Modelling thermal cycles and intermetallic growth during friction melt bonding of ULC steel to aluminium alloy 2024-T3

Abstract

Dissimilar materials, aluminium 2024-T3 and ultralow carbon steel, have been welded by a novel process called friction melt bonding. A finite element thermal model is developed to predict temperature cycles and to estimate the fusion pool geometry and the intermetallic bonding layer thickness. The total mechanical power input in pseudo-steady state is inferred from in situ measurements at the tool torque and rotational speed. Temperature dependent properties, including the latent heat of fusion, and proper contact conditions between the welded plates and the backing plate are included. Predicted temperatures are in agreement with the measurements at various distances from the weld centreline. Molten pool geometries and intermetallic thicknesses, whose control is crucial to insure good weld mechanical performances, are also in accordance with the experimental observations.     

Effect of welding conditions on reduction in angular distortion by welding with trailing heat sink

Abstract

The effectiveness of welding with a trailing heat sink in reducing the angular distortion of a weld has been experimentally investigated with focus on the cooling position. A numerical model of welding with a trailing heat sink is constructed through the comparison of measured values of weld penetration, thermal cycles and angular distortion with those calculated. On the basis of this model, the effect of welding heat input conditions on the reduction in angular distortion is discussed to evaluate the versatility of welding with a trailing heat sink. The results indicate that the choice of an appropriate cooling position behind the welding heat source is essential for the effective reduction in angular distortion. The reduction in angular distortion by the heat sink at the appropriate cooling position increases with the heat input parameter Q_net/h, where Q_net is the weld heat input and h is the thickness of the plate.     

Application of discrete distribution point heat source model to simulate multipass weld thermal cycles in medium thick plates

Abstract

An analytical solution to the heat flow differential equation has been proposed, which allows an adequate and simplified approach to multipass welding conditions in the heat affected zone. The solution is based on the medium thick plate temperature distribution initially proposed by Rosenthal. The model applies a discrete distribution of point heat sources localised on any point of the plate that is being welded. This approach allows changing the position of the heat sources in the groove from one pass to another, reproducing multipass welding conditions. Actual thermal cycles of multipass welding of AISI 304 and 2304 stainless steels were recorded and compared with simulated thermal cycles, which verified the agreement between the simulated and the actual HAZ thermal cycles. Microstructures of alloys UNS S32304 and UNS S32205 were reproduced in the Gleeble system using the calculated thermal cycles and their comparison with actual weld microstructures confirmed the utility of the proposed heat flow model for metallurgical weldability studies involving multipass welding.     

Microstructural evolution of simulated heat affected zone in ASTM A533 type B steel plates

Abstract

In order to investigate the weldability of ASTM A533 type B steel plates, simulated heat affected zone (HAZ) experiments with heat inputs of 20, 50, and 80 kJ cm^?1 were carried out. The thermal cycles used corresponded to the actual thermal cycles that occur in the coarse grained region of the real HAZ. The microstructural development and the toughness of the simulated HAZ were studied. It was found that with a heat input of 20 kJ cm^?1 the simulated HAZ microstructure gives larger amounts of lower bainite with significant amounts of auto tempered martensite. In the cases of heat inputs of 50 and 80 kJ cm^?1 both the simulated HAZ microstructures were composed chiefly of upper bainite. The specimen obtained from the simulation of a heat input of 20 kJ cm^?1 possesses higher toughness than those obtained from the simulations of heat inputs of 50 and 80 kJ cm^?1 Furthermore, the results also indicate that the tempering response is more sensitive in autotempered martensite than in bainite.     

Measurement and calculation of arc power and heat transfer efficiency in pulsed gas metal arc welding

Abstract

This study investigates the heat transfer efficiency of the pulsed gas metal arc welding (GMAW-P) process. The arc power and heat input were calculated from arc current and voltage measurements and the heat input was also measured with a liquid nitrogen calorimeter. The measured heat transfer efficiency for GMAW-P varied slightly over a wide range of pulse parameters, with an average value of 70%, a maximum of 72% and a minimum of 68%. Welding heat transfer efficiency based on arc power calculated as the product of average current and voltage was too high (averaging 82%), while that calculated using the product of the root mean square (RMS) of the average current and voltage was too low (averaging 61%). Both also varied significantly with pulsing parameters. Mathematical analysis shows that average instantaneous power values must be used when current and voltage vary significantly with time. The experimental differences between the average instantaneous power and the other calculated values are explained by the relative phases of the pulsed current, voltage and arc resistance waveforms.     

