Time filter

Source Type

Hyderabad, India

Singh S.K.,GRIET | Gupta A.K.,BITS Pilani
CIRP Journal of Manufacturing Science and Technology | Year: 2010

In this paper, a new data mining technique support vector regression (SVR) is applied to predict the thickness along cup wall in hydro-mechanical deep drawing. After using the experimental results for training and testing, the model was applied to new data for prediction of thickness strains in hydro-mechanical deep drawing. The prediction results of SVR are compared with that of artificial neural network (ANN), finite element (FE) simulation and the experimental observations. The results are promising. It is found that SVR predicts the thickness variation in the drawn cups very accurately especially in the wall region. © 2010 CIRP. Source

Gupta A.K.,BITS Pilani | Krishnamurthy H.N.,BITS Pilani | Singh Y.,BITS Pilani | Prasad K.M.,BITS Pilani | Singh S.K.,GRIET
Materials and Design | Year: 2013

The experimental stress-strain data from isothermal tensile tests over a wide range of temperatures (623-923K at an interval of 50K), strains (0.02-0.30 at an interval of 0.02) and strain rates (0.0001, 0.001, 0.01, 0.1s-1) were employed to determine the Dynamic Strain Aging (DSA) regime and to formulate a suitable constitutive model to predict the elevated-temperature deformation behavior in DSA regime of Austenitic Stainless Steel (ASS) 304. Four models, namely, Johnson Cook (JC) model, modified Zerilli-Armstrong (m-ZA) model, modified Arrhenius type equations (m-Arr) and Artificial Neural Networks (ANNs), were investigated. Suitability of these models was evaluated by comparing the correlation coefficient, average absolute error and its standard deviation. It was observed that JC, m-ZA and m-Arr model could not effectively predict flow stress behavior of ASS304 in DSA regime, while the predictions by ANN model are found to be in good agreement with the experimental data. © 2012 Elsevier Ltd. Source

Gupta A.K.,BITS Pilani | Anirudh V.K.,BITS Pilani | Singh S.K.,GRIET
Materials and Design | Year: 2013

Strain, strain rate and temperature have a significant impact on the flow stress of a material. To study the impact of these factors on flow stress, quite a few empirical, semi-empirical constitutive models have been reported. In this work, four such models are being presented for estimation of flow stress on Austenitic Stainless Steel 316. While the Johnson Cook model, Modified Zerilli-Armstrong model are semi-empirical models, the Arrhenius type equation is a physical based equation. The Artificial Neural Network model on the other hand is trained based on the training data and employed to predict the flow stress on the testing data. The experiments for these data were conducted at various strain rates (0.1-0.0001s-1) and at various temperatures (323-623K). Values of stress were taken at strain intervals of 0.05 (from 0.05 to 0.3) to evaluate the material constants of the constitutive models. A comparative study on the reliability of the four models has also been made at the end. The correlation coefficient values observed were 0.9423 (JC model), 0.9879 (modified ZA model), 0.9852 (modified Arrhenius type equation) and 0.9930 (ANN model). © 2012 Elsevier Ltd. Source

Kotkunde N.,BITS Pilani | Krishnamurthy H.N.,BITS Pilani | Puranik P.,BITS Pilani | Gupta A.K.,BITS Pilani | Singh S.K.,GRIET
Materials and Design | Year: 2014

A reliable and accurate prediction of flow behavior of metals in industrial forming process considering the coupled effects of strain, strain rate and temperature is crucial in understanding the workability of the metal and optimizing parameters for hot forming process. In this study, the tensile fracture behavior of the Ti-6Al-4V alloy is examined with scanning electron microscope (SEM) over the range of magnifications. SEM study revealed that microvoids and shallow dimples are observed at the fracture surface which indicates the fracture is predominately ductile in nature. Also, an investigation on flow behavior of Ti-6Al-4V alloy is done using constitutive models. Four constitutive models; modified Johnson-Cook (m-JC), modified Arrhenius type equations (m-Arr), modified Zerilli-Armstrong (m-ZA) and Rusinek-Klepaczko (RK) models are developed to predict the flow stress. The predictions of these constitutive models are compared with each other using statistical measures like correlation coefficient, average absolute error and its standard deviation. Comparing the statistical measures, m-Arr model is a better model for predicting the flow stress, but considering the fact that m-ZA model is a physical based model, m-ZA model is preferred over the m-Arr model. © 2013 Elsevier Ltd. Source

Singh S.K.,GRIET
International Journal of Material Forming | Year: 2010

Anisotropy and strain rate sensitivity index (m) plays a very important role in the formability of materials. In the present investigation strain ratios in 0°, 45° and 90° to the rolling direction and the strain rate sensitivity index were calculated at different temperatures. After developing the data from experiments, Artificial Neural Network (ANN) models are trained for different properties. Trained ANN models are used to calculate different strain ratios and sensitivity index at unknown temperatures. The results are promising and the percentage error in ANN prediction is found to be around 10%. © 2010 Springer/ESAFORM. Source

Discover hidden collaborations