Existing methods of joining automotive aluminium alloys are either expensive (Self Pierce Rivets) or di cult to implement (Resistance Spot Welding). Ultrasonic spot welding (USW) is a new alternative method using ~2% of the energy of resistance spot welding. USW is a solid state welding process that combines vibration and pressure at the interface of a joint to produce a weld. Much of the existing research focuses testing under laboratory conditions, using simple coupon sample geometry, and has proven to be an extremely robust process. This thesis shows a detailed investigation into the implementation of USW on automotive body panels, in collaboration with Jaguar Land Rover. Weld performance, bonding mechanisms and temperature gradients found in AA5754 align well with other research conducted using 6XXX series aluminium alloys. A laboratory trial was completed to verify all joints could be achieved on a Jaguar XJ dash panel, followed by installation of a USW machine in a production cell. A detailed statistical analysis was performed on strength and sticking data gathered from 60 Jaguar XJ dash panels that were welded in the trial. Results showed difficulty to apply USW in certain areas of the panel, although previous trials had suggested it was possible. A collaboration with Ford Motor Company allowed research to be conducted at the Ford Research and Innovation Center. Experiments were designed to discover which elements of the USW equipment had the most profound effect on weld strength, and a full factorial Design of Experiments was produced to and the most effective method of reducing variation in weld strength. Results showed that the vibrational response of complex geometry parts makes USW very difficult to predict, making it difficult to successfully implement in the automotive industry.

This work has focused on formulating a experimental/numerical framework for the investigation of spot weld properties and performance. An Inverse temperature measurement approach has been established to predict the thermal history of a spot welded joints using remote thermocouples. This method incorporated the experimental data into an Artificial Neural Network (AAN) to predict cooling curves of the HAZ. Advanced modelling programs have been developed to simulate spot welded joints and thermocouples. Using the programs to investigate the effects of the key dimensional or material parameters on the mechanical or thermal response of spot welded joints of steels and different thermocouple joints relevant to their applications. Graphical User Interface Abaqus plug-ins of spot welded joints have developed using Python scripting and are used to investigate the effect of nugget size and sheet thickness on the stress and deformation of spot welded joints of steel. These works are important to establish an integrated approach to study the electrical, mechanical and thermal process of the spot welding process.

Complex heat treatments and manufacturing processes such as welding involve a wide range of temperatures and temperature rates, affecting the microstructure of the material and its properties. In this work, a diffusion based approach to model growth and shrinkage of precipitates in the alpha + beta field of Ti-6Al-4V alloys is established. Experimental heat treatments were used to validate the numerical predictions of the model for lamellar shrinkage, whilst data from literature have been used to evaluate the numerical model for the growth of equiaxed microstructures. The agreement between measurements and numerical predictions was found to be very good. Experimentally-based approaches were used both to describe the growth of alpha lamellae and martensitic needles while cooling down from temperatures above the beta transus, and beta grain growth for temperatures remaining above the beta transus. Such models were coded in the commercial FE software Visual-Weld for the prediction of microstructure evolution during welding simulations. Experimental welding tests were carried out to validate the predictions. The metallurgical models developed were linked with a mechanical physically based model to predict the flow properties and the initial implementation of the coupled models in Visual-Weld is discussed.

A comprehensive study of microstructure, texture and mechanical properties of Ti-6Al-4V joints produced by inertia friction welding is addressed in this thesis. Axis-symmetric tubular welds of different wall thicknesses were produced using different welding parameters. In the present study, there are two thin wall welds and three thick wall welds. A martensitic transformation occurred in the thin wall welds, resulting from the fast cooling rate, whereas it is prevented in the thick wall welds. The acicular microstructure in the Centre Weld Zone is finer and plastic deformation in the Thermal Mechanical Affected Zone is more severe with a higher energy input. In thick wall welds, prior β grain boundaries and grain boundary α are more clearly observed. Based on the reconstruction of prior β grains, the average β grain size can be related to the exponential of the forged pressure over the average input power density. The mechanical behaviour of both thin wall and thick wall welds for the varied welding parameters is also investigated, including tensile tests, fracture toughness tests, low cycle fatigue and fatigue crack growth resistance tests.

