Search

Fusion welding of steel pressure vessels provides one of the most economic methods for creating such structures. However, the highly localised heat distribution, at the fused region gives rise to non-uniform heating / expansion and cooling / contraction of the weld and the surrounding base material, consequently giving rise to residual stresses and distortion. Several techniques are available to minimize welding distortions and residual stresses [1]. One feasible and efficient mitigation technique, that alters the inherent residual stresses developed during welding, includes the use of low transformation start temperature welding wire [2e4]. By altering the filler material and metallurgical transformations developed at the fusion zone and heat affected zone, wielded regions at the weld line can give rise to compressive residual stresses [2]. Consequently by establishing the appropriate chemical composition of the welding wire, a significant reduction in welding distortions can be achieved. This study focuses on establishing the merits of using different filler materials to minimize welding residual stresses and distortion through a thermal-metallurgical-elasto-plastic numerical parametric study. Numerical modelling of butt welded structures using three different filler materials is considered and the applicability of using low transformation temperature welding wire is investigated. The predicted residual stresses and distortions give a significant advance in understanding and control of welding distortion in welded fabrications. The chemical composition of the filler material and ultimately fused zone composition, leading to reductions in residual stresses and distortions is identified.

Welding is one of the most important metal joining techniques due to its advantages in relation to speed and versatility. One of the unwanted effects of gas metal arc welding is residual stress, that arises due to the inherent temperature loadings and can lead to structural integrity or assembly problems. Experimental investigation of these effects is highly desirable but often costly, especially in large-scale construction such as shipbuilding. It is therefore of interest to exploit the potential of computational analysis techniques, such as finite element analysis, for the assessment of these effects. The use of finite element analysis in this context is however not straightforward, due to the complex nature of the problem. The work presented here deals with an experimentally validated modelling strategy applied to two weld joint configurations: butt and fillet welded plates. Simulation techniques are presented which are aimed at predicting residual stresses. These are validated against measurements carried out using the hole drilling method for residual stress determination. Results show a good qualitative and quantitative match in simulated and experimentally measured values, which back up the modelling techniques adopted.

The aim of this study was to investigate the influence of different tack welding fabrication procedures on the final deformations of seam welded plate structures. In this study the length of the tacks, number of tacks, and position of tacks with respect to the final seam weld are altered, and their sensitivity with respect to out-of-plane deformation is established. A parametric study of these three different procedural variations is performed via finite element models to identify which fabrication procedures would lead to minimal out-of-plane distortion.

Fusion welding is the most common and convenient method used for the fabrication of large, thin-plate welded structures. However, the resulting tendency to out-of-plane distortion exacts severe design and fabrication penalties in terms of poorer buckling performance, lack of fairness in external appearance, poor fit-up and frequent requirements for expensive rework. This study forms part of a long-term project that has the aim of modelling welding and related fabrication processes computationally with particular emphasis on the out-of-plane distortion outcomes. Throughout the present work the computational models have been cross-referenced to realistic experimental test cases. A repeated finding of the trials was that minor variations in fabrication procedures, were found to have significant effects on distortion. In particular, the pre-fabrication procedures, including spot and tack welding, have a significant effect on the initial out-of-plane distortion leading to differences in distortion of the post-welded structures. The support and clamping conditions during welding and cooling of welded thin-plate structures were also found to have a considerable influence on the final predicted out-of-plane distortion. The outcomes often result in different buckling instability behaviour. This paper concentrates on these aspects and draws on studies of buttwelding between plates of thicknesses 3-8 mm.

Multiply-stiffened, thin plate, welded fabrications are used in a wide variety of transport fields, however the resulting out-of-plane distortion associated with welding exacts a severe design penalty. Depending on the information required, the size of the structure under investigation and the computer power at hand, three computational strategies may be considered to predict welding distortion. If prediction of the localized residual stresses from welding is of major importance, then a full transient, uncoupled thermo-elastoplastic analysis is preferred. This method is not readily applicable to predict welding distortions in industrial-scale welded structures. More computationally efficient models are required and two other models are suggested in the current study. A series of experimental tests of a realistic nature were performed to validate the proposed computational strategies. Computational and experimental study of butt and fillet welding of small and industrial size fabrications is considered.

