Effect of workpiece vibration frequency on heat distribution and material flow in the molten pool in tandem-pulsed gas metal arc welding

doi:10.21203/rs.3.rs-2496490/v1

Download PDF

Research Article

Effect of workpiece vibration frequency on heat distribution and material flow in the molten pool in tandem-pulsed gas metal arc welding

https://doi.org/10.21203/rs.3.rs-2496490/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 16 Oct, 2023

Read the published version in The International Journal of Advanced Manufacturing Technology →

You are reading this latest preprint version

Versatility, high deposition rate, fine quality, and low equipment cost are known features of the tandem-pulsed gas metal arc welding (TP-GMAW) process. Concurrently, vibration-assisted welding has been considered one of the trends in developing advanced industrial concepts. This study presents a three-dimensional model of the TP-GMAW process to investigate heat transfer and material flow. The competition between the heat distribution beneath the surface and the physical movement caused by the workpiece sine-mode vibration are traced and discussed to understand how the penetration shape change was determined. It is found that applying the vibration extends the heat distribution along the welding direction beneath the weld pool surface, and this trend increases with increasing vibration frequency and effective heat input. In contrast, the heat extending is minimum in the sample without vibration having the highest heat input. This inconsistency can be explained by the physical movement of material in a molten pool due to the workpiece vibration. The vibration also changes the material flow from the surface to the depth in the central rear areas, although it follows an opposite direction in the sample without vibration. Thus, the material flow is vital in improving the penetration shape.

Pulsed gas metal arc welding

Numerical simulation

Tandem welding

In-process vibration

Thermal analysis

The effect of in-process vibration-assisted welding has been taken into consideration lately [1–5] as one of the trends in developing advanced manufacturing technology. However, there have been a few systematic investigations of the possible interactions concerning all aspects of the welding process associated with the workpiece vibration. Coincidentally, it was discovered that a specific kind of vibration mode transposed the finger-shape into pan-bottom penetration in tandem-pulsed gas metal arc welding (TP-GMAW) [6]. The penetration shape change leads to variation in the heat-affected zone (HAZ) thickness and the local mechanical properties. From the industrial point of view, the workpiece vibration is not an on-demand process in heavy processes; however, it has been investigated as an alternative for the enhancement of the mechanical properties of the weld metal. The requirement of the single pass tandem GMAW for joining the thick weld metal is an industrial concept for such cases. The penetration shape is important in improving mechanical properties among all the factors. The formation mechanism of the finger-type penetration has been discussed from different points of view [7–11] and for various welding techniques, such as hybrid TIG-MIG welding [12]. Aucott et al. [13] revealed an internal flow behaviour in arc welding by high-energy synchrotron radiation. They investigated the effective role of surface-active elements in controlling the internal melt flow. The effect of sulfur on creating both negative and positive surface tension gradients during welding was investigated by a comprehensive approach that led to more reliable simulation results [14]. Unlike the well-known ultrasonic frequencies found effective in many welding and joining applications [15–18], the process window of workpiece vibration and its variables are almost uncertain. Usually, the variables are limited by physical constraints. On the other side, optimizing the welding variables via experiments is almost time-consuming. The multiscale welding simulations have been proved in different conditions to visualize and predict the optimal process variables with finite experiments [19, 20].

The numerical simulations of gas metal arc welding (GMAW) have been developed over two decades. Cho et al. [21] proposed a model of weld pool dynamics in the stationary pulsed GMAW. Then, they developed the model to predict the weld bead hump formation in GMAW [22]. The effect of simulation methods on thermal characteristics in GMAW was simulated and visualized by Cheon et al. [23]. Wu et al. developed numerical models for the twin-wire [24] and hybrid GMAW process [25] using a comprehensive CFD-based simulation. The mechanisms influencing the molten pool flow were investigated and simulated in various conditions of driving forces [26], undercut formation [27], transverse magnetic field [28], arc interactions [29], groove shape [30], double-sided [31], double-shielded [32], and double-pulsed [33] GMAW. The improvement of numerical models is still ongoing, especially in the topics such as wire-arc additive manufacturing (WAAM) [34, 35], where new industrial applications are emerging.

Despite significant progress in previous research, adding the extra component, such as a specific workpiece vibration, to a multi-physics welding simulation is still challenging, depending on the numerical method. Most of the reports in this area focused on stress reliving and grain boundary enhancement [36–38] while neglecting the arc behaviour in the presence of the vibration. This research attempts to clarify a ternary relation between the workpiece vibration, the arc interaction, and the mechanical properties. Utilizing the potential of the improved Finite Difference Method (FDM), the analysis of the fusion welding zone was performed in good agreement with the experimental results. The previous research [14] focused on penetration shape change mainly in two samples with (W/) and without (W/O) vibration, and the representation of the change using the FDM simulation succeeded. The vibration increased material flow to weld toes in the rear part of the molten pool due to the Marangoni flow, affecting the penetration shape. Resistance of the arc was increased by the vibration that lowered the current value, followed by agitating the weld-pool surface. The decreasing heat input changed the material flow direction in the molten pool and enhanced the Marangoni flow. In the present study, other vibration frequencies are applied to the TP-GMAW, and the vibration effect on the heat distribution and material flow in the molten pool was investigated through the FDM simulation. The previous study did not focus much on vibration frequency difference. Thus, the effect of the workpiece vibration frequency on the total balance of heat input and material flow in the molten pool is investigated in the present article.

A robotic TP-GMAW machine was employed in this study, supported by the two independent power sources that fed the electrodes through the welding torch. The power sources were synchronized in an anti-phase manner to prevent mutual interaction of the trailing and leading arcs in pulsed conditions. A 6-axis robotic arm, KR30, KUKA make, was interfaced with Fronius make power sources. According to previous studies, a mixed shielding gas containing 82% Ar and 18% CO₂ was used at a rate of 12 L/min per cylinder to ensure spray transfer [14]. The currents at the electrodes were set to 180 A under the control of synergic welding equipment. The torch was set parallel to the welding direction in a flat position, and the bead-on-plate welding was carried out for a length of 0.2 m per sample. A hot-rolled mild steel ASTM A1011 was used as the base metal in the present study consisting of the ferrite-pearlite structure. The wire electrode employed was ER 70-S of 1.2 mm in diameter made by ADOR. The chemical compositions of the base material and welding wire are listed in Table 1. The average feed rate is set at 7.4 m/min per wire. The slip table connected to the Sdyn SEV-240 Electro-dynamic vibration machine was vibrated at the desired frequency ranges parallel to the welding direction. The wedge clamps were used to fix the workpiece tightly to avoid potential distortion during welding.

