ABSTRACT

    Premature fatigue fractures in structural components are a major problem in the manufacturing industry. The challenge for modellers has been to deliver reliable fatigue-analysis tools, because over-designing components is becoming an increasingly unattractive solution to the problem. Currently software packages exist for fatigue simulation of components or systems. However, a common feature of such software is that they all require the fatigue properties of the materials used. When such information is not available, the fatigue simulation cannot proceed until relevant experimental measurements are carried out, which can be both time-consuming and very costly. It is the aim of the current work to help solve this dilemma by developing models that can calculate the strain-life relationship not only at room temperature but also high temperatures. This work extends previous successful models for predicting the monotonic material properties of commercial alloys as a function of alloy chemistry, heat treatment, temperature and strain rate. In the present paper, attempts are made to model the high temperature fatigue properties of some engineering alloys. The effect of strain rate and cyclic loading frequency on fatigue properties are also discussed.

NOMENCLATURE

Δε Total strain range in axial fatigue test

Δεe Elastic strain range in axial fatigue test

Δεp Plastic strain range in axial fatigue test

N Number of cycles to failure

σf' Axial fatigue strength coefficient

εf' Axial fatigue ductility coefficient

b Axial fatigue strength exponent

c Axial fatigue ductility exponent

K' Cyclic strain hardening coefficient

n' Cyclic strain hardening exponent

n monotonic hardening exponent

E Young’s modulus

εf Fracture ductility

σu Ultimate tensile strength

σy Yield strength

ν Frequency of loading in fatigue test

INTRODUCTION

    Premature fatigue fractures in structural components is a major industrial problem. It is often said that 80~90% of all the structural failures occur through a fatigue mechanism (1). The big challenge for modellers has been to deliver reliable fatigue-analysis tools because over-designing com-ponents is becoming an increasingly unattractive solution to the problem. Software packages that can be used for fatigue simulation of compon-ents or systems are available, but they all require the fatigue properties of the materials used as inputs, such as stress-life (S-N) or strain-life (ε-N) curves. However, it is often difficult to gain access to measured cyclic properties as the number of alloys for which such information available is limited, and the chemical composition, heat treatment and microstructure will all change the way in which the material responses to cyclic loading (2,3). Therefore it becomes problematic to experimentally measure all the required cyclic properties for generalised use.

    One way to consider solving this problem is to relate cyclic properties to monotonic tensile properties. If estimation methods with reasonable accuracy can be established, they can serve to provide fast solutions to fatigue problems without the time and cost involved in fatigue testing. Therefore, much effort has been put into finding such methods (4,5,6). All such attempts have been empirically based, providing some approxi-mations that can be useful. However, the ultimate way to solve this problem is through computer modelling where fatigue properties can be calculated as a function of alloy chemistry, processing details and working environment. This would be a significant step towards "true" virtual engineering design, where the design of components/systems and alloy composition/processing route are combined.

    The first step of the present modelling approach is therefore to calculate the monotonic properties, including yield/tensile strength, hardness, Young's modulus and stress-strain curves of commercial alloys as a function of alloy chemistry, processing details, strain rate and temperature using the JMatPro computer software (7,8,9,10,11,12). In the present paper, models used in resembling high temperature strength will be briefly introduced and demonstrated using various alloys as examples. The subsequently calculated monotonic properties will then be linked with em-pirical monotonic-to-cyclic relations, with the aim of seeing if reliable cyclical properties relevant to fatigue can be calculated. Most of the existing empirical methods relate to room temperature behaviour. Whether such methods can be applied at elevated temperatures remains unclear. The aim of the present work is therefore to calculate fatigue properties at high temperatures and compare with detailed experimental results for various commercial alloys.

``````