Aerospace applications typically require electronic products with not only higher levels of reliability than the consumer electronics industry but also longer service lives within demanding working environments. The transition will inevitably mean changes to design and manufacturing procedures, which is likely to incur a significant cost to the business. For example, the best candidate Pb-free solder alloys have been shown to require higher soldering temperatures and have higher surface tensions. Moreover, a reduction in product safety and reliability is not acceptable to the industry. This present work is divided into three sections. Firstly, the effect of increased component soldering temperatures on the integrity of the epoxy laminate material used for manufacture of printed circuit boards (PCB) has been assessed. Secondly, the required changes in soldering process parameters have been investigated for a range of solders and PCB finishes, largely due to the different wetting characteristics brought about by the increased surface tension of the Pb-free solders. Thirdly, the reliability of SnAgCu solder is assessed in comparison to the currently utilised SnPbAg solder alloy. This has been achieved firstly by accelerated thermal cycling, as the dominant mode of failure in a solder joint is typically thermo-mechanical fatigue and as such is already well researched. In addition, the mechanical fatigue properties have been assessed using a novel accelerated vibration test method and then finally, the two individual accelerated environmental tests of thermal cycling and vibration have been combined in a novel way to assess whether the combination is especially dangerous for SnAgCu solder reliability. A secondary objective of the combined environment test was to see if the well established thermal cycling test method for demonstration of product reliability can be further accelerated while still producing solder joint failure representative of those in-service. The present work shows that SnAgCu solder has inferior thermo-mechanical and mechanical fatigue life to SnPbAg solder. A combined environment test has been developed which effectively combines the single environments of thermal and vibration. The combination of thermal cycling with superimposed vibration is especially dangerous for SnAgCu solder, where an 89% reduction in the characteristic life is observed when compared to the equivalent thermal cycling characteristic life. It is suspected that a large reduction in life will be observed in SnPbAg solder, but not as pronounced as SnAgCu due to SnPbAg solders ability to better withstand plastic deformation that is induced by thermal cycling.

Creep failure of the weld structure in P91 steel components in high temperature power plant applications is often a key factor limiting the lifetime of the components. Whilst creep failure in weld heat-affected zone (HAZ) regions has been studied widely, the creep properties of the weld metal itself have been less well documented. In this work, the creep response of P91 weld metal in isolation was investigated in terms of microstructural evolution and mechanical properties. The microstructural examination of P91 multi-pass weld metal revealed a typical weld metal structure including columnar regions and refined regions. The columnar region exhibited high hardness whilst the refined region exhibited lower hardness. The anisotropic creep behaviour of P91 weld metal was observed in creep tests of both longitudinal and transverse specimens at 650ºC and various stress levels. This behaviour can be correlated with the microstructural anisotropy observed, where longitudinal specimens with banded columnar regions and refined regions parallel to the stress axis had longer creep life than transverse specimens with overlapped typical-shape beads. Longitudinal weld specimens showed higher strain to failure than transverse specimens. The microstructural investigation of creep tested P91 weld metal revealed two primary modes of creep fractures. In addition to creep fractures along columnar grain boundaries (typical of weld metal creep failure), creep fractures were also found along creep-weak white-bands which had formed at the inter-bead boundaries. The white-band regions consisted of material where the M23C6 carbides had dissolved during creep testing; the loss of carbides had allowed recrystallisation of the martensitic structure to ferrite and consequently this material was much softer than the bulk weld metal. The element mapping over the weld metal by laser-induced breakdown spectroscopy (LIBS) demonstrated that there was significant inhomogeneity in the distribution of certain elements, most significantly, chromium, manganese and molybdenum. This inhomogeneity resulted in strong activity gradients in carbon (even though the carbon concentration was homogeneous following welding) resulting in carbon loss from the alloy-depleted regions, the associated dissolution of carbides and the recrystallisation that accompanied this, and thus the poor mechanical properties which resulted in creep failure. The inhomogeneity in the distribution of certain alloying elements can be partially attributed to the solute partition of alloying elements during weld solidification which has been confirmed with examination of simulation P91 TIG welds. However, the homogeneity of weld metal in this case required mixing of a base steel (the core rod in the weld consumable) and particles of various ferro-alloys (delivered into the weld pool from the flux). It is argued that poor mixing in the stagnant layer (unmixed zone) at the solid-liquid interface during weld solidification also makes a significant contribution to the formation of alloy-depleted regions. The formation of white-bands has been modelled using Thermo-Calc based on the understanding of the formation mechanism involving solute partition and subsequent carbon diffusion out of the alloy-depleted region. A good correlation to experimental results has been shown in the prediction of limiting carbon concentration and M23C6 carbide content in white-bands. In addition, it was also suggested that depletion of carbides and carbon are strongly linked and that depletion of alloying elements only above a critical value will result in total carbide loss and thus recrystallisation into a white-band.