Table 1

The composition of the base material and filler wire (wt%).
	C	Si	Mn	P	S	Fe
ASTM A1011	0.211	0.206	0.710	0.015	0.018	Bal.
ER 70-S	0.07–0.15	0.80–1.15	＜1.85	＜0.025	＜0.035	Bal.

The sine-mode vibration was applied in parallel to the welding direction. A range of vibration frequencies of 50 to 450 Hz was conducted at a fixed acceleration of 1.2 m/s², as shown in Fig. 1(a). The velocity of sinusoidal vibration can be drawn from the mathematical relation between the vibration parameters; for example, it changes continuously from 0.76 mm/s to -0.76 mm/s when the frequency is set to 250 Hz. The average background and pulsed peak currents of the pulsed waveform were measured in the range of 86 to 96 A and 322 to 330 A, respectively, in the samples W/ and W/O vibration, as shown in Fig. 1(b). One droplet per pulse transition was maintained at around 150 Hz pulse frequency. The tandem interval time and the pulse duty are set to 2.2 and 3.2 ms, respectively. Keeping the contact tip to workpiece distance (CTWD) at 20 mm and the electrode distance of 8 mm, the bead-on-plate welds were laid at a constant welding speed of 1.2 m/min. Cross-sectional microstructure observations of as-welded specimens were obtained using a digital single-lens reflex (DSLR) camera occupied by an optical macro (OM) lens after etching by 93 vol% Methanol + 7 vol% Nitric acid.

A three-dimensional (3D) Cartesian model is presented to analyze the profile and convection in the weld pool following the TP-GMAW process. A laminar flow of incompressible and Newtonian fluid is thought in the computational domain to support non-isothermal phase change between solid and liquid phases. The effect of body forces, such as the electromagnetic and buoyancy forces, are considered in the molten pool, together with Marangoni convection. The effect of plasma heat on the droplets is simplified as a constant value of the droplet’s initial temperature. To visualize the characters of the fusion zone and track the flow, the Finite Difference Method (FDM) is employed besides three governing equations of mass continuity, energy conservation, and momentum (Navier–Stokes) [7]. In addition, the Lagrangian-based volume of fluid technique (VOF) is used for computing the free surface of the molten pool [39]. By neglecting the volume variation of the phase change, the porous media drag concept is applied for modelling the flow of the mushy zone. To handle the computational domain, the commercial code, Flow-3D (v.11.2u6 general package), is used while keeping high accuracy in tracking the moving free surface and the significant optimizations of the simulation cost. A simplified model of the electromagnetic force is used in double-ellipse mode by adding a correction factor [40]. The pressure boundary condition is expressed based on Newton’s viscosity law. Arc pressure in this study is assumed to have a double-ellipse distribution, where the magnitude and the effective radii are estimated using empirical measurement [14]. The plasma jet shear is defined by an analytical solution [41] where the velocity of the jet and the plasma Reynolds number are calculated by the maximum arc pressure employing Bernoulli’s equation.

The surface of the base material is exposed to electrode heat sources while losing heat in the form of convection and radiation. In the presence of trailing and leading heat sources, the mathematical expression of heat flux ($q$) at the surface is given as follows:

$$K\frac{\partial T}{\partial \overrightarrow{n}}={q}_{trail}+{q}_{lead}-{h}_{c}\left(T-{T}_{0}\right)-\epsilon {e}_{S-B}\left({T}^{4}-{T}_{0}^{4}\right)$$

where $K,\overrightarrow{n}, {h}_{c}, {T}_{0}, \epsilon ,$and ${e}_{S-B}$ represent the thermal conductivity, normal vector of the free surface, convection coefficient, ambient temperature, emissivity of the workpiece surface, and the Stefan-Boltzmann constant, respectively. The heat power of each source ($Q$) is assumed to be equal to the sum of the arc heat rate and the droplets’ heat content as follows:

$$\eta Q=\eta {Q}_{arc}+\eta {Q}_{drop}\to \eta \stackrel{-}{UI}=\eta \left(1-{\eta }_{d}\right)\stackrel{-}{UI}+\eta {\eta }_{d}\stackrel{-}{UI}$$

where the $\eta$ is the welding efficiency, $\stackrel{-}{UI}$ represents the instantaneous average power and the ${\eta }_{d}$ represents the ratio of the droplet heat rate and average power [42]. The heat flux on the free surface is modelled as a double-ellipse distribution of the modified fixed Gaussian relation with effective radii in the rear of the $x$ direction (${\sigma }_{qxr}$), the front of the $x$ direction (${\sigma }_{qxf}$), and the $y$ direction (${\sigma }_{qy}$) [25]:

$${q}_{arc}=\frac{\eta \left(1-{\eta }_{d}\right)UI}{2\pi {\sigma }_{q\stackrel{-}{x}}{\sigma }_{qy}}exp\left\{-\frac{{\left(x-{x}_{c}-{V}_{w}t\right)}^{2}}{2{\sigma }_{qx}^{2}}-\frac{{\left(y-{y}_{c}\right)}^{2}}{2{\sigma }_{qy}^{2}}\right\}$$

$${\sigma }_{qx\left(trail \right)}=\left\{\begin{array}{c}{\sigma }_{qxf}, x\ge {x}_{c}-{V}_{w}t \\ {\sigma }_{qxr}, x<{x}_{c}-{V}_{w}t\end{array}\right., {\sigma }_{q\stackrel{-}{x}}=\left(\frac{{\sigma }_{qxf}+{\sigma }_{qxr}}{2}\right)$$

To estimate the heat energy transferred by the droplets to the workpiece (${Q}_{drop}$), the heat capacity equation is employed, assuming the spray transition mode as follows:

$${Q}_{drop}=\frac{4}{3}\pi {r}_{d}^{3}{f}_{d}\left\{{\rho }_{s}{C}_{s}\left({T}_{s}-{T}_{0}\right)+\rho {L}_{f}+{\rho }_{l}{C}_{l}\left({T}_{d}-{T}_{l}\right)\right\}$$

where ${r}_{d}$, ${f}_{d}$, $\rho ,$ $C,$ and ${L}_{f}$ are the initial radius of the droplet, frequency of droplet generation, density, specific heat capacity, and latent heat of fusion, respectively.