This study aims to develop and validate a wide-ranging simulation tool to predict welding distortion in stiffened plates and shells, with particular emphasis on out-of-plane deformation. The approach adopted in this study uncouples the thermal, elasto-plastic and structural effects leading to distortion. The computational models and results are supported by realistic welding tests and appropriate measurements to validate the simulated thermal fields and out-of-plane distortions. The simplest and most computationally efficient model makes use of algorithms, instead of numerical computation, to link the thermal welding strains to the elasto-plastic and structural responses of the welded assembly, via a static, single-load-step analysis. Alternative, more computationally intensive models are explored which simulate the full transient thermal and elasto-plastic structural responses in an uncoupled fashion. These provide a cross-reference for the more rudimentary but computationally efficient models. The experiments and computational strategies are applied to welded assemblies incorporating double-fillet-welded stiffeners.

This study presents a number of finite-element (FE) models aimed at illustrating the effect of using different modelling strategies for the simulation of the thermo-elasto-plastic stages of the welding process. The main challenge with computational approaches is to model the complex and at times indeterminate nature of the welding process, preferably in a simple and transparent manner. The study uses the example of butt welding thin rectangular plates, the results of main interest being the out-of plane distortion and longitudinal residual stresses. The FE model results are compared to experimental data and simplified analytical algorithms that have been developed and validated in the context of a project to improve the prediction and control of welding distortion in thin welded structures.

Prediction and control of thermal distortion is particularly important for the design and manufacture of multiply stiffened welded structures. This study aimed to develop and experimentally validate a comprehensive simulation tool to predict distortion, with particular emphasis on out-of-plane deformation generated in double-sided fillet-welded attachments. Simulation was used to optimise the relative positions of a twin-arc configuration, to give minimum out-of-plane deformation consistent with reasonable production time for single stiffener, double-fillet attachments. The critical buckling load of the structure was approached and exceeded as the arcs were brought closer and simulation allowed the influence of this factor to be determined.

The study's aim is to improve the applicability of finite element analysis to the prediction of welding distortions, with particular emphasis on out-of-plane deformations. Robust simulation strategies are established which take account of material properties and joint configurations, but are at the same time computationally efficient. This is achieved by reducing the full transient thermo-elastoplastic analysis to variants of an uncoupled thermal, elasto-plastic and structural treatment. A two-dimensional cross-section thermal model was used to establish thermal transients. The maximum temperatures, experienced at each node during the welding cycle, were then used to link the thermal welding strains to the elasto-plastic and structural response of the welded structures. Three efficient models have been identified that reduce the transient analysis to a simple multi-load-step analysis and these were applied to sample butt-welded plates. The simulation techniques are supported by full-scale welding tests on steel plates. The evolution of angular distortion can be treated simply through non-linear, stepwise, static analyses. This captures important effects of restraint and material properties. The longitudinal bending distortion can then be established via simple elastic-perfectly plastic algorithms, through a fictitious elastic thermal load analysis.

Major simplifications of the corresponding rigorous thermal and thermo-mechanical analyses are applied in this chapter. Transient temperature fields are replaced by fixed temperature profiles. The ‘computationally efficient’ (CE) methods introduce further simplifications where transverse and longitudinal thermal strains are derived from approximations embodied in the mismatched thermal strain and transverse contraction strain algorithms. In the CE ‘analytical’ approach, these strains are applied to linear elastic, small strain models, using conventional strength of materials techniques. In the CE ‘computational’ approach, the strains are applied to linear-elastic, large-strain, finite-element models. Various hybrid stepwise methods are also shown, where a measure of sequential description is included, together with the use of elasto-plastic, finite-element computation.

