Skip navigation
Please use this identifier to cite or link to this item: http://repository.iitr.ac.in/handle/123456789/4232
Title: ANN based prediction of moment coefficients in slabs subjected to patch load
Authors: Giri V.
Upadhyay A.
Published in: Structural Engineering and Mechanics
Abstract: The problem and its solution of a computer based design procedure for patch loaded interior panel design, in the context of civil engineering, are presented. An Artificial Neural Network (ANN) is trained to predict moment coefficient in short span m 1 for interior panel is proposed. Another ANN is also trained for moment coefficient in long span m2 value. The Bending moment coefficients m1 and m2 for different values of K are calculated and presented in the form of curves. The development of this ANN modeling involves determination of the training set, ANN architecture, training and testing of the network, and generalization capabilities. According to Pigeaud's method the moment along the shorter and longer directions depends on the ratios K, u/B, and v/L. The developed ANN architecture is computationally efficient and has been successfully used in the approximation of Pigeaud's curves.
Citation: Structural Engineering and Mechanics(2006), 24(4): 509-514
URI: https://doi.org/10.12989/sem.2006.24.4.509
http://repository.iitr.ac.in/handle/123456789/4232
Issue Date: 2006
Publisher: Techno Press
ISSN: 12254568
Author Scopus IDs: 57213715712
13608643500
Author Affiliations: Giri, V., Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee, 247 667, India
Upadhyay, A., Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee, 247 667, India
Corresponding Author: Upadhyay, A.; Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee, 247 667, India; email: akhilfce@iitr.ernet.in
Appears in Collections:Journal Publications [CE]

Files in This Item:
There are no files associated with this item.
Show full item record


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.