Machine learning for the improvement of springback modelling

Dezelak, M.; Pahole, Ivo; Ficko, Mirko; Brezočnik, Miran

doi:10.14743/apem2012.1.127

Cited by 8 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…al. studied that by this ML algorithm there is possibility of improving springback prediction, recording all successful solution of springback reduction and compensation [37]. Jabbari and Shokoohi investigated a new method for die angle optimization of V shape bending by using reduced basis technique.…”

Section: Optimizationmentioning

confidence: 99%

A Review on Spring Back and Optimization Process

2018

IJMTER

View full text Add to dashboard Cite

Abstract-Recently forming industries faces spring back drawback in forming process like cold drawing, sheet metal forming, bending etc. These forming process resulted permanent deformation by cold forming to create different shapes. Spring back defect taking place during the forming process has been emphasized. The solution to such defect problems was also given. Literature review shows that considerable amount of researches have been conducted in the field of material forming and spring-back optimization process. The present review focuses on the cold forming processes, spring back defect, factors affecting spring back, the various methods used to analyze and optimize the spring back free process. This review also opens future research work to get dimensional stability and least springback.

show abstract

Section: Optimizationmentioning

confidence: 99%

A Review on Spring Back and Optimization Process

2018

IJMTER

View full text Add to dashboard Cite

show abstract

“…S PA (10) is the average value of the products for the each analysed point deviation -differences between the predicted and actual values. S P (11) and S A (12) are the mean-squared errors of the predicted and actual values. The correlation coefficient C C ranges from 1 for perfectly correlated results, through to 0 when there is no correlation, to -1 when the results are perfectly correlated negatively.…”

Section: Fem Simulationmentioning

confidence: 99%

“…The main goal today is to minimise the number of physical iterations of the forming tool's adaptations in order to achieve geometrical accuracy of the product and sufficient robustness of the production process [4]. A number of researchers during the past twenty years have tried to understand the background of the springback phenomenon and wanted to find ways of overcoming or at least controlling it [5][6][7][8][9][10][11][12][13][14]. The introduction of the new Advanced High Strength Steel (AHSS) into the automotive industry has made springback issues even more prominent.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of Twist Springback Prediction After an AHSS Forming Process

Dezelak¹,

Stepisnik²,

Pahole³

2014

Int. j. simul. model.

Self Cite

View full text Add to dashboard Cite

The springback of a sheet metal part is the change of its shape after removing a forming tool. An accurate prediction of springback is very difficult because sheet metal undergoes a complicated deformation process during the forming process. These are the following types of springback when considering the geometry of the product and the forming regime: angular change, sidewall curl, and twist. The latter is the key issue of this paper. A new definition of twist springback is proposed having been tested on a referenced sheet metal part. An experimental device for twist springback testing has been designed and the effect of blank rolling direction on the twist was investigated for dual-phase (DP) sheet steel. Finite element method (FEM) results of twisting behaviour using AutoForm software for different material models were compared with the experimental and the correlation evaluation was performed.

show abstract

“…In this case, the numerical simulations using FEM were performed to obtain the feed data required to train the ANN. Likewise, Dezelak et al 26 studied the combination of FEM, ANN and multiple linear regression to predict the spring-back in a sheet metal forming process.…”

Section: Dm Techniques For Modelling Non-linear Problemsmentioning

confidence: 99%

Characterization of electrolytic tinplate materials via combined finite element and regression models

Illera

Lostado

Martinez

et al. 2014

The Journal of Strain Analysis for Engineering Design

View full text Add to dashboard Cite

In this article, a new method that combines finite element method with data mining techniques is proposed to obtain the mechanical properties of electrolytic tinplate. Using information provided by two simple and economic tests (hardness and spring-back), already used in industries to classify tinplate, yield stress and tensile parameters of a generic electrolytic tinplate can be estimated. Initially, a group of finite element models based on these simple tests were built and validated against experimental data. The validated finite element models were then used to investigate the effect of different thicknesses and electrolytic tinplate plastic hardening parameters. With the convergent results obtained from these finite element simulations, a database was generated with the new electrolytic tinplate properties. Various types of regression models (model trees, artificial neural networks and support vector machines) based on data mining techniques were used to obtain the yield stress and plastic hardening parameters from a generic sample of electrolytic tinplate. The accuracy of the results demonstrates that this new method may be used to economically predict yield stress and plastic hardening parameters of a generic electrolytic tinplate.

show abstract

Machine learning for the improvement of springback modelling

Cited by 8 publications

References 9 publications

A Review on Spring Back and Optimization Process

A Review on Spring Back and Optimization Process

Evaluation of Twist Springback Prediction After an AHSS Forming Process

Characterization of electrolytic tinplate materials via combined finite element and regression models

Contact Info

Product

Resources

About