Results from weld simulations and experiments for fillet-weld procedures are brought together to provide information on distortion in stiffened plate fabrications. The results of thermocouple and thermographic studies are used to expand on thermal aspects, described initially for small-scale fabrications in Chapter 7 . Computationally efficient solution methods are applied to large-scale fabrication cases and the results compared with experiments. Special attention is given to application of the transverse contraction strain (TCS) algorithm and ways in which it can be fine-tuned to match different models of transverse constraint. Modelling methods for multiple parallel stiffeners are given and the results of full-scale experiments are compared with the results of a computational case study in which different sequences of stiffener attachment were applied.

Modelling strategies and results from previous chapters are brought together in a comprehensive treatment of butt welding procedures. Various practical issues are treated, including effects on distortion of initial out-of-flatness, disposition of supports, tacking assembly and clamping of components. In the case of clamping, it is shown that excessively constrained clamping strategies can lead to greater levels of distortion than if the components are relatively unconstrained. The outcomes of the various simulation methods for prediction of residual stresses in butt welds are described. Effects in multiple longitudinal butt welds where there are interactions between several welds in an assembly are shown.

Distortion and residual stress are driven by the thermal gradients generated by the moving heat source. Heat energy enters the workpiece through the fusion zone primarily and procedures are provided to model heat input as a lumped source, including the effects of weld filler deposition and losses due to convection and radiation. Alternative three-dimensional and two-dimensional thermal models are given. Stiffener fillet welding involves more conduction routes and a double-sided fillet stiffener weld is treated. Methods to determine welding thermal efficiency are given, with values for different processes. The characteristics of thermal cutting are discussed.

Information from all preceding chapters in the book is applied to support design and manufacturing strategies, showing how simulation can be applied to explore alternative design configurations and manufacturing sequences. In terms of design, configuration and manufacturing sequence questions in the fabrication of multi-stiffened plates are examined. Studies also include the use of multiple welding heads, where computation is applied to show that an optimum spacing can be established, avoiding the risk of buckling. Simulation is also applied to investigate three techniques that have been proposed to mitigate distortion.

Contrasting approaches to control and reduction of distortion are described. A method based on the use of artificial neural networks (ANNs) uses statistical techniques to interpret observations of distortions and to discover interactions between different input parameters. The section on computational simulation is based on mathematical modelling of the physical processes in welding. Several processes that can be applied along with the welding operation or in the post-welding phase are described. Thermal and mechanical methods include transient thermal tensioning, roller tensioning, intense cooling in the trailing region behind the weld and the use of special weld filler metal.

The opening chapter contrasts a welding design and fabrication strategy based on trial and error with one that is guided by quantitative welding simulation methods. An outline is given of the main contributing factors to the development of computational welding mechanics as a discipline and the possibilities that such studies offer for interpreting welding phenomena such as distortion during fabrication. The importance of establishing simple models is emphasised and a first mention is made of simplified computational methods. The content of the following chapters is briefly described and the relationships between each of the chapters in the sequence are explained.

A first introduction to welding mechanics is given, based at this stage on uncoupled analytical solutions for thermal and thermomechanical behaviour. Rosenthal’s well-known quasi-static solution for the temperature field caused by a fast-moving, concentrated heat source is used to demonstrate the influence of different material properties and welding process conditions on the temperature fields. An approximate thermo-mechanical solution due to Okerblom is used to demonstrate the roles of the thermal dilatation of the material being welded and the heat input on welding deformations. This solution is also used to illustrate the development of residual stress fields and the potential for buckling.

Experimental strategies and techniques underpinning this book are described and the role of practical studies is highlighted. Thermal investigation of welding heat input and conduction in the workpiece is based on thermocouple arrays and thermography. Welding and measurement rigs to create industrial-scale structures are described. The arrangement integrates welding with thermal and deformation measurements to minimise workpiece handling errors. A special deformation-smoothing procedure is given that ensures appropriate interpolation between point deformation measurements and leads to profiles that can be differentiated to provide accurate surface slopes and curvatures. The use of different statistical measures of out-of-plane deformation is discussed.

The overall industrial context of thin, stiffened-plate fabrication is outlined. A typical fabrication sequence is described, showing the various stages where distortion issues can arise. Stock quality and flatness have a strong influence on outcomes. Some high-risk factors for distortion are identified, which could be removed by better management of the design and fabrication process. Cause and effect are not always easily related in many cases and application of quantitative modelling tools could often resolve ambiguity and improve distortion outcomes. The chapter closes with a brief description of methods used to rectify any distortion that may have arisen.