Simulation of GMAW thermal process based on string heat source model

Abstract

A 3-D string heat source model was developed to simulate the thermal process for gas metal arc welding (GMAW) with a new transient solution of heat transfer. Reflection heat sources, the imaginary heat source and the imaginary heat sink were introduced to make the solution more reliable. The moving welding heat input during GMAW welding process was presented with a moving 3-D string heat source model made up of a group of elementary point heat sources along the moving coordinate axes. Using established models, the whole process from arc starting, quasi-steady state (QSS), to arc extinguishing during welding Q195 steel has been simulated. Meanwhile, some calculations and experiments for JG590 steel have also been made. The predicted weld cross-sections and welding thermal cycles show a good agreement with experimental measurements.     

Application of thermal modelling to laser beam welding of aluminium alloys

Abstract

A thermal model has been developed for laser welding which describes the heat input in terms of point and line sources. The model was used to generate weld profiles which closely matched those found by experiment. Outputs of the model (the thermal gradient G_L and the growth rate R) were used to describe the macroscopic grain structure found along the weld centreline. Columnar structures were predicted at low welding speeds (high G_L/R ratio) and equiaxed structures at high welding speeds (low GL/R ratio). Using the thermal model, cooling rates of ～1500 K s^–1 were estimated for the lowest welding speed, which increased by an order of magnitude for the highest welding speed considered. There was excellent agreement between the dendrite secondary arm spacings measured by experiment and those predicted using the thermal model.     

Modeling cast IN-738 superalloy gas tungsten arc welds

A three-dimensional finite-element thermal model has been developed to generate weld profiles, and to analyze transient heat flow, thermal gradients and thermal cycles in cast IN-738 superalloy gas tungsten arc welds. Outputs of the model (cooling rates, the thermal gradient G and the growth rate R) were used to describe solidification structures found around the weld pool for three different welding speeds at constant heat input. Calculations around the weld pool indicate that the cooling rate increases from the fusion line to the centerline at all welding speeds. It was also observed that the cooling rate (G × R) and the ratio G/R fall with welding speed. For instance, as the welding speed is increased, the cooling rates at the centerline, fusion line and penetration depth decrease. Moreover, it was observed that as the power and welding speed both increase (but keeping the heat input constant), the weld pool becomes wider and more elongated, shifting from circular to elliptical shaped. The calculations were performed using ABAQUS^® FE code on the basis of a time-increment Lagrangian formulation. The heat source represented by a moving Gaussian power density distribution is applied over the top surface of the specimen during a period of time that depends on the welding speed. Temperature-dependent material properties and the effect of forced convection due to the flow of the shielding gas are included in the model. Numerically predicted sizes of the melt-pool zone and dendrite secondary arm spacing induced by the gas tungsten arc welding process are also given.     

Effect of welding parameters on residual stress in type 420 martensitic stainless steel

Abstract

The effect of welding parameters on residual stress induced by shrinkage of weldment and metallurgical phase transformation in type 420 martensitic stainless steel has been investigated. In this study, type 1018 low carbon steel was adopted as the base metal and type 420 martensitic stainless steel was used for the filler metal during submerged arc welding. The thermal cycles at various locations were recorded and dilatometry was used to examine the martensite phase transformation temperatures. The experimental results show that the residual stress increased with the heat input during welding. Using a higher welding heat input increased the amount of heat going into the weldment and elevated the martensite phase transformation temperature. Residual stresses could not be significantly reduced by increasing preheat (interpass) temperature while welding. Using higher preheat temperature conditions could elevate the equilibrium temperature and the martensite phase transformation temperature and increased the heat input to the weldment.     