The momentum of the workpiece vibration is another boundary condition applied in this investigation. The displacement, ${x}_{l}$, and instant velocity, $v$, of the workpiece for a given frequency of sinusoidal vibration can be expressed as:

$${x}_{l}=\left(D/2\right)sin\left(2\pi ft\right)$$

$$v=D\pi fcos\left(2\pi ft\right)$$

where $D$ is the peak-to-peak displacement and $f$ is the frequency of workpiece vibration. The acceleration reaches a peak, ${G}_{p}$, when $sin\left(2\pi ft\right)=1$, as shown in Eq. 8. The vibration is defined as the continuous change in velocity of the workpiece at fixed acceleration. The interaction of the workpiece vibration with the molten pool is investigated during the welding simulation.

$$\left|{G}_{p}\right|=2D{\left(\pi f\right)}^{2}$$

The 3D computational domain is divided into two regions, fluid and air-filled, where the enthalpy relation [14] is used to simulate the phase change of solid and liquid in the pseudo-fluid region. The cell’s temperature is calculated from its enthalpy, considering the convection and conduction. The total symmetric domain of 60x18x7.5 mm (parallel to X, Y, and Z directions, respectively) is assumed to be the solid phase at the start of the simulation by 5 mm thickness and 2.5 mm thickness of the void. The welding direction is parallel to the X direction. To reproduce a clear penetration zone, the domain is meshed by non-uniform rectangular cuboid cells of 0.15 mm in the central region and the utmost 0.2 mm in the surrounding. One of the most challenging parts of the tandem wire’s simulation is the impingement of the droplets from the two sources to the free surface, which occurs relatively closer than the initial distance between the two sources. The wire speed, torch angle, and arc interaction affect the droplet path. To improve the simulation efficiency, the arc interaction is simplified to only two conditions of the current mode: background-pulse and pulse-background, while the transitional effect of arc interaction is fairly ignored due to the 2.2 ms delay between pulse-charging. In addition, it is assumed that the spherical droplets generate at 1.5 mm above the top surface at a 6 mm distance versus an initial 8 mm, alongside the pseudo-parallel sources, as shown in Fig. 2(a). The TP-GMAW produces finger-type penetration, as illustrated in Fig. 2(b). The cross-sectional OM image shows that the penetration shape is not exactly symmetric, and a slight deviation can occur. Once the torch is rotated by 90 degrees around the Z axis, keeping other conditions the same (Fig. 2(c)), the deposited area is similar, while the penetration shape is completely different, and the effective distance between the leading and the trailing torches is estimated to be 6 mm. The mass balance of the feed rate (7.4 m/min) and the droplet generation frequency (150 Hz) is employed to estimate the initial radius of the droplet, 0.59 mm, while it keeps constant in all simulations. On the other hand, the current fluctuation is assumed to affect the superheated droplet’s temperature regardless of its size. The droplet generation frequency follows the current pulse frequency, which means one drop per pulse is produced in a semi-stable condition of the spray mode with respect to shielding gas [43].

The total efficiency of 68% is adopted for the GMAW process, while the ratio of the droplet heat rate per average power is estimated at 36% and consequently, the droplet efficiency, $\eta {\eta }_{d}$, and the arc heat efficiency, $\eta \left(1-{\eta }_{d}\right)$, are set to 24.5% and 43.5%, respectively [44]. The metal evaporation is ignored in this study due to the maximum temperature remains below the boiling temperature (2800 K) during the welding procedure. As one of the most important effects of workpiece vibration, decreasing current values with decreasing the vibration frequency was detected in both background and the pulse mode (Fig. 1(b)). To consider this fluctuation, the average background and pulse current are measured using a histogram of recorded current values during the welding. The initial temperature of the droplets is estimated at around 2500 K for W/O vibration cases based on previous studies [44]. Adopting the current change in Eqs. 2 and 5, the initial temperature of the droplets is set to 2370, 2400, and 2440 K for 50, 250, and 450 Hz of vibration frequency, respectively. The arc efficiency is fixed for both W/ and W/O vibration cases as the initial velocity of the droplets is set to 0.7 m/s equally [45]. The voltage varies from 25 to 39 V, for which the arc length of each electrode is considered to be 4.8 mm in the stable welding mode [21]. In order to estimate the double-ellipse effective radii, the arc interaction phenomenon is employed in this study introduced by Ueyama et al. [46]. Using the current values of the leading electrode at the background (96 A) and the trailing at pulsed mode (330 A), the maximum deflection of the pulsed arc is calculated at 0.56 mm. The thermophysical properties of ASTM A1011 steel applied in this simulation are listed in Table 2.

Table 2
Thermophysical properties of A1011 and the welding environment.
Property	Value
Ambient and the initial material temperature	300 (K)
Convection coefficient	100 (W/m².K)
Density*	6.4–7.8 (g/cm³)
Density of plasma argon (10,000 K, 1 atm)	4.6$\times$10⁻⁵ (g/cm³)
Drag coefficient constant	1 (1/s)
Emissivity of weld material	0.5
Latent heat of phase change (fusion)	2.75×10⁵ (J/kg)
Magnetic permeability of material	1.26$\times$10⁻⁶ (H/m)
Maximum plasma arc pressure (330 A)	908 (Pa)
Solidus and Liquidus temperatures	1698, 1778 K
Specific Heat*	0.50–0.95 J/g.K
Thermal Conduction*	26–50 (W/m.K)
Viscosity*	2.0-5.5 (g/m.s)
Viscosity of plasma argon (10,000 k)	0.05 (g/m.s)
* Temperature-dependent

To improve the simulation accuracy, the physical properties are considered temperature-dependent [14]. The surface tension of the free surface is defined in a pseudo-binary Fe-S system [47] as follows:

$$\gamma =1.94-4.30\times {10}^{-4}\left(T-{T}_{m}\right)-RT\times 1.30\times {10}^{-5}\times ln\left[1+3.18\times {10}^{-3}\times {a}_{s}{e}^{\left(1.66\times {10}^{5}/RT\right)}\right]$$

The average sulfur of 0.028 wt% is approximated in terms of the dilution corresponding to the base metal and the filler material content, 0.018 and 0.035 wt%, respectively. Note that the gradient of surface tension $\left(\partial \gamma ∕\partial T\right)$ changes its sign from positive at lower temperatures to negative at elevated temperatures, making a clear change in result compared to the model with a constant value of the negative gradient. The oxygen effect on the surface tension gradient is ignored due to the high amount of manganese and silicon in the filler composition [48]. A range of vibration frequencies of 50 to 450 Hz at a fixed acceleration of 1.2 m/s² was applied to the workpiece. For generating the pulsed waveform in compliance with recorded results, a trigonometric relationship is suggested as follows:

F $\left(I,t\right)={I}_{1}+\left(\frac{{I}_{2}-{I}_{1}}{2}\right)(1+sin(\frac{2\pi }{{P}_{d}}t-\frac{\pi }{2}\left)\right)$ (10)

where the ${I}_{1}$, ${I}_{2}$, and ${P}_{d}$ are the background current, the pulsed current, and the pulse duty of the total waveform, respectively. A similar relation is applied to generate the voltage waveform in the simulation. The recorded average power per electrode, 4.96 kJ/s, compared with the simulated average power of 4.74 kJ/s, admits the model’s accuracy within 5% error. The generation time of the droplets is synchronized with the waveform once the current pulse ends up at each cycle during the total time of the simulation. The generalized minimum residual (GMRES) method is employed as the pressure-velocity solver (with the subspace of 30 and the split Lagrangian VOF technique) is used with the explicit solver for the viscous stress. The implicit solver was considered for the heat transfer with a 1.5 factor of over-relaxation. The maximum time step is set to ${10}^{-5}$ s is limited by the convergence criteria of both surface tension and advection. Fluid fraction clean-up is the other numerical option used to decrease the possible divergence created by tiny droplets with a volume less than 5% of cell volume in the air-filled region.

