Fatigue Assessment of Welded Joints by Local Approaches

Dieter Radaj, C M Sonsino, W Fricke

• Publisher: Woodhead Publishing
• Year: 2006
• Print ISBN: 9781855739482
• eBook ISBN: 9781845691882
• Edition: 2
• Topic: Technology & Engineering
• Subtopic: Mechanical Engineering
• Index: Technology & Engineering

Local approaches to fatigue assessment are used to predict the structural durability of welded joints, to optimise their design and to evaluate unforeseen joint failures. This standard work provides a systematic survey of the principles and practical applications of the various methods. It covers the hot spot structural stress approach to fatigue in general, the notch stress and notch strain approach to crack initiation and the fracture mechanics approach to crack propagation. Seam-welded and spot-welded joints in structural steels and aluminium alloys are also considered.This completely reworked second edition takes into account the tremendous progress in understanding and applying local approaches which has been achieved in the last decade. It is a standard reference for designers, structural analysts and testing engineers who are responsible for the fatigue-resistant in-service behaviour of welded structures.- Completely reworked second edition of a standard work providing a systematic survey of the principles and practical applications of the various methods- Covers the hot spot structural stress approach to fatigue in general, the notch stress and notch strain approach to crack initiation and the fracture mechanics approach to crack propagation.- Written by a distinguished team of authors

1

Introduction

1.1 Fatigue strength assessment of welded joints

1.1.1 Present state of the art

Fatigue failure of structural members, comprising crack initiation, crack propagation and final fracture is an extremely localised process in respect of its origin. Therefore, the local parameters of geometry, loading and material have a major influence on the fatigue strength and service life of structural members. They must be taken into account as close to reality as possible when performing fatigue strength assessments and especially so when optimising the design in respect of fatigue resistance.

Design rules for fatigue-resistant structures, on the other hand, take local effects only roughly into account. They are based mainly on the nominal stress approach, which is a global concept in principle. The permissible nominal stresses depend on the ‘notch class’, ‘detail class’ or ‘fatigue class’ (FAT) of the welded joint being considered. They are supplemented by general design recommendations.

The code-related state of the art is unsatisfactory in those fields of engineering where structural members are subjected to fatigue-relevant variable load amplitudes with appreciable numbers of cycles or where nominal stresses cannot be meaningfully defined. Local concepts are applied in these areas based on local strain measurements, mainly by strain gauges, and by local stress calculations mainly based on the finite element method. Both the testing engineer and the structural analyst urgently need well-founded methods for evaluating these local stresses and strains in respect of fatigue strength and service life.

These needs can be met only insufficiently if at all. The multitude of proposals on how to assess the fatigue resistance of structural members based on local parameters is difficult to overview and evaluate.¹ Different fields of engineering, ‘schools’ of researchers and national communities prefer different approaches. All proposals are more or less incomplete in respect of user demands, and the local parameter data, for the most part, lack statistical proof. As a result, the application of local approaches lags behind the possibilities provided by computerised structural analysis.

1.1.2 Demands from industrial product development

The demands originating from industrial product development concerning local approaches are twofold:

– An overview covering methods and data available for application is needed.

– Standardisation of the procedures and their incorporation into design codes are required.

This book is intended mainly to satisfy the first demand. Industrial users should obtain all the available information so that they can decide on the best way to treat their individual fatigue problems on the basis of local approaches. They must then supplement the information available from the book and the quoted literature by their own empirical and experimental data.

The second demand can only partly be satisfied. There is no generally acknowledged theory of local fatigue strength available on which a uniform analytical scheme could be based. On the one hand there are manifold procedural variants and data sets and on the other hand there are innumerable fatigue problems in industry. Any general standardisation of the local approaches would interfere with the development of further methods, which must always be adapted to the application being considered. Only carefully selected parts of the procedure are suited to standardisation or at least to defining a guideline. Substantial progress with regard to the standardisation of analytical strength assessments based on local stresses (structural or notch stresses) has been achieved by the IIW recommendations³ and by the FKM guideline.¹

The subjects in this book are restricted to welded joints, which are of paramount economic relevance. Additionally, welded joints show peculiarities in respect of fatigue behaviour which make a separate treatment of the fatigue assessment methods desirable. Finally, part of the local approaches has been developed for welded joints independently of the methods developed for non-welded members. Restriction to welded joints is therefore well justified.