Arc power and efficiency in gas tungsten arc welding of aluminium

Abstract

A calorimetric study of gas tungsten arc welding of aluminium is described. The present study comprised experiments in which autogenous welding runs were each made on a block of electrical conductor grade aluminium. The blocks were all approximately cubic in shape which, when combined with the high thermal conductivity of aluminium, ensured that their temperature equalised soon after the completion of a run. Each sample was immersed in insulating material before welding so that heat losses to the surroundings were minimised. Thermocouples were attached to the block in each experiment and the bulk temperature rise was related to the energy input associated with the welding run. The effects of arc polarity, alternating current balance, shielding gas composition, arc length and welding current on the arc power and arc efficiency were investigated. The results obtained with alternating current are compared to those for direct current, and the differences are explained.     

Angular distortion of fillet welded T joint using low transformation temperature welding wire

Abstract

A newly developed low transformation temperature welding wire, of which the transformation start temperature is lower than that of conventional welding wires, was applied to fabrication of fillet welded T joints. The welding angular distortion and the temperature profile of the weld metal were continuously measured during the welding process. The angular distortion of the fabricated T joint was reduced when the weld metal reached the martensitic transformation start temperature. The residual angular distortion was less with the low transformation temperature welding wire than that with the conventional welding wires. The welding distortion of T joints was calculated by a numerical simulation with consideration of the effect of phase transformation under weld thermal cycles. The welding distortion was reproduced with high accuracy in the numerical simulation. Results of the numerical simulation also determined that there was a direct correspondence between the transformation expansion of the weld metal and the angular distortion.     

Electroslag cladding of low alloy steel with stainless steel

Abstract

Cladding of a low alloy steel slab with stainless steel was carried out using a modified electroslag remelting technique. It is shown that the thickness of the cladding that can be achieved via electroslag remelting is dependent on the fill ratio used. The effect of power input on the joint profile obtained is reported. A combination of low fill ratio and relatively low power input is essential to minimise penetration of the base slab by the liquid metal. A satisfactory joint profile and defect free joint can be obtained via the optimisation of these process parameters. The clad product was successfully forged and rolled, which indicates satisfactory strength of the clad joint.     

Influence of laser beam variables on AZ91D weld fusion zone microstructure

Abstract

The influence of the laser beam variables on rapidly solidified magnesium alloy AZ91D weld microstructures was investigated using a continuous 1.2 kW Nd-YAG laser. To generate two- or three-dimensional heat flow thermal cycles, 2 mm thickness specimens were bead on plate welded using a focused 0.7 mm beam, 600 and 900 W power, and travel speeds from 10 to 110 mm s^-1. The fusion zones were examined via microscopy, electron dispersive spectroscopy (EDS), and X-ray diffraction (XRD) and tested for microhardness. Fusion zone dimensions were first measured and correlated with laser beam variables, and then used to estimate weld solidification times from well known formulations. Measurements showed that fusion zone hardness increased considerably for short solidification times (of the order of 15 ms). Hardness was lower near the fusion line, but greater where the last liquid solidified. Using EDS and XRD it was demonstrated that short solidification times led to greater fractions of redistributed aluminium in the last liquid left at the weld surface when heat flow was three-dimensional. Hardness was correlated not only with β-Mg₁₇Al₁₂ fractions, but also with finer microstructures, consistent with the Mg-Al phase equilibria and concepts of solidification.     

Dimensionless correlation to estimate peak temperature during friction stir welding

Abstract

A dimensionless correlation has been developed based on Buckingham's π-theorem to estimate the peak temperature during friction stir welding (FSW). A relationship is proposed between dimensionless peak temperature and dimensionless heat input. Apart from the estimation of peak temperature, it can also be used for the selection of welding conditions to prevent melting of the workpiece during FSW. The correlation includes thermal properties of the material and the tool, the area of the tool shoulder and the rotational and translation speeds of the tool. The peak temperatures reported in the literature during FSW of various materials and welding conditions were found to be in fair agreement with the proposed correlation.     