4.1. Emphasis on validation of the weld pool in the present study

To validate the simulated weld pool, the cross-sectional OM image of the experimental weld is used and the formation mechanism is discussed in detail. As a typical method, a single cut of the simulated cross-section is used for the comparison; however, it may not necessarily be a comprehensive validation. The weld pool with a noticeable volume of melted material and a longer solidified time shows a dynamic pattern as a simultaneous function of position and time. On the other hand, the boundary between liquid and solid phases in the experiments may have been created at a consecutive time during the welding instead of generation at an instance. This phenomenon is one of the suggested mechanisms for the creation of finger-type penetration. The bottom of the puddle is formed at the very beginning, followed by penetration deepening at the weld toes in the next step. Analyzing a specific time frame can help to solve the problem in the desired cross-section (in the YZ plane), although, in this study, the heat and material flow change along the welding direction (X) pose a technical limitation. In fact, it is impossible to select a single unique position in all four simulated domains corresponding to the W/ and W/O vibration samples to represent the final melted area for the desired time steps. To overcome this limitation and scalability of the results, it was assumed that all the simulations reach ‘DYNAMIC STABILITY’ in the late simulation time (1.5 s < t < 2.2 s). In the following, the fusion boundary in the solid phase and the boundary between liquid and void in the air-filled are extracted from the four unique positions simultaneously. The boundaries overlaid on a 2D cross-sectional image together, the outmost boundary in accumulated boundaries can represent the final geometry of penetration. Note that comparing high-resolution simulation and experimental results and consequent validating the numerical results are the challenges the proposed numerical model overcomes.

4.2. Penetration shape change at four unique positions

In the present study, three of the four unique X positions of 39.08, 41.78, and 43.43 mm at the 1.96 s, reaching the dynamic stability in the simulation sample, are chosen for comparison and validation. The 3D simulation result of the W/O vibration sample is shown in Fig. 3. The criteria for selecting X positions are the fusion boundaries along the welding direction. Due to the dynamic nature of melting, these X positions do not necessarily show similar melting progress in all the simulation samples, and an intermediate state has been considered. The maximum height of the weld bead was estimated based on the cross-section at the rear section (X = 14.03 mm, the remaining one of the four unique positions) due to the limitation of both domain dimensions and simulation time.

Figure 4 illustrates a comparison of the simulation results with the OM images (left half of Fig. 4(a), (c), (e), and (g)) of the samples; 50, 250, 450 Hz of vibration, and W/O vibration, respectively. The simulated boundaries in the solid phase refer to the cells that partially melted by 50% in the puddle and the outer boundaries in the air-filled void with 95% of the filled material. The results show a reliable trend as the overprediction of the weld width is limited to a maximum of 4%, except about 8% in the 450 Hz vibration. The prediction of the height using the rear section at 14.03 mm made another prominence due to the relatively narrow width compared to the solidified beads. Changing the effective current depending on vibration frequency during welding of different experiments of W/ and W/O vibration (Fig. 1(b)) affects the fusion zone's depth as predicted by the numerical results. The maximum deviation of the numerical results is observed in the 450 Hz sample (Fig. 4(f)), considering the OM images are relatively symmetric compared with the other W/ and W/O vibration samples. In summary, the simulation almost successfully predicts the fusion zone with an acceptable margin of error. Simplified assumptions, including the position of heat sources and the material properties, are suggested as the main factors that cause the deviance. Considering the HAZ thickness, weld penetration depth, and symmetric pattern, the 250 Hz sample shows an almost optimal condition. The mechanism of the formation of the finger-type penetration has been well-studied in different cases of GMAW [7, 14]. With the collision of the droplets produced by the leading electrode and exposing the molten-pool surface to the arc, a shallow melting layer is created, which immediately deepens with the impact of the droplets from the trailing electrode.

The simulation results indicate that the deepest melted zone in the OM images was formed at this stage and the deepest in the W/O vibration sample with the maximum droplet temperature. On the other hand, the toe sides of the weld pool can change shape and melt more by the heat and material flow behind this pool position, such as Marangoni flow and the longitudinal flow of material along the welding direction that is an effect of the workpiece vibration along the welding direction. With respect to the estimated transition temperature of the Marangoni flow (2150 K, Eq. 9) in the presence of surface tension active elements, the negative gradient of surface tension below the torch is considerably weak, although it starts to change to positive values and increase the inward flow away from the torch. The penetration shape is determined by the overlap of these two melting stages just below the droplets and later near the weld toes. The penetration curvature at the intersection of fusion boundaries in the two melting stages is the minimum in the 250 Hz of vibration in both experiment and the simulation, resulting in the formation of the thinnest HAZ.

4.3. Heat distribution beneath the molten pool surface varied with vibration frequency

The following discusses the heat and material flow patterns in different vibration-frequency conditions. Figure 5 shows temperature maps of the free surface in the top view images of the simulated results around 1.96 s. The results confirm that the heat distribution in the upper part begins to extend along the welding direction by applying the vibration, and this trend increases with increasing vibration frequency. This is a unique and common trend in the W/ vibration samples found in the present study since the trend is not observed in the W/O vibration sample. As shown in Fig. 1, vibration velocity and distance along the welding direction decreased at the fixed acceleration, while heat input increased with increasing vibration frequency (the maximum heat input is shown in the W/O vibration sample). This suggests that the trend is a kind of result caused by the balance between workpiece vibration and heat input. This change can be considered due to the change in the material flow inside the weld pool. At this point, it is necessary to consider a series of small changes. First, extending the heat in the presence of vibration has reduced the effective width of the heat distribution. Next, by increasing the vibration frequency, the heat in the areas far from the torch is concentrated in the centre of the pool, which can change the Marangoni flow pattern. The images taken from the surface of the experimental welds can confirm the heat extending in the central parts. The solidified morphology of the weld cap is demonstrated in Fig. 6. In the 250 Hz sample (Fig. 6(a)), the freezing lines are rotated by 9 degrees, which is a little lower than 10 degrees in the W/O vibration sample (Fig. 6(b)), in addition to reducing the width of the weld pool, confirming the suggested change in heat distribution. However, the width of the weld pool in the W/ vibration samples did not decrease with increasing vibration frequency since effective heat input increased. It suggests that competition between heat distribution and effective heat input resulted in the width change.

