scholarly journals Parameter identification of a mechanical ductile damage using Artificial Neural Networks in sheet metal forming

2013 ◽  
Vol 45 ◽  
pp. 605-615 ◽  
Author(s):  
Fethi Abbassi ◽  
Touhami Belhadj ◽  
Sébastien Mistou ◽  
Ali Zghal
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Parameter Identification ◽  
Sheet Metal ◽  
Sheet Metal Forming ◽  
Metal Forming ◽  
Ductile Damage ◽  
Artificial Neural

Fracture in sheet metal forming: Effect of ductile damage evolution

2007 ◽  
Vol 85 (3-4) ◽  
pp. 205-212 ◽  
Author(s):  
M. Khelifa ◽  
M. Oudjene ◽  
A. Khennane
Keyword(s):  
Sheet Metal ◽  
Sheet Metal Forming ◽  
Metal Forming ◽  
Damage Evolution ◽  
Ductile Damage

Numerical Prediction of Forming Limit in Hemispherical Punch Bulging by Lemaitre's Ductile Damage Model

2011 ◽  
Vol 110-116 ◽  
pp. 1437-1441 ◽  
Author(s):  
Farhad Haji Aboutalebi ◽  
Mehdi Nasresfahani
Keyword(s):  
Sheet Metal ◽  
Sheet Metal Forming ◽  
Metal Forming ◽  
Damage Mechanics ◽  
Thin Sheet ◽  
Damage Model ◽  
Ductile Damage ◽  
Forming Limit ◽  
St14 Steel ◽  
Hemispherical Punch

Prediction of sheet metal forming limits or analysis of forming failures is a very sensitive problem for design engineers of sheet forming industries. In this paper, first, damage behaviour of St14 steel (DIN 1623) is studied in order to be used in complex forming conditions with the goal of reducing the number of costly trials. Mechanical properties and Lemaitre's ductile damage parameters of the material are determined by using standard tensile and Vickers micro-hardness tests. A fully coupled elastic-plastic-damage model is developed and implemented into an explicit code. Using this model, damage propagation and crack initiation, and ductile fracture behaviour of hemispherical punch bulging process are predicted. The model can quickly predict both deformation and damage behaviour of the part because of using plane stress algorithm, which is valid for thin sheet metals. Experiments are also carried out to validate the results. Comparison of the numerical and experimental results shows good adaptation. Hence, it is concluded that finite element analysis in conjunction with continuum damage mechanics can be used as a reliable tool to predict ductile damage and forming limit in sheet metal forming processes.

Multiple Regression and Neural Network Based Characterization of Friction in Sheet Metal Forming

2014 ◽  
Vol 1051 ◽  
pp. 204-210 ◽  
Author(s):  
Hirpa G. Lemu ◽  
Tomasz Trzepieciński
Keyword(s):  
Neural Networks ◽  
Regression Analysis ◽  
Friction Coefficient ◽  
Multiple Regression ◽  
Sheet Metal ◽  
Sheet Metal Forming ◽  
Metal Forming ◽  
Process Conditions ◽  
Wide Range ◽  
Friction Models

This article proposes a frictional resistance description approach in sheet metal forming and the objective is to characterize the friction coefficient value under a wide range of friction conditions without performing time-consuming experiments. To describe the friction condition in sheet metal forming simulations, the friction coefficient should be quantified using friction models. Realistic friction models must account for the influence of surface roughness and surface topography on the lubricant flow and dry friction conditions. Due to considerable amount of factors that affect the friction coefficient value, building the analytical friction model for specified process conditions is too demanding. Thus, mathematical models that describe the friction behaviour using multiple regression analysis (MRA) and artificial neural networks (ANN) are utilized. The regression analysis was performed using the user subroutine in the MATLAB, while the ANN model was built in STATISTICA Neural Networks. As input variables for regression model and training of multilayer perceptron (MLP) the results of strip drawing friction test were utilized.

Sign in / Sign up

Export Citation Format

Share Document