Showing papers by "Tarasankar Debroy published in 2006"

Numerical simulation of three-dimensional heat transfer and plastic flow during friction stir welding

Abstract: Three-dimensional visco-plastic flow of metals and the temperature fields in friction stir welding have been modeled based on the previous work on thermomechanical processing of metals. The equations of conservation of mass, momentum, and energy were solved in three dimensions using spatially variable thermophysical properties and non-Newtonian viscosity. The framework for the numerical solution of fluid flow and heat transfer was adapted from decades of previous work in fusion welding. Non-Newtonian viscosity for the metal flow was calculated considering strain rate, temperature, and temperature-dependent material properties. The computed profiles of strain rate and viscosity were examined in light of the existing literature on thermomechanical processing. The heat and mass flow during welding was found to be strongly three-dimensional. Significant asymmetry of heat and mass flow, which increased with welding speed and rotational speed, was observed. Convective transport of heat was an important mechanism of heat transfer near the tool surface. The numerically simulated temperature fields, cooling rates, and the geometry of the thermomechanically affected zone agreed well with independently determined experimental values.

Numerical modelling of 3D plastic flow and heat transfer during friction stir welding of stainless steel

Abstract: Three-dimensional (3D) viscoplastic flow and temperature field during friction stir welding (FSW) of 304 austenitic stainless steel were mathematically modelled. The equations of conservation of mass, momentum and energy were solved in three dimensions using spatially variable thermophysical properties using a methodology adapted from well established previous work in fusion welding. Non-Newtonian viscosity for the metal flow was calculated considering strain rate and temperature dependent flow stress. The computed profiles of strain rate and viscosity were examined in light of the existing literature on thermomechanical processing of alloys. The computed results showed significant viscoplastic flow near the tool surface, and convective transport of heat was found to be an important mechanism of heat transfer. The computed temperature and velocity fields demonstrated strongly 3D nature of the transport of heat and mass indicating the need for 3D calculations. The computed temperature profiles agreed well with the corresponding experimentally measured values. The non-Newtonian viscosity for FSW of stainless steel was found to be of the same order of magnitude as that for the FSW of aluminium. Like FSW of aluminium, the viscosity was found to be a strong function of both strain rate and temperature, while strain rate was found to be the most dominant factor. A small region of recirculating plasticised material was found to be present near the tool pin. The size of this region was larger near the shoulder and smaller further away from it. Streamlines around the pin were influenced by the presence of the rotating shoulder, especially at higher elevations. Stream lines indicated that material was transported mainly around the pin in the retreating side.

Liquid metal expulsion during laser spot welding of 304 stainless steel

Abstract: During laser spot welding of many metals and alloys, the peak temperatures on the weld pool surface are very high and often exceed the boiling points of materials. In such situations, the equilibrium pressure on the weld pool surface is higher than the atmospheric pressure and the escaping vapour exerts a large recoil force on the weld pool surface. As a consequence, the molten metal may be expelled from the weld pool surface. The liquid metal expulsion has been examined both experimentally and theoretically for the laser spot welding of 304 stainless steel. The ejected metal droplets were collected on the inner surface of an open ended quartz tube which was mounted perpendicular to the sample surface and co-axial with the laser beam. The size range of the ejected particles was determined by examining the interior surface of the tube after the experiments. The temperature distribution, free surface profile of the weld pool and the initiation time for liquid metal expulsion were computed based on a three-dimensional transient heat transfer and fluid flow model. By comparing the vapour recoil force with the surface tension force at the periphery of the liquid pool, the model predicted whether liquid metal expulsion would take place under different welding conditions. Expulsion of the weld metal was also correlated with the depression of the liquid metal in the middle of the weld pool due to the recoil force of the vapourized material. Higher laser power density and longer pulse duration significantly increased liquid metal expulsion during spot welding.

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.

Improving reliability of heat and fluid flow calculation during conduction mode laser spot welding by multivariable optimisation

Abstract: Several uncertain parameters affect the reliability of heat transfer and fluid flow calculations during conduction mode laser spot welding because their values cannot be prescribed from fundamental principles. These parameters include absorptivity of the laser beam, effective thermal conductivity and effective viscosity of liquid metal in the weld pool. Values of these parameters are usually adjusted by trial and error so that the computed results agree with the corresponding experimental values. Here it is shown that by integrating multivariable constrained optimisation with convective heat transfer and fluid flow calculations, the values of the uncertain parameters can be obtained from a limited volume of experimental data. The optimisation technique requires numerically calculated sensitivity values of weld dimensions with respect to absorptivity, effective thermal conductivity and effective viscosity and minimises the discrepancy between the predicted and the measured weld dimensions. The nume...