It is suggested that the arc length decreases. Consequently, the resistance of the arc also decreases by increasing the vibration frequency (decreasing the vibration distance (Fig. 1(a))), which leads to agitating the surface of the weld pool. These changes lead to an increase in the current value. The current change is not too significant to make the arc unstable, while the frequencies (and corresponding vibration distances) used in this study are adequate to change the penetration shape by affecting the balance of the weld pool. It should be noted that in a dynamic balance among the forces acting on the weld pool, a cyclic pattern may detect between two dominant forces. In simple words, the molten material follows a reciprocating flow depending on the time of the simulation in a specific zone before solidifying. Due to the technical limitations of simulation, it is possible to assume constant time and check different cross-sections. This hypothesis is based on the that the movement of the melt has a relatively steady state. At the same time, an arbitrary cross-section can experience all three simulated sections over time.

4.4. Visualizing material flow in molten pool varied with vibration frequency

The images from the rear part of the molten pool are used to visualize how the Marangoni flow controls this region and affects the penetration shape, especially in the weld toes (Fig. 7). In order to track the paths of the fluid element, the streamlines are superimposed at an arbitrary location. As a general trend, the nearest cross-section at 43.43 mm to the trailing arc shows a finger-type shape in all samples. The W/O vibration sample clearly showed heat spreading towards the weld toe near the surface instead of the central area (Fig. 7(j)-(l)), and this causes finger-type penetration. On the other hand, the two vortexes near the centre and toe inside a molten pool are affected by the balance between extending heat distribution beneath the surface and heat input, depending on vibration frequency. The vortex near the toe disappeared at 41.78, 39.08 mm in the W/O vibration sample, while the vortex disappeared similarly in the 50 Hz sample (Fig. 7(b), (c)), appeared in the middle position between the weld centre and toe at the 41.78 mm in the 250 Hz sample (Fig. 7(e)), and appeared near the weld toe at the 41.78, 39.08 mm in the 450 Hz sample (Fig. 7(h), (i)). It is noted that the centre vortex disappeared from the finger-type puddle in the cross-section, and a simple one-way pattern can be recognized at 43.43 mm in the 250 Hz sample (Fig. 7(d)). The vortex in the middle position between the weld centre and toe at the 41.78 mm deepened middle penetration (Fig. 7(e)) and material flow at the 39.08 mm deepened toe penetration, similar to the W/O vibration sample (Fig. 7(f)). Consequently, the overlap of these two melting stages made the penetration curvature around the intersection of the two stages minimum, and the HAZ thinned in the 250 Hz sample among the W/ vibration samples. The penetration shape formation, depending on the vibration frequency suggested by the simulation, agrees with the experimental observation of penetration shape.

The longitudinal central plane in the simulation along the welding direction (Y = 0) can help to understand better the changes made by applying vibration. Figure 8 illustrates the longitudinal cross-sections in the four W/ and W/O vibration samples at 1.96 s in simulation time, with white boundaries indicating 50% in the solid fraction. Similar to Fig. 7, the streamlines of material flow added to the heat distribution pattern can create a good vision of fluid flow at the rear part of the molten pool before solidifying. In these images, the application of vibration along the welding direction has changed the material flow from the surface to the depth in the central rear areas (left side of Fig. 8(a)-(c)). In contrast, it follows an opposite direction in the W/O vibration sample (left side of Fig. 8(d)). As discussed earlier, the heat distribution in the W/ vibration sample extended along the welding direction, while the W/O vibration sample expanded towards the weld toes. Here again, the difference in the heat distribution pattern has caused a different material flow between W/ and W/O vibration samples. Although material flow in the W/ vibration samples was similar, the centre vortex position and size differed. The simulations indicate that the length of the white boundary between white arrows at maximum depth is less in the 250 Hz sample than in other W/ vibration samples in Fig. 8(a) to (c). The length can affect the HAZ shape under the molten pool, resulting in the thinnest HAZ.

Here, a possible contradiction between the heat input and heat distribution needs to be discussed. The recorded current values show the effective heat input increased monotonically by increasing the vibration frequency. In contrast, in the highest heat input (W/O vibration) case, the heat extending along the welding direction is minimum in the topmost layer of the molten pool. This inconsistency can be explained by the physical movement of material in a molten pool due to vibration. It is suggested that with the increase in the frequency of vibration, the role of the physical movement of the molten pool becomes dominant, and the dynamic balance goes towards the part of the high-temperature material being transferred to the central and rear areas of the pool.

According to the streamlines of material in plane Y = 0, the way the molten material returns to the rear of the molten pool is a key difference between the W/ and W/O vibration samples. In the presence of vibration, the surface areas consist of material with a higher temperature, which causes a change in the heat distribution pattern near the free surface. From the other point of view, in the presence of vibration, the rear surface (left side of Fig. 8) consists of material supplied from the top of the molten pool. In contrast, in the absence of vibration (Fig. 8(d)), the surface consists of material supplied from the depth of the molten pool. The results confirm that the material flow toward the toes (Y<, > 0) plays a more important role in supplying material to the rear zone of the weld pool and keeping the mass balanced during welding.

The 3D streamlines of material flow can derive a better vision in the W/O vibration sample. As illustrated in Fig. 9, the contour of the average macroscopic internal energy in the entire domain represents an energetic view of the welding process besides the projected 3D streamlines.

A large vortex can be detected at the centre associated with a large long stream to the rear of the weld pool (A₃) and a small stream to the front surface (A₁). The specific zones with higher internal energy show a more chaotic pattern. As a result, the only dominant vortexes associated with narrow streamlines are shown in these areas for simplification, and complicated vortexes and much narrower streams are not shown. The traces of the dominant vortex are observable near the centre and the bottom of the weld pool in Fig. 7(j), (k), and Fig. 8(d). In agreement with the discussion on Fig. 8(d), Fig. 9(b) shows how the material moves upwards in the rear region of the molten pool in two paths, A₂ and A₃. In the A₂ path, the material moves upwards to the central surface (Y = 0 mm); in contrast, the material moves backwards along the molten-pool bottom in the A₃ path, followed by moving upward to weld toes as demonstrated in a transparent mode in Fig. 9(b). On the other hand, due to the higher concentration of internal energy in the central zone of the W/ vibration samples, the scalable results could not be extracted for comparison in an isolated section. Plotting the streamlines in the W/ vibration samples reveals extremely complicated lines compared with the W/O vibration sample.

