A numerical model for deep penetration welding processes

The general features of a numerical model, and of its extensions, for calculating the temperature and fluid velocity field in a three-dimensional workpiece undergoing deep penetration laser beam welding are described. In the current model, the deposition of power from the beam is represented by time-dependent boundary conditions on the equations of energy and momentum transfer. These boundary conditions are specified at each timestep on a surface whose configuration can change with time and upon which energy is deposited according to a specified power distribution. This model also includes the effects of the buoy-ancy force on the melt pool and of the surface tension gradient on the surface of the fluid. The coupled equations of energy, momentum transfer, and continuity combined with the time-dependent boundary conditions representing the keyhole and the moving boundaries of the workpiece are solved by using a specific implementation of the SIMPLE algorithm. The important features of the numerical methods used in the model are discussed. Isotherms and convection patterns calculated using the current model are presented, and their significance for predicting weldment properties is discussed. A significant result of the simulations is that they demonstrate the overwhelming influence of the keyhole vapor/liquid inter-face on fluid convection and conduction in deep penetration welding.     

凝固过程中的颗粒推斥

颗粒推斥是材料在固液相变过程中的基础问题之一，界面自由能的传统理论难以解释在金属基复合材料制备过程中的颗粒推斥现象，作者提出了固液界面前沿的流体流动是颗粒推斥的主要原因之一，并给出颗粒滚动及颗粒速度两种作用机制，推导出颗粒推斥条件判据，并以实验验证流体流动的作用机制及颗粒推斥理论判据。     

Mechanisms increasing welding efficiency during new development of double shielded TIG process

Abstract

A new welding method has been developed for the tungsten inert gas process which can both improve the weld depth/width (D/W) ratio and protect the electrode from oxidation during the welding process. The effects of the welding parameters and the Marangoni convection patterns on the weld pool shapes are discussed. Results showed that the change of welding parameters directly change the temperature distribution on the pool surface and the concentration addition of the surface active element, hence, eventually affect the pattern and strength of the Marangoni convection. For the liquid pool on stainless steel, the weld shapes depends to a large extent on the pattern and strength of the Marangoni convection. It is possible to generate a heavy oxide layer on pool surface under a relatively high oxygen content to affect the surface tension stress. The heavy oxide layer may inhibit the heat flow on the weld pool surface, and reduce the D/W ratio.     

Error analysis of forward and reverse heat conduction and convection calculations considering uncertainties in welding

Abstract

Numerical models of fusion welding traditionally compute temperature field for a given set of welding conditions in a forward manner. The reliability of computed temperature profile depends on the accuracy of a number of model input parameters, values of which are uncertain in nature. Here, the authors show that a genetic algorithm (GA) assisted integrated numerical model, following either convection or conduction based calculations, can identify the suitable values of the uncertain model input parameters and in turn provide reliable computed results. Powered with GA, the integrated model is used further in a reverse manner to predict multiple sets of welding conditions for a target weld geometry. The convection based calculations have been able to provide more reliable multiple welding variables in reverse calculations.     

Effect of oxygen content in He–O2 shielding gas on weld shape in ultra deep penetration TIG

Abstract

The effect of the shielding gas concentration on the weld shape was studied for the moving bead on plate TIG welding of SUS304 stainless steel under He–O₂ mixed shielding. The small addition of oxygen to the helium base shielding gas can precisely control the oxygen content in a liquid pool and the weld shape. Oxygen is a surface active element for stainless steel. When the oxygen content in the liquid pool is above the critical value of ～ 70 ppm, the weld shape suddenly changes from a wide shallow type to a deep narrow one due to the change in the Marangoni convection from the outward to inward direction on the liquid pool surface. Weld shape variations influenced by the welding parameters including welding speed, welding current and electrode tip work distance under pure He and He–0.4%O₂ mixed gas shielding were systematically investigated. The investigation results showed that the final shape of the TIG weld depends to a large extent on the pattern and magnitude of the Marangoni convection on the pool surface, which is governed by the combined effect of the oxygen content in a liquid pool, temperature coefficient of the surface tension (dσ s/dT) and the temperature gradient on the pool surface (dT/dr, r is the radius of the weld pool surface). It is considered that the change in welding parameters alters the temperature distribution and gradient on the pool surface, and thus, affects the magnitude of the Marangoni convection and final weld shape.     