The following books give additional guidance on fatigue assessment of welded joints by local approaches: Haibach² (in-service fatigue strength, highly related to design, emphasis on analysis and statistics) and Radaj⁴ (also related to design, covers early stage of development of local approaches). The contribution by Seeger⁸ in a more general handbook lays emphasis on assessment methods with inclusion of variable-amplitude and multiaxial loading conditions. The fundamentals of the analysis of fatigue strength and its application to non-welded members are presented in books by Haibach,² Dowling⁹⁵¹ and Radaj.⁶ The analysis of welding residual stresses and distortion is found in Radaj’s book.⁷

1.2 Basic aspects of assessment procedures

1.2.1 Multitude of parameters governing fatigue failure

The local approaches to fatigue assessment reviewed in this book aim to cover the dominating parameters of extremely complex physical processes in order to make them controllable by the engineer. These processes comprise primarily microstructural phenomena (moving dislocations, micro-crack initiation on slip bands and further crack growth by local slip mechanisms at the crack tip) but can be approximately described by a macroscopic elastic or elastic-plastic stress and strain analysis according to continuum mechanics which refers to the cyclic deformation causing initiation and propagation of the ‘technical crack’ with inclusion of the final fracture, Fig. 1.1. A technical crack is considered to have been initiated (usually at the surface) if its surface length reaches values which can be detected by common technical means, e.g. 1 mm, and its depth 0.5 mm.

Fig. 1.1 Micro- and macrophenomena of material fatigue.

The initiation of the technical crack by cyclic loading under definite local material conditions is primarily governed by the amplitudes of the cyclic stress and strain components at the notch root, with the volume of the highly stressed material, the multiaxiality of the cyclic stress state and its static mean value (possibly fluctuating) also being of importance. The total number of parameters influencing the critical values of the cyclic stress and strain components which describe crack initiation are summarised in Table 1.1 which refers to the local approach insofar as local stresses and strains are introduced to characterise the loading type. The number of influencing parameters is large, but can be handled within the procedure of strength assessment. However, a problem arises from the restricted possibilities of decoupling the effects of these influencing parameters in the case of engineering tasks.

Table 1.1

Parameters governing fatigue crack initiation; after Radaj⁵

Structural member	Surface	Material
Shape	Roughness	Type
Size	Hardness	Alloy
Dimensions	Residual stress	Microstructure
Loading type	Loading course	Environment
Stress amplitude	Amplitude spectrum	Temperature
Mean stress including residual stress	Amplitude sequence	Corrosion
Multiaxiality including phase angle	Rest periods

Crack propagation by cyclic loading is primarily governed by the amplitudes of the cyclic stress intensity factor or of the cyclic J-integral at the crack tip. Most of the parameters which determine the critical value of stress, strain or energy at the crack tip causing crack propagation are identical to those which cause crack initiation. Only the influence of the surface diminishes whereas crack shape, crack size and crack path gain in importance.

The multitude of parameter constellations governing fatigue are advantageously structured according to Haibach,² based on the main testing and analysis procedures used to obtain the above-mentioned critical values for fatigue strength or service life assessments, Fig. 1.2.

Fig. 1.2 Field of parameter constellations governing fatigue failure, structured on the basis of the main testing and analysis procedures; after Haibach.²

The description of fatigue strength proceeds from the S–N curve (nominal stress amplitude versus number of cycles) of the unnotched specimen (a). The S–N curve of the notched specimen (b) is gained therefrom by considering the stress concentration factor and the notch radius. Finally, the S–N curve of the structural component (c) results from additionally considering size and surface effects (including residual stresses). This path a–b–c or e–f–g is connected with the problem of strength dependent on shape and size (German idiom ‘Gestaltfestigkeit’). On the other hand, the fatigue life curve resulting from variable-amplitude loading can be derived from the S–N curve resulting from constant-amplitude loading by introducing a damage accumulation hypothesis. This is the path a–e, b–f or c–g from conventional fatigue strength to service fatigue strength. The problem of damage accumulation can be partly solved by determining the fatigue life curve of the notched specimen under standard load sequences, path d–f–g instead of c–g.

The structuring of the parameter field and procedures mentioned above does not mean that every fatigue strength assessment starts with the S–N curve of the unnotched specimen and ends with the life curve of the structural c...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright page
• Foreword
• Preface
• Author contact details
• 1: Introduction
• 2: Nominal stress approach for welded joints
• 3: Structural stress or strain approach for seam-welded joints
• 4: Notch stress approach for seam-welded joints
• 5: Notch strain approach for seam-welded joints
• 6: Crack propagation approach for seam-welded joints
• 7: Notch stress intensity approach for seam-welded joints
• 8: Local approaches applied to a seam-welded tubular joint
• 9: Structural stress or strain approach for spot-welded and similar lap joints
• 10: Stress intensity approach for spot-welded and similar lap joints
• 11: Notch- and crack-based approaches for spot-welded and similar lap joints
• 12: Significance, limitations and potential of local approaches
• Bibliography
• Index