4.5. Heat conduction history in HAZ depending on vibration frequency

Another outcome is the prediction of cooling rate using a point probe in a simulated HAZ. In this analysis, two fixed probes are used in the HAZ, explaining that probe B has a longer distance in a diagonal direction on the YZ plane from the fusion boundary than probe A. The time variation of a probe temperature, together with the relative position of the probes, is given in Fig. 10. In the W/O vibration sample, which exhibits the highest input heat, a two-step pattern can be traced in both probes. After two seconds of simulations, the temperature goes through a short-term stable condition, unlike the other samples except for the 450 Hz sample, which drops immediately after reaching the peak temperature. The heat input in the 450 Hz sample is close to the W/O vibration sample, indicating that it has a relatively similar time variation of a probe temperature to the W/O vibration sample. These graphs show the importance of considering several cross-sections to predict the fusion boundary in Fig. 4. In both 50 and 250 Hz samples, the cooling rate is almost the same. Still, the peak temperature is 40 to 50 K higher in the 50 Hz sample, despite the lower input heat. The reason can be primarily due to the difference in the distance from the fusion boundary (penetration shape) that can form depending on a dynamic balance of heat distribution and material flow in the molten pool. The balance detected in the 250 Hz sample is believed to be one of the optimal balances, which can benefit the mechanical properties of HAZ [6]. According to the results and the simulated several YZ planes, as shown in Fig. 7, it is concluded that the 50 Hz sample has a more uniform penetration shape than the 250 Hz sample. Still, in addition to the width of the HAZ, which depends on the penetration shape, it should be noted that the thermal history and the peak temperature are also important in forming this zone. Regarding the role of heat input in the W/O vibration sample, it is well known that despite increasing the distance from the fusion boundary, the temperature peak in the W/O vibration sample is the highest in the samples in probe B. Thus, workpiece vibration with 50 and 250 Hz along the welding direction improves penetration shape and HAZ compared with the W/O vibration.

The article presents a comprehensive three-dimensional numerical model of the TP-GMAW process that is simulated successfully to investigate the weld pool formation, heat transfer, and material flow in the presence of the workpiece vibration. OM images of the W/ vibration samples with different workpiece vibration frequencies are employed to validate the simulation results. The previous study did not focus much on vibration frequency difference. Thus, the effect of the workpiece vibration frequency on the total balance of heat input and material flow in the molten pool is investigated in the present article. The following conclusions are drawn based on the results:

(1) The simulation in the W/ and W/O vibration samples makes a common feature that a shallow melting layer is created by the droplets from the leading electrode, which immediately deepens with the impact of the droplets from the trailing electrode. The W/O vibration sample with the maximum droplet temperature exhibited the deepest penetration. The toe sides of the weld pool can change shape and melt more by the heat and material flow behind this pool position, such as Marangoni flow and the longitudinal flow of material along the welding direction, depending on vibration frequency.

(2) The heat distribution in the upper molten pool begins to extend along the welding direction by applying the vibration, and this trend increases with increasing vibration frequency. This is a unique and common trend in the W/ vibration samples. Similarly, the effective heat input increased monotonically with increasing the vibration frequency. In contrast, the heat extending along the welding direction is minimum in the W/O vibration sample with the highest heat input. This inconsistency can be explained by the physical movement of material in a molten pool due to workpiece vibration.

(3) The application of vibration along the welding direction has changed the material flow from the surface to the depth in the central rear areas in the W/ vibration samples, although it follows an opposite direction in the W/O vibration sample. The two vortexes near the centre and toes inside a molten pool are affected by the balance between extending heat distribution beneath the surface and heat input, depending on vibration frequency. Consequently, the penetration curvature around the intersection of the two melting stages was minimum and the HAZ thinned in the 250 Hz sample.

(4) Tracing the material flow using a set of 3D streamlines shows the creation of a large vortex at the centre of the molten pool. It is found that applying the vibration relieves the vortex and increases internal energy concentration in the central zone. The scalable results could not be extracted for comparison in an isolated section. Plotting the streamlines in the W/ vibration samples with the dense energetic zone reveals extremely complicated lines compared with the W/O vibration sample.

(5) The probe analyses confirm that the specific vibration condition led to a narrow HAZ with a lower peak temperature during TP-GMAW than the W/O vibration specimen, regardless of the distance from the fusion boundary.

Acknowledgements: The investigation results from a joint research program between the Joining and Welding Research Institute (JWRI), Osaka University and the Indian Institute of Technology. The author, Habib Hamed Zargari, is grateful to the Indian Institute of Technology Hyderabad (IITH) for providing the facilities.

Funding: This research was funded by The Joint Usage/Research Center on Joining and Welding, Osaka University, by Project (20990018) to Create Research and Educational Hubs for Innovative Manufacturing in Asia, Osaka University, funded by the Special Budget Project of the Ministry of Education, Culture, Sports, Science and Technology, and also by JSPA KAKENHI Grant Number JP21K04677 (K. Ito).

Author contribution

Habib Hamed Zargari: writing-original draft, conceptualization, formal analysis, visualization, and data curation.

Kazuhiro Ito: project administration, review, and editing. Abhay Sharma: resources, review, and editing.

Consent for publication: All authors have read and agreed to the published version of the manuscript.

Competing interest: The authors declare no competing interest.

