P.V. Sivaprasad

Artificial neural network modeling to evaluate and predict the deformation behavior of stainless steel type AISI 304L during hot torsion

The deformation behavior of type 304L stainless steel during hot torsion is investigated using artificial neural network (ANN). Torsion tests in the temperature range of 600-1200^oC and in the (maximum surface) strain rate range of 0.1-100s^-^1 were carried out. These experiments provided the required data for training the neural network and for subsequent testing. The input parameters of the model are strain, log strain rate and temperature while torsional flow stress is the output. A three layer feed-forward network was trained with standard back propagation (BP) and Resilient propagation (Rprop) algorithm. The paper makes a robust comparison of the performances of the above two algorithms. The network trained with Rprop algorithm is found to perform better and also needs less number of iterations for convergence. The developed ANN model employing this algorithm could efficiently track the work hardening, dynamic softening and flow localization regions of the deforming material. Sensitivity analysis showed that temperature and strain rate are the most significant parameters while strain affects the flow stress only moderately. The ANN model, described in this paper, is an efficient quantitative tool to evaluate and predict the deformation behavior of type 304L stainless steel during hot torsion.

Activation energy for serrated flow in a 15Cr5Ni Ti-modified austenitic stainless steel

Serrated flow in a 15Cr15Ni titanium-modified austenitic stainless steel has been investigated at a wide range of temperatures (300–1023 K) and strain rates (10−5−10−2s−1). Three regimes of serrated flow have been identified as low temperature type A (LT-A), high temperature type A (HT-A) and high temperature type C (HT-C). Different methods suggested in the literature for the determination of activation energy, Q were employed and Q was obtained aas 115 ± 9, 140 ± 7 and 178 ± 7 kJ/mol for LT-A, Ht-A and HT-C regimes respectively. Based on the Q values, the mechanisms responsible for the three regimes of serrated flow were identified as vacancy migration for LT-A, C (or N) as controlling diffusing species to form a pair with vacancy causing Schoeck-Seeger locking for HT-A regime and interaction of Ti with C (or N) for HT-C regime. Titanium was found to influence the serrated flow in all regimes.

Constitutive flow behaviour of austenitic stainless steels under hot deformation: artificial neural network modelling to understand, evaluate and predict

An artificial neural network (ANN) model is developed to predict the constitutive flow behaviour of austenitic stainless steels during hot deformation. The input parameters are alloy composition and process variables whereas flow stress is the output. The model is based on a three-layer feed-forward ANN with a back-propagation learning algorithm. The neural network is trained with an in-house database obtained from hot compression tests on various grades of austenitic stainless steels. The performance of the model is evaluated using a wide variety of statistical indices. Good agreement between experimental and predicted data is obtained. The correlation between individual alloying elements and high temperature flow behaviour is investigated by employing the ANN model. The results are found to be consistent with the physical phenomena. The model can be used as a guideline for new alloy development.

Evolution and characterization of dynamically recrystallized microstructure in a titanium-modified austenitic stainless steel using ultrasonic and EBSD techniques

A C-scan ultrasonic imaging system was used to investigate the microstructural evolution during dynamic recrystallization (DRX) of a 15Cr–15Ni–2.2Mo–Ti modified austenitic stainless steel (alloy D9). Four specimens were forged at 1273 K to different strains in the range 0.1–0.5. Specimens with true strains of 0.2 or lower did not show any variation in the amplitude of the first back-wall echo. However, a visible variation in the C-scan image was observed at and above the 0.3 strain level. This variation was attributed to the evolution of fine grains. The formation of fine grains was related to DRX, as indicated by electron backscattered diffraction. This study also revealed the characteristics of the DRX or ‘necklace grains’, as opposed to the so-called parent grains or rest of the microstructure.

Capability of a Feed-Forward Artificial Neural Network to Predict the Constitutive Flow Behavior of As Cast 304 Stainless Steel Under Hot Deformation

A model is developed to predict the constitutive flow behavior of as cast 304 stainless steel during hot deformation using artificial neural network (ANN). The inputs of the neural network are strain, strain rate, and temperature, whereas flow stress is the output. Experimental data obtained from hot compression tests in the temperature range 1023-1523 K, strain range 0.1-0.5, and strain rate range 10 -3 -10 2 s -1 are employed to develop the model. A three-layer feed-forward ANN is trained with standard back propagation and some upgraded algorithms like resilient propagation (Rprop) and superSAB. The performances of these algorithms are evaluated using a wide variety of standard statistical indices. The results of this study show that Rprop algorithm performs better as compared to others and thereby considered as the most efficient algorithm for the present study. It has been shown that the developed ANN model can efficiently and accurately predict the hot deformation behavior of as cast 304 stainless steel. Finally, an attempt has been made to quantify the extrapolation ability of the developed network.

