Springback analysis in air bending process through experiment based artificial neural networks

Özgu Şenol, Volkan Esat, Haluk Darendeliler

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Sheet metal bending is one of the most frequently used sheet metal forming processes in manufacturing industry. This study investigates bending parameters and springback phenomenon of a stainless-steel sheet in air bending process. In most of the applications, springback is determined either by trial and error procedures or by using numerical methods. Artificial Neural Network (ANN) approach has proved to be a helpful tool for the engineers. ANN is used in this study to predict the springback amounts of stainless steel sheets through experiment based networks. Air bending process is first modeled and analyzed by a commercial finite element code. Springback amounts for different sheet thicknesses and bend angles are computed. In addition to computational modeling, experimentation of the air bending processes is carried out and experimental results are used in artificial neural network development to show the feasibility of ANN based on experimentation. Experimental outcome is also used for validation of the FE analysis of the process, which demonstrates good agreement. It is observed that ANN can be applied effectively to determine springback in air bending process, which embodies significant potential to determine air bending process parameters for industrial applications such as punch stroke.

Original languageEnglish
Title of host publicationICTP 2014 - 11th International Conference on Technology of Plasticity, Nagoya, Japan
Subtitle of host publicationProcedia Engineering
Place of PublicationISSN: 1877-7058
Pages999-1004
Number of pages6
Volume81
DOIs
Publication statusPublished - 2014
Externally publishedYes
Event11th International Conference on Technology of Plasticity, ICTP 2014 - Nagoya, Japan
Duration: 19 Oct 201424 Oct 2014

Conference

Conference11th International Conference on Technology of Plasticity, ICTP 2014
Country/TerritoryJapan
CityNagoya
Period19/10/1424/10/14

Keywords

  • Air bending
  • Finite element method
  • Neural network

Fingerprint

Dive into the research topics of 'Springback analysis in air bending process through experiment based artificial neural networks'. Together they form a unique fingerprint.

Cite this