M.N. Ilman, R.A. Sriwijaya, M.R. Muslih, N.A. Triwibowo, Sehono, Strength and fatigue crack growth behaviours of metal inert gas AA5083-H116 welded joints under in-process vibrational treatment, J. Manuf. Process. 59 (2020) 727–738. https://doi.org/10.1016/j.jmapro.2020.10.035.
A. V. Kolubaev, O. V. Sizova, S. V. Fortuna, A. V. Vorontsov, A.N. Ivanov, E.A. Kolubaev, Weld structure of low-carbon structural steel formed by ultrasonic-assisted laser welding, J. Constr. Steel Res. 172 (2020) 106190. https://doi.org/10.1016/j.jcsr.2020.106190.
H. Ohrdes, S. Nothdurft, C. Nowroth, J. Grajczak, J. Twiefel, J. Hermsdorf, S. Kaierle, J. Wallaschek, Influence of the ultrasonic vibration amplitude on the melt pool dynamics and the weld shape of laser beam welded EN AW-6082 utilizing a new excitation system for laser beam welding, Prod. Eng. 15 (2021) 151–160. https://doi.org/10.1007/s11740-020-01008-0.
J. Wang, Y. Chen, A Pathway to Grain Structure Control of Gas Tungsten Arc Welded Duplex Stainless Steel Through Ultrasonic Vibration, Metall. Mater. Trans. A. 52 (2021) 2667–2675. https://doi.org/10.1007/s11661-021-06262-0.
Y. Ou, Q. Lu, C. Li, H. Yan, P. Zhang, J. Jin, Effect of Vibration Frequency on Microstructure and Properties of Laser-Welded Inconel 718 Nickel-Base Superalloy, J. Mater. Eng. Perform. 30 (2021) 2399–2407. https://doi.org/10.1007/s11665-020-05392-7.
H. Hamed Zargari, K. Ito, T. Miwa, P.K. Parchuri, H. Yamamoto, A. Sharma, Metallurgical Characterization of Penetration Shape Change in Workpiece Vibration-Assisted Tandem-Pulsed Gas Metal Arc Welding, Materials (Basel). 13 (2020) 3096. https://doi.org/10.3390/ma13143096.
J. Cheon, D.V. Kiran, S.J. Na, CFD based visualization of the finger shaped evolution in the gas metal arc welding process, Int. J. Heat Mass Transf. 97 (2016) 1–14. https://doi.org/10.1016/j.ijheatmasstransfer.2016.01.067.
N. Arif, H. Chung, Alternating current-gas metal arc welding for application to thick plates, J. Mater. Process. Technol. 222 (2015) 75–83. https://doi.org/10.1016/j.jmatprotec.2015.02.041.
R. Zong, J. Chen, C. Wu, A comparison of double shielded GMAW-P with conventional GMAW-P in the arc, droplet and bead formation, J. Mater. Process. Technol. 285 (2020) 116781. https://doi.org/10.1016/j.jmatprotec.2020.116781.
W. Zhang, C.H. Kim, T. DebRoy, Heat and fluid flow in complex joints during gas metal arc welding - Part II: Application to fillet welding of mild steel, J. Appl. Phys. 95 (2004) 5220–5229. https://doi.org/10.1063/1.1699486.
Z. Zhang, J. Xue, Profile Map of Weld Beads and Its Formation Mechanism in Gas Metal Arc Welding, Metals (Basel). 9 (2019) 146. https://doi.org/10.3390/met9020146.
R. Zong, J. Chen, C. Wu, A comparison of TIG-MIG hybrid welding with conventional MIG welding in the behaviors of arc, droplet and weld pool, J. Mater. Process. Technol. 270 (2019) 345–355. https://doi.org/10.1016/j.jmatprotec.2019.03.003.
L. Aucott, H. Dong, W. Mirihanage, R. Atwood, A. Kidess, S. Gao, S. Wen, J. Marsden, S. Feng, M. Tong, T. Connolley, M. Drakopoulos, C.R. Kleijn, I.M. Richardson, D.J. Browne, R.H. Mathiesen, H. V. Atkinson, Revealing internal flow behaviour in arc welding and additive manufacturing of metals, Nat. Commun. 9 (2018) 1–7. https://doi.org/10.1038/s41467-018-07900-9.
H. Hamed Zargari, K. Ito, M. Kumar, A. Sharma, Visualizing the vibration effect on the tandem-pulsed gas metal arc welding in the presence of surface tension active elements, Int. J. Heat Mass Transf. 161 (2020) 120310. https://doi.org/10.1016/j.ijheatmasstransfer.2020.120310.
Y.B. Zhong, C.S. Wu, G.K. Padhy, Effect of ultrasonic vibration on welding load, temperature and material flow in friction stir welding, J. Mater. Process. Technol. 239 (2017) 273–283. https://doi.org/10.1016/j.jmatprotec.2016.08.025.
C. Chen, S. Lin, C. Fan, C. Yang, L. Zhou, Feasibility analysis of pulsed ultrasonic for controlling the GMAW process and weld appearance, Int. J. Adv. Manuf. Technol. 97 (2018) 3619–3624. https://doi.org/10.1007/s00170-018-2202-8.
C. Zhang, C.S. Wu, S. Tian, Effect of ultrasonic vibration on current density and keyholing capability of plasma arc, Sci. Technol. Weld. Join. 25 (2020) 422–430. https://doi.org/10.1080/13621718.2020.1735062.
S. Kumar, C.S. Wu, G.K. Padhy, W. Ding, Application of ultrasonic vibrations in welding and metal processing: A status review, J. Manuf. Process. 26 (2017) 295–322. https://doi.org/10.1016/j.jmapro.2017.02.027.
Y. Lei, J. Xiong, R. Li, Effect of inter layer idle time on thermal behavior for multi-layer single-pass thin-walled parts in GMAW-based additive manufacturing, Int. J. Adv. Manuf. Technol. 96 (2018) 1355–1365. https://doi.org/10.1007/s00170-018-1699-1.
G. Xu, L. Li, J. Wang, J. Zhu, P. Li, Study of weld formation in swing arc narrow gap vertical GMA welding by numerical modeling and experiment, Int. J. Adv. Manuf. Technol. 96 (2018) 1905–1917. https://doi.org/10.1007/s00170-018-1729-z.
M.H. Cho, Y.C. Lim, D.F. Farson, Simulation of Weld Pool Dynamics in the Stationary Pulsed Gas Metal Arc Welding Process and Final Weld Shape, Suppl. To Weld. J. 85 (2006) 271–283.
M. H. CHO, D. F. FARSON, Simulation Study of a Hybrid Process for the Prevention of Weld Bead Hump Formation, Weld. J. Vol. 86 (2007) 253–262.
J. Cheon, S.-J. Na, Influence of simulation methods of temperature distribution on thermal and metallurgical characteristics in GMA welding, Mater. Des. 108 (2016) 183–194. https://doi.org/https://doi.org/10.1016/j.matdes.2016.06.090.
D. Wu, X. Hua, D. Ye, X. Ma, F. Li, Understanding of the weld pool convection in twin-wire GMAW process, Int. J. Adv. Manuf. Technol. 88 (2017) 219–227. https://doi.org/10.1007/s00170-016-8775-1.
D. Wu, S. Tashiro, Z. Wu, K. Nomura, X. Hua, M. Tanaka, Analysis of heat transfer and material flow in hybrid KPAW-GMAW process based on the novel three dimensional CFD simulation, Int. J. Heat Mass Transf. 147 (2020) 118921. https://doi.org/10.1016/j.ijheatmasstransfer.2019.118921.