Mathematical modeling of heat transfer, fluid flow, and solidification during linear welding with a pulsed laser beam

Abstract: The evolution of temperature and velocity fields during welding of 304 stainless steel with a pulsed laser beam was simulated using a three dimensional numerical heat transfer and fluid flow model. The weld pool solidified between pulses and regions of the weld bead melted and solidified several times during welding. Short laser pulses restricted the width of the weld track and velocities in the weld pool. However, convection still remained an important mechanism of heat transfer in the weld pool. The computed high cooling rates during linear welding with neodymium-doped yttrium aluminum garnet pulsed laser operated at 140W average power, 20Hz frequency, and 5ms pulse duration were consistent with those observed in typical laser welding. After the laser beam was switched off, the mushy zone expanded, reaching its maximum size when no pure liquid region remained. Calculations of solidification parameters indicated that the criterion for plane front solidification was not satisfied. The results demonstrate ...

Non-isothermal grain growth in metals and alloys

Abstract: A large portion of the previous research on grain growth has been focused on isothermal conditions and excellent reviews on this topic are available in the literature. However, most of the materials processing operations such as casting, rolling and welding take place under non-isothermal conditions. Grain growth in materials under significant spatial and temporal variation of temperature exhibits many important special characteristics. The present paper critically examines these special features of grain growth in several important materials processing operations. Various experimental and theoretical approaches used to study grain growth in non-isothermal systems are examined and compared. The classical isothermal grain growth theories are reviewed, because they are sometimes applied to non-isothermal systems by incorporating thermal history in the calculations. Various numerical techniques and their applications to non-isothermal systems are also examined and compared. Finally, progress made in ...

Toward a unified model to prevent humping defects in gas tungsten arc welding

Abstract: During gas tungsten arc (GTA) welding, high welding speed and current can lead to a serious weld defect with a bead-like appearance known as humping. Currently, there is no unified model to predict the formation of humping defects in GTA welding. Here we propose and test a new comprehensive computational model that can predict and prevent the formation of humping defects considering the values of arc current, welding speed, nature of the shielding gas, electrode geometry, ambient pressure, torch angle, and external magnetic field during gas tungsten arc (GTA) welding. The model considers stability of the waves on the weld pool surface due to relative motion between the shielding gas and the liquid metal based on the Kelvin-Helmholtz instability theory. The main factors for the instability were found to be the velocities of the shielding gas and the weld metal, densities of the molten metal and shielding gas, weld pool size, and surface tension of the molten weld metal. The weld pool size and weld metal velocities were calculated by a numerical heat transfer and fluid flow model, and the shielding gas velocity was calculated from an analytical relation. Good agreement between the model predictions of humping and the independent experimental results from various sources show that the model can be used to prevent humping considering the effects of arc current, welding speed, nature of the shielding gas, electrode geometry, ambient pressure, torch angle, and external magnetic field during GTA welding. Recommendations are provided for the use of special electrodes and an external magnetic field and, where practical, controlled pressure and careful selection of shielding gas to prevent humping under conditions when high welding speed and current are needed to sustain productivity goals.

Tailoring weld geometry during keyhole mode laser welding using a genetic algorithm and a heat transfer model

Abstract: Tailoring of weld attributes based on scientific principles remains an important goal in welding research The current generation of unidirectional laser keyhole models cannot determine sets of welding variables that can lead to a particular weld attribute such as specific weld geometry Here we show how a computational heat transfer model of keyhole mode laser welding can be restructured for systematic tailoring of weld attributes based on scientific principles Furthermore, the model presented here can calculate multiple sets of laser welding variables, ie laser power, welding speed and beam defocus, with each set leading to the same weld pool geometry Many sets of welding variables were obtained via a global search using a real number-based genetic algorithm, which was combined with a numerical heat transfer model of keyhole laser welding The reliability of the numerical heat transfer calculations was significantly improved by optimizing values of the uncertain input parameters from a limited volume of experimental data The computational procedure was applied to the keyhole mode laser welding of the 5182 Al–Mg alloy to calculate various sets of welding variables to achieve a specified weld geometry The calculated welding parameter sets showed wide variations of the values of welding parameters, but each set resulted in a similar fusion zone geometry The effectiveness of the computational procedure was examined by comparing the computed weld geometry for each set of welding parameters with the corresponding experimental geometry The results provide hope that systematic tailoring of weld attributes via multiple pathways, each representing alternative welding parameter sets, is attainable based on scientific principles

Neural network model of heat and fluid flow in gas metal arc fillet welding based on genetic algorithm and conjugate gradient optimisation

Abstract: Although numerical calculations of heat transfer and fluid flow can provide detailed insights into welding processes and welded materials, these calculations are complex and unsuitable in situations where rapid calculations are needed. A recourse is to train and validate a neural network, using results from a well tested heat and fluid flow model to significantly expedite calculations and ensure that the computed results conform to the basic laws of conservation of mass, momentum and energy. Seven feedforward neural networks were developed for gas metal arc (GMA) fillet welding, one each for predicting penetration, leg length, throat, weld pool length, cooling time between 800uC and 500uC, maximum velocity and peak temperature in the weld pool. Each model considered 22 inputs that included all the welding variables, such as current, voltage, welding speed, wire radius, wire feed rate, arc efficiency, arc radius, power distribution, and material properties such as thermal conductivity, specific heat and temperature coefficient of surface tension. The weights in the neural network models were calculated using the conjugate gradient (CG) method and by a hybrid optimisation scheme involving the CG method and a genetic algorithm (GA). The neural network produced by the hybrid optimisation model produced better results than the networks based on the CG method with various sets of randomised initial weights. The CG method alone was unable to find the best optimal weights for achieving low errors. The hybrid optimisation scheme helped in finding optimal weights through a global search, as evidenced by good agreement between all the outputs from the neural networks and the corresponding results from the heat and fluid flow model.

