A Method to Predict the Sheet Metal Deformation under Different Locator Layout

Abstract:

"N-2-1" principle is widely recognized in the fixture design for deformable sheet metal workpieces, where N, the locators on primary datum, is the key to sheet metal fixture design. However, little research is done on how to determine the positions of N locators when considering normal deviation. This paper, concerning the low efficiency of the locating point optimization method using the finite element model, proposes a GA-BP neural network to predict the sheet metal deformation under different locator layout, by which to calculate the clamp normal deviation. Finally, a case is used to illustrate the flowchart of GA-BP network. Results show that the method has good performance and prediction ability.

• Periodical: Applied Mechanics and Materials (Volumes 395-396)
• Pages: 1146-1149
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.395-396.1146
• Online since: September 2013
• Authors: Hong Wang Zhao, Cong Lu, Wan Yu Li
• Keywords: Deformation, GA-BP Neural Network, Locators, Prediction

References

[1] W. Cai, S.J. Hu and J.X. Yuan: Journal of Manufacturing Science and Engineering, 1996, 118(8), 318-324.

[2] J.A. Camelio., S.J. Hu, D. Ceglarek.: Journal of Manufacturing Systems, 2004, 23(3), 182-193.

[3] B. Li, B.W. Shiu: ASME Journal of Manufacturing Science and Engineering, 2003, 125(1): 120-127.

[4] Y. Xing, S. Jin, X. Lai: Journal of Computer-Aided Design & Computer Graphics, 2009, 21(1): 101-106 (In Chinese).

[5] Y. Xing, Y. Wang: Journal of Computer-Aided Design & Computer Graphics, 2012, 24(5): 677-682 (In Chinese).