W.I. Cho, S.J. Na, Impact of driving forces on molten pool in gas metal arc welding, Weld. World. 65 (2021) 1735–1747. https://doi.org/10.1007/s40194-021-01138-8.
R. Zong, J. Chen, C. Wu, D. Lou, Numerical analysis of molten metal behavior and undercut formation in high-speed GMAW, J. Mater. Process. Technol. 297 (2021) 117266. https://doi.org/https://doi.org/10.1016/j.jmatprotec.2021.117266.
L. Wang, C. Wu, J. Chen, J. Gao, Experimental measurement of fluid flow in high-speed GMAW assisted by transverse magnetic field, J. Manuf. Process. 56 (2020) 1193–1200. https://doi.org/https://doi.org/10.1016/j.jmapro.2020.06.006.
J. Chen, Z. Han, L. Wang, C. Wu, Influence of arc interactions on heat and mass transfer during a two-arc hybrid welding, Int. J. Heat Mass Transf. 148 (2020) 119058. https://doi.org/https://doi.org/10.1016/j.ijheatmasstransfer.2019.119058.
A. Ebrahimi, A. Babu, C.R. Kleijn, M.J.M. Hermans, I.M. Richardson, Y. Adonyi, J. Nadzam, The Effect of Groove Shape on Molten Metal Flow Behaviour in Gas Metal Arc Welding, Materials (Basel). 14 (2021) 1–17. https://doi.org/10.3390/ma14237444.
H. Zhang, C. Wang, S. Lin, Molten pool behaviors in double-sided pulsed gmaw of t-joint: A numerical study, Metals (Basel). 11 (2021) 1–18. https://doi.org/10.3390/met11101594.
R. Zong, J. Chen, C. Wu, A comparison of double shielded GMAW-P with conventional GMAW-P in the arc, droplet and bead formation, J. Mater. Process. Technol. 285 (2020) 116781. https://doi.org/https://doi.org/10.1016/j.jmatprotec.2020.116781.
L.L. Wang, H.L. Wei, J.X. Xue, T. DebRoy, Special features of double pulsed gas metal arc welding, J. Mater. Process. Technol. 251 (2018) 369–375. https://doi.org/https://doi.org/10.1016/j.jmatprotec.2017.08.039.
W. Ou, G.L.L. Knapp, T. Mukherjee, Y. Wei, T. DebRoy, An improved heat transfer and fluid flow model of wire-arc additive manufacturing, Int. J. Heat Mass Transf. 167 (2021) 120835. https://doi.org/10.1016/j.ijheatmasstransfer.2020.120835.
W. Zhao, Y. Wei, J. Long, J. Chen, R. Liu, W. Ou, Modeling and simulation of heat transfer, fluid flow and geometry morphology in GMAW-based wire arc additive manufacturing, Weld. World. 65 (2021) 1571–1590. https://doi.org/10.1007/s40194-021-01123-1.
X. Li, D. Bai, Y. Wang, F. Liu, H. Zhang, S. Liu, High-nitrogen steel laser-arc hybrid welding in vibration condition, Mater. Sci. Technol. 36 (2020) 434–442. https://doi.org/10.1080/02670836.2019.1706907.
A. Krajewski, Mechanical vibrations in welding processes, Weld. Int. 30 (2016) 27–32. https://doi.org/10.1080/09507116.2014.937606.
E. Ingram, O. Golan, R. Haj-Ali, N. Eliaz, The Effect of Localized Vibration during Welding on the Microstructure and Mechanical Behavior of Steel Welds, Materials (Basel). 12 (2019) 2553. https://doi.org/10.3390/ma12162553.
M. Barkhudarov, Lagrangian VOF Advection method for FLOW-3D, Flow Sci. Inc. FSI-03-TN6 (2004) 1–11.
A. Kumar, T. DebRoy, Calculation of three-dimensional electromagnetic force field during arc welding, J. Appl. Phys. 94 (2003) 1267–1277. https://doi.org/10.1063/1.1587006.
D.J. Phares, G.T. Smedley, R.C. Flagan, The wall shear stress produced by the normal impingement of a jet on a flat surface, J. Fluid Mech. 418 (2000) 351–375. https://doi.org/10.1017/S002211200000121X.
Y. Wang, H.L. Tsai, Impingement of filler droplets and weld pool dynamics during gas metal arc welding process, Int. J. Heat Mass Transf. 44 (2001) 2067–2080. https://doi.org/10.1016/S0017-9310(00)00252-0.
M.S. Min Hyun Cho, Numerical Simulation of Arc Welding Process and Its Application, The Ohio State University, 2006.
C. McIntosh, J. Chapuis, P. Mendez, Effect of Ar-CO2 Gas Blends on Droplet Temperature in GMAW :Droplet temperature can be lowered while still maintaining a similar deposition rate, Weld. J. 95 (2016) 273s-279s.
J. Hu, H.L. Tsai, Modelling of transport phenomena in 3D GMAW of thick metals with v groove, J. Phys. D. Appl. Phys. 41 (2008). https://doi.org/10.1088/0022-3727/41/6/065202.
T. Ueyama, T. Ohnawa, M. Tanaka, K. Nakata, Occurrence of arc interaction in tandem pulsed gas metal arc welding, Sci. Technol. Weld. Join. 12 (2007) 523–529. https://doi.org/10.1179/174329307x173715.
P. Sahoo, T. Debroy, M.J. McNallan, Surface tension of binary metal-surface active solute systems under conditions relevant to welding metallurgy, Metall. Trans. B. 19 (1988) 483–491. https://doi.org/10.1007/BF02657748.
M.R.U. Ahsan, M. Cheepu, R. Ashiri, T.H. Kim, C. Jeong, Y. Do Park, Mechanisms of weld pool flow and slag formation location in cold metal transfer (CMT) gas metal arc welding (GMAW), Weld. World. 61 (2017) 1275–1285. https://doi.org/10.1007/s40194-017-0489-y.

Download PDF

Journal Publication

published 16 Oct, 2023

Read the published version in The International Journal of Advanced Manufacturing Technology →

Reviewers agreed at journal
12 Feb, 2023
Reviewers invited by journal
29 Jan, 2023
Editor assigned by journal
20 Jan, 2023
First submitted to journal
19 Jan, 2023

You are reading this latest preprint version

Effect of workpiece vibration frequency on heat distribution and material flow in the molten pool in tandem-pulsed gas metal arc welding

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Experimental Setup

3. Mathematical Formulation And Assumptions

4. Results And Discussion

4.1. Emphasis on validation of the weld pool in the present study

4.2. Penetration shape change at four unique positions

4.3. Heat distribution beneath the molten pool surface varied with vibration frequency

4.4. Visualizing material flow in molten pool varied with vibration frequency

4.5. Heat conduction history in HAZ depending on vibration frequency

5. Conclusion

Declarations

References

Status:

Journal Publication

Version 1