Modeling And Characterization Of High Carbon Nanobainitic Steels

Gaganpreet Sidhu

doi:10.32920/ryerson.14645598

Abstract

Analytical models have been developed for the transformation kinetics, microstructure analysis and the mechanical properties in bainitic steels. Three models are proposed for the bainitic transformation based on the chemical composition and the heat treatment conditions of the steel as inputs: (1) thermodynamic model on kinetics of bainite transformation, (2) improved thermo-statistical model that eliminates the material dependent empirical constants and (3) an artificial neural network model to predict the volume fraction of bainite. Neural networks have also been used to model the hardness of high carbon steels, subjected to isothermal heat treatment. Collectively, for a steel of given composition and subjected to a particular isothermal heat treatment, the models can be used to determine the volume fraction of bainitic phase and the material hardness values. The models have been extensively validated with the experimental data from literature as well as from three new high carbon experimental steels with various alloying elements that were used in the present work. For these experimental steels, data on the volume fraction of phases (via X-ray diffraction), yield strength (via compression tests) and hardness were obtained for various combinations of isothermal heat treatment times and temperatures. The heat treated steels were subjected to compression and hardness tests and the data have been used to develop a new correlation between the yield stress and the hardness. It was observed that while all three experimental steels exhibit a predominantly nanostructured bainite microstructure, the presence of Co and Al in one of the steels accelerated and maximized the nano-bainitic transformation within a reasonably short isothermal transformation time. Excellent yield strength (>1.7 GPa) and good deformability were observed in this steel after isothermal heat treatment at a low temperature of 250C for a relatively short duration of 24 hours.

Highlights

1.1 Background and MotivationThe ultimate goal of steel researchers is to obtain a product that is cheap to manufacture but at the same time meets certain requirements on mechanical properties
The validation of the model was performed with respect to the following progressively stringent criteria: (a) The model must work for the six steels that were used for the derivation of the coefficients in Eq (3.18) and (3.19), i.e., the model development data. (b) for these steels, the model must give good predictions at other isothermal transformation conditions whose data were not used for model development. (c) The model must be able to capture the transformations in steels when the parameters such as austenite grain size are varied
This is because it is known that with such variations, there is a change in the bainitic kinetics and thereby the volume fraction of bainite [43]. (d) the model must work for other steels that have a carbon content that is within the range of the carbon concentration of the steels used for the development of the model

Summary

Introduction

1.1 Background and MotivationThe ultimate goal of steel researchers is to obtain a product that is cheap to manufacture but at the same time meets certain requirements on mechanical properties. As a primary objective, in this chapter, an ANN approach has been presented to study the hardness of isothermally heat treated high-carbon bainitic steels containing as many as nine alloying elements. The key features of this model include: (1) a temperature dependent expression for the number density of potential nucleation sites Ni, and (2) a temperature dependent scaling function in the expression for volume fraction of bainite (f) to account for the incomplete reaction of austenite into bainite [1] This differentiates the current model from that of Van Bohemen, whose expression for f predicts 100% transformation requiring the experimental data to be normalized. Without the scaling function there would be a 100% transformation as is noticed in the model of Van Bohemen et al [37]

Objectives

Methods

Results

Discussion

Conclusion