A comparative assessment of cyclic deformation behaviour in SA333 Gr.6 steel using solid, hollow specimens under axial and shear strain paths

Highlights

• Examined cyclic deformation due to specimen geometry and strain path.

• C–Mn steel used for the study using solid and hollow specimens.

• Differences in the deformation behaviour has been noticed.

• Hysteresis loops and a probability density function (PDF) method employed for analyses.

• Explanation given through the banded microstructure of the steel investigated.

Abstract

Strain controlled cyclic deformation behaviour in a carbon–manganese steel has been studied using solid and thin walled tubular specimens. Both solid and tubular specimens were used for axial strain experiments, while only tubular specimens were used for shear strain experiments. Axial fatigue life in tubular specimens were lower than that in solid specimens. A probability density based analyses of hysteresis loop was employed to understand local yield level changes in solid/hollow specimens under axial or shear strain deformation. The observed difference in the local yield levels have been reasoned to the banded microstructure of the material.

Introduction

Conventionally, low cycle fatigue (LCF) behaviour of metallic materials is studied under uniaxial loading conditions using solid specimens. Such experiments are sufficient to gather the basic information on the cyclic deformation behaviour of materials. However, the actual in-service conditions in many engineering materials involve complex loading schemes and often data from a simple uniaxial experiment is inadequate to understand cyclic deformation behaviour under such situations. More attention has therefore been laid of late to study the cyclic deformation behaviour of materials under complex loading situations such as thermo mechanical fatigue or multiaxial fatigue using thin walled hollow specimens. One of the main focus here is the extension of the basic understanding on uniaxial fatigue to more complex situations through a parametric development of a damage parameter that relates the uniaxial and multiaxial situations. One argument put forth by Fash et al. [1] in this regard is that an appropriate establishment be obtained to prove that the data and damage behaviour using a smooth uniaxial specimen represents that of a thin wall tubular specimen. It thus calls for a comparative assessment of cyclic deformation behaviour in solid and hollow specimens. This paper is aimed to share some of our recent experimental observations on this issue.

Unfortunately, barring a few, not many investigations are available on this issue. Research by Fash et al. [1] and few others [2], [3] have shown that even after considering the influence of all multiaxiality factors, the fatigue lives in solid and hollow specimens are not the same. It has been reasoned that the difference basically arises due to the variations in the stress/strain gradients brought forth by the specimen geometry. In another investigation [4] on low cycle fatigue using tubular specimens of varying wall thickness, it has been noticed that fatigue lives decreased with decreasing wall thickness and in general, fatigue lives in solid specimens were higher when compared to the results from hollow specimens under identical experimental conditions. Evidences for a shorter fatigue life in hollow specimens than their companion solid geometry is also available in [2]. Through simulation experiments Shatil et al. [2] have shown that the difference in lives is connected to the manner in which the radial stress/strain gradient varies from surface to mid section in the two geometries. Recently, Keun and Soon Bok [5] have also observed a similar variation in the fatigue lives of tubular and solid specimens. Their experimental and simulation results show that the ratio of the internal plastic strain between hollow and solid specimens were consistently higher than unity indicating that the internal plastic strain energy density in hollow specimens is more and thus explaining the shorter live in these geometries.

All these investigations clearly bring out the fact that the fatigue lives in solid and hollow specimen geometries need not be essentially the same. While these investigations provide sufficient evidence for the same, no major attempt has been made so far to understand the influence that this geometry effect can have on the cyclic deformation process itself. It is this aspect that we are reporting here. In this work, we provide a comparative assessment of the cyclic deformation behaviour in a carbon–manganese steel using solid and thin walled tubular specimens in axial strain path. Additionally, we also compare the influence of axial and shear strain paths on the cyclic deformation process using tubular specimens.

The process of cyclic deformation can be best represented by the analyses of the stress–strain hysteresis loops. We have adopted this technique to understand the influence of specimen geometry and/or strain path on cyclic deformation process such as hardening/softening behaviour, Masing/non-Masing behaviour, alterations in the proportional limit. We also employ a probability density function (PDF) based analyses of the branches of the hysteresis loops to understand the local alterations in cyclic deformation. The overall objective is to provide a fundamental understanding of changes in cyclic plastic phenomena due to specimen geometry and/or strain path. The work is aimed at providing a knowledge base on this issue and will be helpful for further evaluation of various plasticity models.