Effect of high magnetic field on solidification microstructure evolution of a Cu-Fe immiscible alloy

The liquid phase separation behavior and the evolution of the solidification microstructure of a binary Cu₅₀Fe₅₀ alloy were investigated under the conditions of without and with a 10 T magnetic field, with different undercooling during the solidification process. Results show that the combined effect of Stokes motion and Marangoni convection leads to the formation of the core-shell structure under the condition without the magnetic field. In addition, specific gravity segregation is reinforced by increasing the undercooling, resulting in Fe-rich phase drifts towards the sample edge. In the 10 T magnetic field, the Fe-rich phase is elongated in the parallel direction of the magnetic field under the action of demagnetization energy due to the difference of static magnetic energy and surface energy. In the vertical direction, through the action of Lorentz force, the convection in the melt is inhibited and Fe-rich phase becomes more dispersed. Meanwhile, the diffusion of the two phases and the coagulation of the Fe-rich phases are also restrained under the magnetic field, therefore, the phase volume fraction of the Fe-rich phase decreases at the same undercooling in the 10 T magnetic field. The magnetic field inhibits the segregation behavior in the vertical direction of the magnetic field, and at the same time, improves the gravitational segregation to a certain extent, which has a very important impact on microstructure regulation.

     

Probabilistic modelling of fatigue crack initiation from pits and pit clusters in aluminium alloys

Abstract

A numerical model of crack initiation under high cycle fatigue loading from pits is investigated in this paper. A probability based pit growth model, which takes into account the influence of mechanical cyclic load and particle clusters present in alloys, is used for investigations. Critical pit sizes, calculated using linear elastic fracture mechanics principles, are used to determine the probability of crack initiation for different conditions. The results are critically compared to extract an insight on the parameters that control the pit growth behaviour and thereby the fatigue crack initiation.     

Physical and mathematical modelling of argon shroud for protection of steel stream during casting

Abstract

Physical modelling of argon shroud was carried out using a two-dimensional 1 : 1 scaled model in order to gain a basic understanding of air entrainment during casting. The flow pattern of protective gas was determined with particle image velocimetry technique. Computational fluid dynamics calculations were also conducted. The results of the particle image velocimetry measurements and computational fluid dynamics modelling showed mutual agreement. The main findings of the present study are the following. First, the use of vertical gas inlet outside the collar on the top of the vertical runner should be avoided. Second, the distance between the ladle and the collar should be minimised to reduce the oxygen entrainment by convection and diffusion. Third, a modest flowrate of argon gas should be used. Further increase in the gas flow does not seem to improve the protection.     

Numerical simulation of the effect of fluid flow on solute distribution and dendritic morphology

Abstract

The presence of bulk and interdendritic flow during solidification can alter the microstructure, potentially leading to the formation of defects. In this paper, a numerical model is presented for the direct simulation of dendritic growth in the presence of fluid flow in both liquid and mushy zones. The Navier–Stokes equations are solved for multiphase flow using a projection method. The energy conservation and solute diffusion equations are solved via a combined stochastic nucleation approach and finite difference solution to simulate dendritic growth. The predicted microstructures illustrate typical asymmetric dendritic growth behaviour under forced convection, which is consistent with prior similar simulations of a single dendrite during unconstrained growth (both 2D and 3D). The micromodel was coupled with a macromodel to investigate the effects of forced fluid flow on equiaxed dendritic growth and micro-segregation during vacuum arc remelting.     

