scholarly journals Improving Prediction of Springback in Sheet Metal Forming Using Multilayer Perceptron-Based Genetic Algorithm

Materials ◽  
2020 ◽  
Vol 13 (14) ◽  
pp. 3129
Author(s):  
Tomasz Trzepieciński ◽  
Hirpa G. Lemu
Keyword(s):  
Genetic Algorithm ◽  
Multilayer Perceptron ◽  
Sheet Material ◽  
Rolling Direction ◽  
Back Propagation ◽  
Bending Test ◽  
Uniaxial Tensile ◽  
Multilayer Perceptrons ◽  
Input Parameters ◽  
Springback Coefficient

This paper presents the results of predictions of springback of cold-rolled anisotropic steel sheets using an approach based on a multilayer perceptron-based artificial neural network (ANN) coupled with a genetic algorithm (GA). A GA was used to optimise the number of input parameters of the multilayer perceptron that was trained using different algorithms. In the investigations, the mechanical parameters of sheet material determined in uniaxial tensile tests were used as input parameters to train the ANN. The springback coefficient, determined experimentally in the V-die air bending test, was used as an output variable. It was found that specimens cut along the rolling direction exhibit higher values of springback coefficient than specimens cut transverse to the rolling direction. An increase in the bending angle leads to an increase in the springback coefficient. A GA-based analysis has shown that Young’s modulus and ultimate tensile stress are variables having no significant effect on the coefficient of springback. Multilayer perceptrons trained by back propagation, conjugate gradients and Lavenberg–Marquardt algorithms definitely favour punch bend depth under load as the most important variables affecting the springback coefficient.

scholarly journals Modelling Anisotropic Phenomena of Friction of Deep-Drawing Quality Steel Sheets Using Artificial Neural Networks

2021 ◽  
Vol 21 (3) ◽  
pp. 31-42
Author(s):  
Tomasz Trzepieciński ◽  
Hirpa G. Lemu ◽  
Łukasz Chodoła ◽  
Daniel Ficek ◽  
Ireneusz Szczęsny
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Coefficient Of Friction ◽  
Multilayer Perceptron ◽  
Deep Drawing ◽  
Sheet Material ◽  
Rolling Direction ◽  
Steel Sheets ◽  
The Coefficient Of Friction ◽  
Artificial Neural

Abstract This paper presents a method of determining the coefficient of friction in metal forming using multilayer perceptron based on experimental data obtained from the pin-on-disk tribometer. As test material, deep-drawing quality DC01, DC03 and DC05 steel sheets were used. The experimental results show that the coefficient of friction depends on the measured angle from the rolling direction and corresponds to the surface topography. The number of input variables of the artificial neural network was optimized using genetic algorithms. In this process, surface parameters of the sheet, sheet material parameters, friction conditions and pressure force were used as input parameters to train the artificial neural network. Some of the obtained results have pointed out that genetic algorithm can successfully be applied to optimize the training set. The trained multilayer perceptron predicted the value of the friction coefficient for the DC04 sheet. It was found that the tested steel sheet exhibits anisotropic tribological properties. The highest values of the coefficient of friction under dry friction conditions were registered for sheet DC05, which had the lowest value of the yield stress. Prediction results of coefficient of friction by multilayer perceptron were in qualitative and quantitative agreement with the experimental ones.

scholarly journals Experimental and Computational analysis of springback in dual phase steels

2021 ◽  
Vol 26 (2) ◽  
pp. 137-145
Author(s):  
Rodolfo Rodríguez Baracaldo ◽  
Yeison Parra-Rodríguez ◽  
José Manuel Arroyo-Osorio
Keyword(s):  
Elastic Recovery ◽  
Plastic Anisotropy ◽  
Microstructural Analysis ◽  
Rolling Direction ◽  
Experimental Tests ◽  
Tensile Tests ◽  
Bending Test ◽  
Uniaxial Tensile ◽  
Dual Phase ◽  
Characteristic Region

In this work the comfortability of dual-phase automotive steel DP600 is studied through uniaxial tensile tests and V-die bending tests in different directions relative to the rolling direction. A microstructural analysis was also carried out in each characteristic region of the deformation zone, evidencing the changes in the morphology of the microstructure grains. Additionally, the plastic anisotropy of the material was studied by implementing the constitutive anisotropy models known as Hill-48 and Barlat-89. The results showed an increase in elastic recovery at 45 ° and 90 ° from the rolling direction. This variation can be attributed to the morphology of the martensite that created preferential location zones within the material during the rolling process. The two models Hill-48 and Barlat-89 correctly describe the yield surface and the plastic anisotropy obtained in the experimental tests carried out. The simulation using the finite element method and the Hill-48 model gave satisfactory results in the prediction of the elastic recovery as compared to the experimental results obtained with the V-die bending test.

scholarly journals Safety Warning Model of Coal Face Based on FCM Fuzzy Clustering and GA-BP Neural Network

Symmetry ◽  
2021 ◽  
Vol 13 (6) ◽  
pp. 1082
Author(s):  
Fanqiang Meng
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Early Warning ◽  
Back Propagation ◽  
Back Propagation Neural Network ◽  
Coal Face ◽  
Early Warning Model ◽  
Working Face ◽  
The Neural Network ◽  
Warning Model