Ultrasonic metal welding (USMW) has received significant attention in the past few years, and has become more reliable and suitable for a wide range of applications. In recent years, the technique has been extensively used due to the advent of component miniaturisation and improvements in producing lightweight components. There are a number of advantages for USMW, including greater efficiency and speed, longer tool life, higher accuracy and no filler or flux needed to be used. Thus the technique can be viewed as being environmently friendly. However, the technique is not inexpensive, primarily due to the high cost of welding tools. Therefore, the design and construction of a lateral-drive USMW system which is capable of joining thin metals is presented in this thesis. The fundamental aspect of this study is the design of an integrated spot welding horn, along with other welding components such as a stationary anvil, mounting holder, welding bed, as well as the relevant fixing tools and fixtures. High precision is required in the design of the components, and in particular the welding horn. Because the horn is responsible for transferring energy to the welding zone, specimens must be prevented from sliding during the joining process, and an appropriate clamping force must be applied which will ensure acceptable bonding. Many criteria have been examined to enhance the performance of a working horn. The horn excitation frequency has been matched to the transducer frequency, ensuring that the horn will be vibrated longitudinally close to 20 kHz, thereby allowing the tuned mode to be isolated from other non-tuned modes, which guarantees uniformity of the vibration amplitude at the horn working surface, high gain factor of 4.108, and the avoidance of any stress initiated at the points between connecting components. Examining of these criteria is essential in order to optimise the excitation of the horn and to transmit the energy with minimum dissipation. The analytical studies and the finite element (FE) modelling of the welding components were successfully simulated, from which the vibrational behaviour and dynamical characteristics of the system were precisely verified using experimental modal analysis (EMA). The welding stack (the horn connected to the transducer), welding components and fixtures were then set-up on the driving machine. The device was examined prior to welding to ensure the excitation at high vibration. Many tests were successfully conducted on the welding together of aluminium and copper in a number of different configurations using the ultrasonic metal spot welding system. Weld strength and quality were shown to depend on complex relations of process parameters such as clamping force, amplitude of vibration, welding time and input power. A series of weld combinations with different thicknesses and ii variations in metal conditions were studied. The results of the lap tested specimens suggest that the bond strength is sensitive to the relationships between clamping force and vibration amplitude. Overall, the weld strength results suggest that the Al-Al welds are stronger and more consistent in terms of weldability than the Cu-Cu welds. In the welding of dissimilar metals, stronger welds are produced when the aluminium specimen is placed on top and in contact with the horn tip, rather than the copper. The thickness and surface condition of the metals such as hardness, surface roughness and oxides, are significantly affect the weld strength. In welding of Al-Cu or Cu-Al, an increase in energy and time was necessary to generate an acceptable bond. The use of stepped amplitude profiling results in a pronounced increase in the weld strength improves consistency and enhances weldability. However, horn tip/specimen adhesion and specimen marking did not occur under certain conditions. The results of the FE simulation and experimental tensile tests, for the load displacement curves profiles, allow for good estimation of the maximum load and therefore weld strength. Weld quality of aluminium and copper specimens were observed through investigation of the deformed surfaces using Nomarsky optical microscopy and scanning electron microscopy (SEM). The results illustrate that good quality welds can be achived by joining specimens, regardless of the surface condition of the metal. The SEM confirmed that no mixing occurred by melting or fusion between intimate surfaces, which indicates that USMW occurs due to adhesion and cohesion mechanisms. Furthermore, xray diffraction confirms the percentage of morphology between Al and Cu, which indicates that largest weld formations are prevalent for those specimens that are softer and lower in hardness and surface roughness, regardless of the type of tempering.