The many different physical phenomena present in welding operations are described. It is shown that the many interactions between these can be simplified to a smaller number, similarly for material properties. An outline of thermal computation approaches is given. Thermo-mechanical computational approaches are described, ranging from ‘reduced solution’ methods, where substantial simplifications are made, to full elasto-plastic methods that make greater demands on computing resources. The advantages of reduced solutions are exemplified through a butt-welding simulation treated via a ‘computationally efficient’ method based on the Okerblom approach and two variants of a hybrid method.

The main aim of the work was to investigate a simplified finite element simulation of the out-of-plane distortion caused by fusion butt welding. The thermal transient part of the simulation made use of a finite element analysis of the two-dimensional cross-section of the weld joint and the thermoelastic–plastic treatment was based on analytical algorithms describing transverse and longitudinal deformations, leading to predictions of transverse angular deformation and longitudinal contraction force. These results were then applied to a non-linear elastic finite element model to provide predictions of the final angular and overall deformations of the butt-welded plates. The validity of the simulation was investigated via full-scale tests on 4 m 1.4 m 5 mm steel plates, butt welded using a flux-cored Ar–CO2 metal–inert gas process. Thermography and thermocouple arrays were used to validate the thermal transient computations and out-of-plane deformations were measured using displacement transducers for transient deformations and a laser scanning system to measure the profiles of the whole plates before and after welding. The results of six full-scale tests are given and comparison with the simulations shows that the procedure provides good prediction of the angular and overall out-of-plane deformations. Prediction accuracy requires account to be taken of initial shape, gravity loading, and support conditions., peer-reviewed

A shaft failure in a 634 kW diesel generator after 4000 hours of operation required investigation, especially when similar sets began to show signs of excessive bearing housing wear. The failure was found to be due to torsional fatigue caused by operation at an unforeseen resonant condition. The damage to the bearing housing of this type of set was also attributed to this behaviour. Replacement of the bearings and housings and redesign of the flexible coupling has been undertaken to prevent further failures. Modifications to design procedures are now being adopted to ensure similar problems do not occur in other existing or new machines.

Meaningful testing of stab resistant body armour requires the use of realistic body tissue simulants. A device for the determination of the force–displacement behaviour of materials in high impact loading situations has been developed for the testing of such simulants. Force measurement is achieved with the use of electrical foil strain gauges applied to a cylindrical load cell. A piezo-resistive accelerometer (±500 g) is used to calculate the displacement of the device through double integration of its signal, with the impact velocity used as a boundary condition. The signals from the strain gauge circuit and the accelerometer are sampled at 2500 Hz. The data are transmitted to a receiver via telemetry using a 418 MHz FM transmitter and from the receiver to a laptop PC via the serial port. Calibration of the device is described and sample results showing forces up to 2500 N and displacements up to 0.04 m are presented.

Many industries, such as shipbuilding, concerned with the fabrication of fusionwelded plate structures, face increasing challenges to produce lightweight structures. This design requirement is commonly met by using thin-plate, multiply stiffened, welded structures, but severe difficulties and high rectification costs are frequently incurred, related to the evolution of out-of-plane deformations. The overall scope of this study is to improve the applicability of computational prediction of distortion by providing simple and adaptable methodologies, which can be readily validated through experience of application in the industrial context. These methods are designed to be computationally economic and robust, and they are also generic with respect to material properties, welding processes, and thickness. The aim is to provide design engineers with the tools to explore alternative structural and process parameters and hence to find out if the outcomes will be acceptable, prior to embarking on manufacturing operations typical of large-scale welded structures. The validity of the simulations was investigated via full-scale tests where several filletwelded 100 mm × 6 mmstiffeners were attached to 4 m × 1.5 m × 5 mmthick plates, according to different sequences. The computational models were used to optimize the welding scheme with respect to minimum out-of-plane deformation and welding sequence.