greater reliability in network prediction.

Paper Details

Date Published: 1 May 1994
PDF: 10 pages
Proc. SPIE 2189, Smart Structures and Materials 1994: Smart Materials, (1 May 1994); doi: 10.1117/12.174079
Show Author Affiliations
M. Allen Matteson, Naval Surface Warfare Ctr. (United States)
R. D. Payne, Naval Surface Warfare Ctr. and Johns Hopkins Univ. (United States)
Craig Madden, Naval Surface Warfare Ctr. (United States)
A. L. Moran, Naval Surface Warfare Ctr. (United States)

Published in SPIE Proceedings Vol. 2189:
Smart Structures and Materials 1994: Smart Materials
Vijay K. Varadan, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?