Dissimilar welding has been subject of several investigations due to its potential importance in various industrial fields such as transportation, energy generation and management. Dissimilar welding can increase the design efficiency, by the use of complementary alloys with different properties, cost cutting and light weighting structures. The use of different materials within a component or structure to best suit a particular task, requirement or increase its life and performance has always been an ambition of several designers and engineers. This project investigated the joining steel and aluminium for the automotive industry and also stainless steel and titanium to be applied in the civil nuclear energy generation industry. These dissimilar metallic combinations are metallurgically incompatible and the formation of brittle intermetallic phases (IMC) need to be controlled or eliminated. To join steel to Al, laser spot welding process was selected, to avoid the bulk melting of steel and Al at the joint interface that enhance the formation of brittle IMC. This part of the work was focused in controlling the joining process to control the IMC formation of galvanized and uncoated steel to Al and verify if it was possible to have a sound and reliable joint in the presence of an IMC layer. In the second part of this study, stainless steel to titanium joining, a different approach was taken with the application of weld metal engineering to modify or eliminate the IMC formation. Several metals were evaluated as potential interlayers to use and laser welding with a Ni interlayer was evaluated with moderate success, due to the modified IMC with improved mechanical properties and the good compatibility between Ni and the stainless steel. A further improvement was achieved when Cu was brazed between stainless steel and Ti using CMT (Cold Metal Transfer) a low heat input MIG process. The final attempt was to use a different interlayer that was 3D printed and deposited in several layers. This interlayer was composed Cu and Nb that were selected as candidates to avoid the IMC formation between the stainless steel and Ti. With this approach it was possible to build an IMC free component and possibly improve and avoid IMC formation in several other dissimilar metallic combinations.

Laser welding of metals and alloys is extensively used in industry due to its advantages of controlled heating, narrow weld bead, low heat affected zone (HAZ) and its ability to weld a wide range of metals and dissimilar metals. Laser welding of dissimilar metals such as carbon steels and stainless steel is still a challenging task, particularly due to the formation of brittle phases in the weld, martensitic formation in the HAZ and solidification cracking in the fusion zone. These issues can significantly deteriorate the strength of the welded joint. The aim of this work is to investigate the fundamental phenomena that occur inside the dissimilar weld zone and their effect on weld quality. In order to establish the key process variables, an initial study concentrated on the effect of different laser process parameters on dissimilar weld quality. In the second part of the work, a comprehensive study was performed to understand and subsequently control the alloying composition in laser dissimilar welding of austenitic stainless steel and low carbon steel. A dissimilar weld that is predominantly austenitic and homogeneous was obtained by controlling the melt pool dynamics through specific point energy and beam alignment. The significance of dilution and alloying elements on joint strength was established. A coupled CFD and FEM numerical model was developed to assist in understanding the melt pool dynamics and transportation processes of alloying elements. The model has been validated by a series of laser welding experiments using various levels of specific point energy. The laser welding characteristics in terms of geometric dimensions, surface morphology, alloying concentration, and dilution, were compared, and it is concluded that the specific point energy and laser beam position are the key parameters that can be controlled to obtain a weld bead with characteristics most suitable for industrial applications. In the third part of the work, a comparative study was performed to understand the significance of cooling rate, and alloying composition on the microstructure and phase structure of the dissimilar weld zone. Results show that the HAZ within the high carbon steel has significantly higher hardness than the weld area, which severely undermines the weld quality. A new heat treatment strategy was proposed based on the results of the numerical simulation, and it is shown to control the brittle phase formation in HAZ of high carbon steel. A series of experiments was performed to verify the developed thermo-metallurgical FEA model and a good qualitative agreement of the predicted martensitic phase distribution is shown to exist. Although this work is presented in the context of dissimilar laser welding of mild steel to stainless steel, the concept is applicable to any dissimilar fusion welding process.