A genetic algorithm and gradient-descent-based neural network with the predictive power of a heat and fluid flow model for welding

Abstract: In recent years, numerical heat and fluid flow models have provided significant insight into welding processes and welded materials that could not have been achieved otherwise. However, these calculations are complex and time consuming, and are unsuitable in situations where rapid calculations are desired. A practical solution to this problem is to develop a neural network that is trained with the data generated by a numerical heat and fluid flow model. Apart from providing high computational speed, the results of this neural network conform to the basic laws of conservation of mass, momentum, and energy. In the present study, six feed-forward neural networks have been developed for the gas tungsten arc (GTA) welding of low-carbon steel. Each network provides one of the six output parameters of GTA welds, i.e., depth, width, and length of the weld pool, peak temperature, cooling time from 800° to 500°C, and maximum liquid velocity. The networks require values of 17 input parameters including the welding variables like current, voltage, welding speed, arc efficiency, arc radius, and power distribution factor, and material properties like thermal conductivity and specific heat. The weights of the neural networks were calculated using two optimization schemes, first using the gradient descent (GD) method with various sets of randomized initial weights, and then applying a hybrid optimization scheme where a genetic algorithm (GA) is used in combination with the GD method. The neural networks produced by the hybrid optimization approach gave better results than all the networks based on only the GD method. Unlike the GD method alone, the hybrid optimization scheme could find the significantly better weights, which is illustrated by the good agreement between all the outputs from the neural networks and the corresponding results from the heat and fluid flow model.

Novel Optimization Methodology for Welding Process/Consumable Integration

Abstract: Advanced materials are being developed to improve the energy efficiency of many industries of future including steel, mining, and chemical, as well as, US infrastructures including bridges, pipelines and buildings. Effective deployment of these materials is highly dependent upon the development of arc welding technology. Traditional welding technology development is slow and often involves expensive and time-consuming trial and error experimentation. The reason for this is the lack of useful predictive tools that enable welding technology development to keep pace with the deployment of new materials in various industrial sectors. Literature reviews showed two kinds of modeling activities. Academic and national laboratory efforts focus on developing integrated weld process models by employing the detailed scientific methodologies. However, these models are cumbersome and not easy to use. Therefore, these scientific models have limited application in real-world industrial conditions. On the other hand, industrial users have relied on simple predictive models based on analytical and empirical equations to drive their product development. The scopes of these simple models are limited. In this research, attempts were made to bridge this gap and provide the industry with a computational tool that combines the advantages of both approaches. This research resulted in the development ofmore » predictive tools which facilitate the development of optimized welding processes and consumables. The work demonstrated that it is possible to develop hybrid integrated models for relating the weld metal composition and process parameters to the performance of welds. In addition, these tools can be deployed for industrial users through user friendly graphical interface. In principle, the welding industry users can use these modular tools to guide their welding process parameter and consumable composition selection. It is hypothesized that by expanding these tools throughout welding industry, substantial energy savings can be made. Savings are expected to be even greater in the case of new steels, which will require extensive mapping over large experimental ranges of parameters such as voltage, current, speed, heat input and pre-heat.« less

Experimental and computational investigation of fusion zone geometries during autogenous keyhole mode laser welds

Abstract: Here we report experimental and numerical modeling efforts targeted at autogenous keyhole mode laser welding of a low-carbon steel (A131 grade EH-36). In order to quantitatively understand the heat transfer and fluid flow processes, a mathematical model involving numerical solution of the equations of conservation of heat, mass and momentum in three dimensions was developed. The model considered formation of a keyhole, liquid steel flow in the weld pool driven by Marangoni convection at the weld pool surface, and heat transfer in the entire weldment. The computed results provide a detailed description of the temperature and velocity fields in the weldment, the shape and size of the keyhole, and the geometry of the fusion zone. There is excellent agreement between the experimental and modelled fusion zones.Here we report experimental and numerical modeling efforts targeted at autogenous keyhole mode laser welding of a low-carbon steel (A131 grade EH-36). In order to quantitatively understand the heat transfer and fluid flow processes, a mathematical model involving numerical solution of the equations of conservation of heat, mass and momentum in three dimensions was developed. The model considered formation of a keyhole, liquid steel flow in the weld pool driven by Marangoni convection at the weld pool surface, and heat transfer in the entire weldment. The computed results provide a detailed description of the temperature and velocity fields in the weldment, the shape and size of the keyhole, and the geometry of the fusion zone. There is excellent agreement between the experimental and modelled fusion zones.