Numerical modelling of temperature distribution during multipass welding of plates

Abstract

Welding is a reliable, cost effective, and efficient metal joining process. Manual metal arc welding (MMAW) is a widely used welding process in industry and multipass welding is very frequently used in the MMAW process. The temperature distribution that prevails during multipass welding affects the material microstructure and hardness of the regions near the weld, and the residual stresses that will be present in the material after cooling to room temperature. These changes will consequently affect the performance of the welded joint. In the present work, a computer model based on the control volume method has been developed to predict the temperature distribution during multipass welding of plates using the MMAW process. To validate the computer model, the temperature distribution was measured experimentally in a 12 mm thickness stainless steel weld pad during multipass welding. The welding parameters were used as input data for the computer model. The computer predictions and the experimentally measured temperature distributions agree well.     

Determination of optimal brazing frequency by solution of thermal and electromagnetic models

Abstract

Control of the thermal cycles in induction heating has now been achieved, as a result of the use of software that simulates electrical and thermal phenomena. In brazing, however, the results are more disappointing. This is because, in comparison with the classical cases of heating before forging, or of thermal processing, the charge consists of at least three more or less separate parts and the thermal cycles of these must be simultaneous. First, relying on the resolution of Maxwell equations applied to a homogeneous cylinder, the present paper describes the problem of determining the optimal brazing frequency, which is a fundamental parameter in solving the multiple charge heating problem. Then a model is presented that will allow solution in the general case of both electromagnetic and thermal equations in a coupled manner. The general method is illustrated with the example of brazing carbide blades onto a tempered steel body.     

Solidification and microstructure modelling of welds in aluminium alloys 5754 and 6111

Abstract

A solidification and microstructure modelling approach has been developed to predict weld metal and heat affected zone (HAZ) characteristics. The freezing range and phase evolution in the weld metal region were predicted using thermodynamic and diffusion controlled growth calculations. The calculated freezing range was correlated with the weld solidification cracking tendency. A simplified analytical model was suggested to describe thermal cycles that are experienced by the HAZ. This analytical model was coupled with a published microstructure model for age hardenable alloys to predict the hardness variations across the HAZ. The above integrated approach was evaluated using experimental welds made on non­age hardenable 5754 (Al–Mg) and age hardenable 6111 (Al–Mg–Si) alloys using gas tungsten arc, electron beam, and gas metal arc welding processes.     

A Finite Element Approach for Die Thermal Design in High-pressure Diecasting and its Experimental Verification

Summary

The paper describes an investigation into the cooling transients in a high-pressure diecasting. A finite element model of the die and casting is presented and the predicted results are compared with thermocouple readings taken at strategic points in the die. As well as illustrating the complex thermal patterns in the die, the agreement between measured and predicted results are shown to be good, which confirms the applicability of the numerical modelling procedure.     

In process heat input control in arc welding

Abstract

An algorithm for in process optimum heat input control of arc welding is discussed, in which an analytical model of heat conduction is used to evaluate the temperature field during welding and the convex programming method is used in the optimisation of the heat input. The thermal response calculated using an analytical model corresponds well with that measured during welding and it is possible to identify the heat efficiency using the model. Furthermore, an in process optimising algorithm which includes the real time identification of heat efficiency is examined to estimate the optimum heat input for the required temperature field in gas tungsten arc welding under conditions where there is a change in the arc length. It is ascertained that the optimising algorithm is sufficient to determine the optimum heat input in real time during welding.     

Thermal-cycling creep of γ-TiAl-based alloys

In two-phase TiAl-based alloys, the coexisting α₂ and γ phases exhibit a thermal expansion mismatch, so that increased creep rates during thermal cycling may be expected. Creep deformation of two γ-TiAl-based alloys was investigated during thermal cycles between 900 and 300 or 350°C with applied tensile stresses of 32.5 or 37.0 MPa. Measured thermal cycling creep rates were compared with isothermal creep rates calculated at the effective average temperature. No creep enhancement was measured upon thermal cycling within uncertainties of 1.6×10⁻³% strain per cycle.