Thick-section austenitic stainless steels have widespread industrial applications, especially in nuclear power plants. The joining methods used in the nuclear industry are primarily based on arc welding processes. However, it has recently been shown that the Narrow Gap Laser Welding (NGLW) technique can be used to join materials with thicknesses that are well beyond the capabilities of single pass autogenous laser welding. The heat input for NGLW is much lower than that of arc welding, as are the expected levels of residual stress and distortion. The multi-pass laser welding technique, based on the narrow gap approach, is an emerging welding technology which can be applied to thick-section welds using a relatively low-power laser, but the process is more complicated than autogenous laser welding, since it is necessary to introduce filler wire to narrow gap weld configurations. Despite this complexity, the technique is very promising for improving the penetration capabilities of the laser welding process. However a limited amount of research has been conducted on the development of the NGLW technique; the control and optimization of weld bead quality inside the narrow gap is still an area of weakness. The research described in this thesis involves investigations on NGLW of AISI grade 316L austenitic stainless steel, and the performance of the resulting welds. Design-of-experiments and statistical modelling techniques were employed to understand and optimize the welding process. A statistical model was used in order to understand the significant process parameters and their interactions, allowing improved control of the weld quality in ultra-narrow gap (1.5 mm gap width) welds. The results show a significant improvement in weld quality can be achieved through the use of statistical modelling and multi-variable optimisation. The microstructure characteristics and mechanical properties (e.g. tensile strengths, fatigue, bending strength and fracture toughness) of the NGLW samples were examined and compared with those of other welding techniques - autogenous laser welding and gas-tungsten arc welding (GTAW). The work shows that NGLW of 316L steel sheets up to 20 mm thickness have generally better or comparable mechanical properties than those of GTAW but with much higher welding productivity. The results of detailed investigations of the 2D residual stress distributions, material distortions, and plastic strain characteristics of the NGLW technique are described. The contour method was employed for residual stress evaluation of the NGLW technique, and the results were validated using X-Ray and neutron diffraction measurements. The results were compared with those obtained with GTAW. The results suggest that the longitudinal tensile residual stresses in NGLW joints are 30-40% lower than those for GTAW joints. The influence of the laser power and number of passes for the NGLW technique, on the developed residual stress and plastic strain has been investigated, and the influence of welding strategy and the use of restraint during welding were also investigated. To understand the thermal history in NGLW and its effect on residual stress, finite element analysis was carried out using ABAQUS to numerically model the behaviour of residual stress across the multipass NGLW weld joints. The model has been validated with the experiments using temperature measurements and in terms of residual stresses the model is compared with neutron diffraction and the contour method. There is a very good correlation between the model and experimental results. The influence of both the laser power and welding speed on the induced residual stress during the NGLW process was also investigated using the model. The aqueous, pitting and stress corrosion cracking behaviour of the NGLW joints were investigated, and the results compared to those for GTAW joints under the same conditions. The results show that NGLW joints have better resistance to pitting corrosion than the GTA welds. Preliminary results also suggest that NGLW has better resistance to stress corrosion cracking.