Dynamic Recrystallization in a Ti Modified Austenitic Stainless Steel During High Strain Rate Deformation

The article discusses the dynamic recrystallization (DRX) behavior in a 15Cr-15Ni-2.2Mo-Ti modified austenitic stainless steel (commonly known as alloy D9) during high strain rate deformation. Hot compression tests were conducted in a Gleeble thermomechanical simulator in the temperatures range 1173 K–1373 K with a strain rate 10 and 100 s−1 to different strains. Hot deformed microstructures were characterized using Electron Back Scatter Diffraction (EBSD) technique. Recrystallized grains were partitioned from the deformed grains employing grain orientation spread approach. Extent of dynamic recrystallization (DRX) was found to be minimal at 1173 K. DRX was predominantly found to happen at and above 1273 K which increases with increase in strain. The fraction of Σ3 and Σ9 boundaries was found to increase with %DRX. The origin and the possible role of these boundaries on DRX in alloy D9 are also explained.

An Overview of Neural Network Based Modeling in Alloy Design and Thermomechanical Processing of Austenitic Stainless Steels

This overview reports some of the research works carried out by us in recent past on the application of artificial neural network (ANN) based modeling in alloy design and thermomechanical processing of austenitic stainless steels. Different ANN models were created in order to simulate various correlations and phenomena in austenitic stainless steels. These include: prediction of mechanical properties of alloy D9 from its alloy content, modeling constitutive flow behavior of austenitic stainless steels, and prediction of torsional flow behavior of type 304L stainless steel. Attempt has been made to explain the simulated results by relevant fundamental metallurgical phenomena.

Strain dependent constitutive analysis of hot deformation behaviour in 9Cr–1Mo ferritic steel

Abstract Following the sine – hyperbolic Arrhenius equation, constitutive analysis has been performed on true stress – strain data obtained from hot compression tests on plain 9Cr –1Mo steel over a wide range of temperatures (1223 – 1373 K) and strain rates (0.01 – 100 s–1). On incorporating the correction for shear modulus and diffusivity into this equation, the power-law plot exhibited a distinct deviation at higher stresses and was accounted for by considering the contribution from dislocation pipe diffusion. The correlation between stress, strain rate and temperature was found to follow the rate equation of the form /DL = constant[sinh(αLσ/G)]nh, where DL is lattice diffusivity, G is the shear modulus and, αL and nh are constants. The material constants were observed to be strain dependent and this was incorporated into the constitutive equation for predicting flow stress. The higher correlation coefficient (R = 0:996) and a lower average absolute relative error (3.264%) associated with prediction revealed that the developed strain dependent constitutive equation could predict flow stress over the investigated hot working domain.

Modeling constitutive behavior of a 15Cr-15Ni-2.2Mo-Ti modified austenitic stainless steel under hot compression using artificial neural network

In this paper, an artificial neural network (ANN) model has been suggested to predict the constitutive flow behavior of a 15Cr-15Ni-2.2Mo-Ti modified austenitic stainless steel under hot deformation. Hot compression tests in the temperature range 850°C1250°C and strain rate range 10-10 s were carried out. These tests provided the required data for training the neural network and for subsequent testing. The inputs of the neural network are strain, log strain rate and temperature while flow stress is obtained as output. A three layer feed-forward network with ten neurons in a single hidden layer and back-propagation learning algorithm has been employed. A very good correlation between experimental and predicted result has been obtained. The effect of temperature and strain rate on flow behavior has been simulated employing the ANN model. The results have been found to be consistent with the metallurgical trend. Finally, a monte carlo analiysis has been carried out to find out the noise sensitivity of the developed model. International Conference on Statistical Mechanics of Plasticity and Related Instabilities Indian Institute of Science, Bangalore August 29 – September 2, 2005

Multiscaling in Normal Grain Growth: A Monte Carlo Study

We describe Monte Carlo simulation of the phenomenon of normal grain growth employing q state Potts model. We find that at long times after quench the grain size distribution is multiscaling.