Risk and security are two symmetric descriptions of the uncertainty of the same system. If the risk early warning is carried out in time, the security capability of the system can be improved. A safety early warning model based on fuzzy c-means clustering (FCM) and back-propagation neural network was established, and a genetic algorithm was introduced to optimize the connection weight and other properties of the neural network, so as to construct the safety early warning system of coal mining face. The system was applied in a coal face in Shandong, China, with 46 groups of data as samples. Firstly, the original data were clustered by FCM, the input space was fuzzy divided, and the samples were clustered into three categories. Then, the clustered data was used as the input of the neural network for training and prediction. The back-propagation neural network and genetic algorithm optimization neural network were trained and verified many times. The results show that the early warning model can realize the prediction and early warning of the safety condition of the working face, and the performance of the neural network model optimized by genetic algorithm is better than the traditional back-propagation artificial neural network model, with higher prediction accuracy and convergence speed. The established early warning model and method can provide reference and basis for the prediction, early warning and risk management of coal mine production safety, so as to discover the hidden danger of working face accident as soon as possible, eliminate the hidden danger in time and reduce the accident probability to the maximum extent.

scholarly journals Coupled Thermomechanical Response Measurement of Deformation of Nickel-Based Superalloys Using Full-Field Digital Image Correlation and Infrared Thermography

Materials ◽  
2021 ◽  
Vol 14 (9) ◽  
pp. 2163
Author(s):  
Krzysztof Żaba ◽  
Tomasz Trzepieciński ◽  
Sandra Puchlerska ◽  
Piotr Noga ◽  
Maciej Balcerzak
Keyword(s):  
Surface Temperature ◽  
Strain Rate ◽  
Rolling Direction ◽  
Image Correlation ◽  
Uniaxial Tensile ◽  
Uniaxial Tensile Test ◽  
Response Measurement ◽  
Sheet Rolling ◽  
Inconel Alloys ◽  
The Relationship

The paper is devoted to highlighting the potential application of the quantitative imaging technique through results associated with work hardening, strain rate and heat generated during elastic and plastic deformation. The aim of the research presented in this article is to determine the relationship between deformation in the uniaxial tensile test of samples made of 1-mm-thick nickel-based superalloys and their change in temperature during deformation. The relationship between yield stress and the Taylor–Quinney coefficient and their change with the strain rate were determined. The research material was 1-mm-thick sheets of three grades of Inconel alloys: 625 HX and 718. The Aramis (GOM GmbH, a company of the ZEISS Group) measurement system and high-sensitivity infrared thermal imaging camera were used for the tests. The uniaxial tensile tests were carried out at three different strain rates. A clear tendency to increase the sample temperature with an increase in the strain rate was observed. This conclusion applies to all materials and directions of sample cutting investigated with respect to the sheet-rolling direction. An almost linear correlation was found between the percent strain and the value of the maximum surface temperature of the specimens. The method used is helpful in assessing the extent of homogeneity of the strain and the material effort during its deformation based on the measurement of the surface temperature.

scholarly journals Classification of ETM+ Remote Sensing Image Based on Hybrid Algorithm of Genetic Algorithm and Back Propagation Neural Network

2013 ◽  
Vol 2013 ◽  
pp. 1-8 ◽  
Author(s):  
Haisheng Song ◽  
Ruisong Xu ◽  
Yueliang Ma ◽  
Gaofei Li
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Genetic Algorithm ◽  
Hybrid Algorithm ◽  
Local Minimum ◽  
Back Propagation ◽  
Remote Sensing Image ◽  
Back Propagation Neural Network ◽  
Bp Algorithm ◽  
Improved Bp Algorithm

The back propagation neural network (BPNN) algorithm can be used as a supervised classification in the processing of remote sensing image classification. But its defects are obvious: falling into the local minimum value easily, slow convergence speed, and being difficult to determine intermediate hidden layer nodes. Genetic algorithm (GA) has the advantages of global optimization and being not easy to fall into local minimum value, but it has the disadvantage of poor local searching capability. This paper uses GA to generate the initial structure of BPNN. Then, the stable, efficient, and fast BP classification network is gotten through making fine adjustments on the improved BP algorithm. Finally, we use the hybrid algorithm to execute classification on remote sensing image and compare it with the improved BP algorithm and traditional maximum likelihood classification (MLC) algorithm. Results of experiments show that the hybrid algorithm outperforms improved BP algorithm and MLC algorithm.

COMPUTING AN ADAPTIVE MESH IN FLUID PROBLEMS USING NEURAL NETWORK AND GENETIC ALGORITHM WITH ADAPTIVE RELAXATION

2008 ◽  
Vol 17 (06) ◽  
pp. 1089-1108 ◽  
Author(s):  
NAMEER N. EL. EMAM ◽  
RASHEED ABDUL SHAHEED
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Numerical Experiments ◽  
Back Propagation ◽  
Adaptive Mesh ◽  
Learning System ◽  
Back Propagation Algorithm ◽  
Adaptive Smoothing ◽  
Different Types ◽  
Optimal Values

A method based on neural network with Back-Propagation Algorithm (BPA) and Adaptive Smoothing Errors (ASE), and a Genetic Algorithm (GA) employing a new concept named Adaptive Relaxation (GAAR) is presented in this paper to construct learning system that can find an Adaptive Mesh points (AM) in fluid problems. AM based on reallocation scheme is implemented on different types of two steps channels by using a three layer neural network with GA. Results of numerical experiments using Finite Element Method (FEM) are discussed. Such discussion is intended to validate the process and to demonstrate the performance of the proposed learning system on three types of two steps channels. It appears that training is fast enough and accurate due to the optimal values of weights by using a few numbers of patterns. Results confirm that the presented neural network with the proposed GA consistently finds better solutions than the conventional neural network.

Sign in / Sign up

Export Citation Format

Share Document