Stationary GMAW-P weld metal deposit spreading

Abstract

Non-isothermal spreading of non-travelling pulsed gas metal arc weld (GMAW-P) deposits was studied by experiments and numerical simulation. After an initial transient period, metal from the melting welding wire was deposited as a regular stream of droplets into a molten weld pool whose spreading was enhanced by direct arc heating of the solid substrate and weld pool convection. Accurate predictions of the final weld cross-section and time histories of base metal temperature and weld pool radius were produced by this simulation, which included mass, energy and momentum of transferring filler metal droplets. Dimensional analysis showed that speed of weld pool spreading was initially dominated by the momentum of liquid droplets while thermal convection by surface tension-driven flow became important at intermediate times. Eventually, the rate of spreading was matched to the thermal diffusion rate in the liquid. Capillary forces were never important in weld pool spreading. The weld metal deposit shape was well-approximated as a spherical cap with increasing volume and constant contact angle.     

Anodic Behaviour of Indium in Alkali

Abstract

Galvanostatic polarisation experiments have been made on indium electrodes in alkaline solutions. Long time and transient measurements are described. The anodic reaction conforms to passivation due to the progressive growth of a solid film on the electrode surface: under some conditions a precursory solution reaction may be observed. The polarisation behavlour is influenced by semiconductor properties of oxide formed in the indium hydroxide layer. ApplzcatlOn to possIble energy conversion devices is mentioned.     

Simulation of deep penetration welding of stainless steel using geometric constraints based on experimental information

Results of a numerical simulation of deep penetration welding of 304 stainless steel are presented. This numerical model calculates the temperature and fluid velocity fields in a three-dimensional workpiece undergoing deep-penetration electron beam welding. The deposition of power from the beam and energy outflow at the model-system boundaries is effected by means of time-dependent boundary conditions on the equations of energy and momentum transfer. The vapor-liquid interface defining the keyhole is represented by a surface whose temperature is that of vaporization for the steel. On this surface, are specified boundary conditions for the momentum transfer equations such that the component of the velocity normal to the keyhole vapor-liquid interface is zero. In addition, this study introduces two new numerical procedures. These procedures are based on the inclusion of experimental information concerning beam spot size and weld pool geometry into the model system via constraints and the deduction of effective keyhole shape via an inverse mapping scheme.     

Modelling of dendritic growth under forced convection in solidification of Al–7Si alloy

Abstract

An integrated microscale model is used to describe the dendritic growth morphology of Al–7Si alloy under forced convection, which combines the two-dimensional finite difference method (FDM) with the cellular automaton (CA) model. The FDM is used to simulate the induced fluid flow and solute transfer. The CA model is used to depict the dendritic morphology. Simulations are performed to investigate the influences of processing variables on morphological evolution under stirring. Calculated results reveal that at the intermediate undercooling, the growth morphology changes from dendritic to rosette as the rotation speed increases. The dendritic growth is promoted by increasing the undercooling. The rotation speed has a minor influence on microstructure formation for the cases of lower and higher undercoolings. A globulitic structure is formed as the nucleation density is increased. Changing the rotation speed is found to have a negligible influence on morphological evolution for the grain refined alloy.     

Numerical modelling of columnar to equiaxed transition – application to microgravity experiments

Abstract

A multiphase model based on statistical averaging is applied for simulation of the columnar-to-equiaxed transition (CET) during solidification of Al–7wt-%Si non-refined and refined alloys under microgravity conditions. Simulations are performed in one-dimensional approximation, taking into account real boundary conditions through thermal sources in the energy equation. The comparison between calculated positions of CET and measured positions in the MACE experiments is made. An agreement is obtained when using values of the nucleation undercooling of ΔT_Ne =6 K and ΔT_Ne =5 K for non-refined and refined alloy, respectively.     

Effect of surfactant redistribution on weld pool shape during gas tungsten arc welding

Abstract

In previous evaluations of GTA welding on stainless steel plates the surfactant content in the workpiece has been assumed to be in a state of equilibrium. However, surfactant concentration usually had to be adjusted in order to achieve agreement between the weld pool shapes in experiments and those in simulations. In this study a physicochemical approach to the redistribution of surfactant at the surface and in the bulk is presented for the first time. Sulphur was considered as a representative surfactant. The model allows surfactant molecular transport via convection, diffusion, and sorption. It is shown that sulphur atoms accumulate at the surface at stagnation points, affecting the coefficient of surface tension and thus the final weld pool shape. Thus, the sulphur content in the simulation approaches the nominal sulphur content in the steel plate. The theory is strengthened by both electron probe microanalysis and Auger electron spectroscopy.     

铜闪速炉内颗粒受热过程的数学建模与数值优化

为了研究闪速炉冶炼颗粒着火延迟问题,建立反应塔内颗粒受热过程的数学模型,并进行计算.结果表明,在塔高0.6 m范围内,辐射换热对于颗粒的加热过程影响较大,而在反应塔0.6 m以下区域,对流换热占主导作用.鉴于强化对流换热过程对提高颗粒加热速度进而提高着火高度更有效,因此,建议采用高速热氧射流技术代替闪速炉内的天然气,使...     

生长速率对NiAl-7.8Mo亚共晶合金定向凝固组织及界面形态的影响

采用高温度梯度定向凝固装置制备了NiAl-7.8Mo亚共晶合金,系统研究了生长速率对合金凝固时界面形态、枝晶生长及初生相析出的影响.随生长速率的增大,固液界面依次呈现平、胞枝、枝的形貌转变,初生β相的一次枝晶间距λ1在胞晶生长阶段逐渐增大,而在枝晶生长阶段λ1又逐渐减小;二次枝晶间距λ2随生长速率的增加一直减小.初生β相的析出量随生长速率的增加而增加.分析结果表明:生长速率增加导致熔体过冷度增加,使得NiAl初生相形核率增加,最终导致初生β相的析出量随生长速率的增加而增加.     

基于改进元胞自动机方法的强制对流作用下三维枝晶生长的数值模拟

建立了一种改进的元胞自动机模型来模拟熔体对流条件下的二元合金三维枝晶的生长。模型中考虑了界面能各向异性和溶质扩散对固/液界面推移的影响,在同一套网格中耦合求解质量传输和液相流动方程,从而可以模拟溶质扩散和熔体对流之间的相互作用。使用该模型模拟了一定过冷度条件下,强制对流对Al-7%Si(质量分数,下同)合金三维枝晶生长形貌的影响。模拟结果表明,熔体强制对流导致迎流侧尖端溶质富集层减薄,枝晶生长出现了迎流生长现象。将模拟得到的溶质过饱和度与Oseen-Ivantsov解析解进行对比,当流速较大时两者吻合较好。同时模拟了三维和二维强制对流作用下枝晶生长形貌的演化,由于三维条件下对流使得熔体能够绕过垂直于对流方向的一次枝晶臂主干将溶质带到背流侧,而二维条件下只能绕过垂直方向一次臂的尖端,因此三维MCA模型能更准确地反映强制对流对枝晶生长的影响。     

Electropolishing of horizontal tubes under conditions of natural convection mass transfer control

The kinetics of electropolishing of the inside surface of horizontal copper tubes in phosphoric acid was studied by measuring the limiting current of the process under natural convection conditions. Phosphoric acid concentration and tube diameter were varied to provide a range of ScGr extending from 3.3 × 10⁹ to 1.6 × 10¹¹. Under these conditions, the mass transfer coefficient of the electropolishing process was correlated to other variables by the dimensionless equation Sh = 0.62(ScGr)^0.28. Limiting current distribution measurement inside the tube revealed that the rate of polishing decreases from the top of the tube to the bottom.     

Decomposition of acid dissolved titanium slag from Australia by sodium hydroxide

The kinetics of the decomposition of acid dissolved titanium slag with a sodium hydroxide system under atmospheric pressure was studied. The effect of reaction temperature, particle size and NaOH-to-slag mass ratio on titanium extraction was investigated. The results show that temperature and particle size have significant influence on titanium extraction. The experimental data of titanium extraction show that the shrinking core model with chemical reaction controlled process is most applicable for the decomposition of slag, with an apparent activation energy of 62.4 kJ·mol⁻¹. Approximately 85 wt.%−90 wt.% of the titanium can be extracted from the slag under the optimal conditions. In addition, the purity of titanium dioxide obtained in the product is up to 98.5 wt.%.