NUMERICAL SIMULATION OF HEAT TRANSFER AND FLUID FLOW IN DOUBLE ELECTRODES TIG ARC-WELD POOL
WANG Xinxin1, FAN Ding1,2(), HUANG Jiankang1,2, HUANG Yong1,2
1 School of Materials Science and Engineering, Lanzhou University of Technology, Lanzhou 730050 2 State Key Laboratory of Advanced Processing and Recycling of Nonferrous Metals, Lanzhou University of Technology, Lanzhou 730050
Cite this article:
WANG Xinxin, FAN Ding, HUANG Jiankang, HUANG Yong. NUMERICAL SIMULATION OF HEAT TRANSFER AND FLUID FLOW IN DOUBLE ELECTRODES TIG ARC-WELD POOL. Acta Metall Sin, 2015, 51(2): 178-190.
Based on a developed unified three dimension (3D) mathematical model including two tungsten electrodes arc and weld pool for double electrode TIG arc heat source, the temperature, velocity, current density, magnetic flux and Lorentz force of the double electrodes TIG arc and the weld pool are obtained for SUS304 stainless steel. The simulated results are in fair agreement with the experimental results available. Buoyance, Lorentz force, plasma drag force, Marangoni shear stress and turbulent effect are taken into account to formulate the weld pool behavior and the effects of the each force on the flow of the weld pool are studied respectively. The heat flux and shear stress at the weld pool surface are analyzed as well. A dimensionless number Pe is used to compare the relative importance of convective heat and conductive heat in the weld pool. It is shown that non-axisymmetric double electrode arc results in the non-axisymmetric characteristics of the current density, heat flux, plasma drag force and Marangoni shear at the weld pool, and thus produces non-axisymmetric weld pool profiles. The evolution of the weld pool has little effect on the arc behavior. The plasma drag force of the double tungsten electrode TIG arc decreases significantly compared with that of the TIG arc. The Marangoni stress determines the weld pool flow and the heat convected dominates the heat transfer in the weld pool, their combination effect determines the heat transfer in the weld pool, which is the essential reason for the formation of the different weld pool profiles。
Fig.1 Schematic of computation domain and boundary conditions
Area
v / (m·s-1)
T / K
?? / V
A / (Wb·m-1)
A
T=T(x, y)
B
A=0
C
-
Eq.(18)
A=0
D
-
Eq.(18)
0
E
-
1800
Table 1 Boundary conditions
Parameter
Value
Unit
Radiative emissivity er
0.4
Boltzmann constant kB
1.381×10-23
J·K-1
elementary charge e
1.6×10-19
C
Stefan-Boltzmann constant s
5.67×10-8
W·m-2·K-4
Permeability m0
4p×10-7
H·m-1
Latent heat L
2.47×105
J·kg-1
Ambient temperature T∞
300
K
Solidus temperature Ts
1673
K
Liquidus temperature Tl
1723
K
Density r
7200
kg·m-3
Thermal expansion coefficient β
10-4
K-1
Convection heat transfer coefficient hc
80
W·m-2·K-1
Table 2 Physical parameters used in the model[19,25,43]
Fig.2 Temperature and flow fields in xz section (a, c) and yz section (b, d) of arc plasma and weld pool with pure Ar shielding (a, b) and Ar+O2 shielding (c, d) after spot welding for 2 s (T—temperature; F—voltage drop; Tm and vm—maximum temperature and velocity, respectively)
Fig.3 Distributions of current density (a) and Lorentz force (b) of the arc and weld pool in xz section
Fig.4 Temperature and flow fields (a) and magnetic flux (b) at a location about 0.15 mm above the anode
Fig.5 Temperature fields at the anode surface (xy section) with pure Ar shielding (a) and Ar+O2 shielding (b) after spot welding for 2 s (Isotherms are in the unit of K)
Fig.6 Temperature distributions in x and y directions at the anode surface (xy section) with pure Ar shielding (a) and Ar+O2 shielding (b) after spot welding for 2 s (Wx and Wy —weld widths in x direction and y direction; d— position in x or y directions )
Fig.7 Current density and heat flux distributions in x direction (a) and y direction (b) at the anode surface (qe and qc are heat flux to the anode due to the electron absorption and the heat conduction, respectively; qa is the total heat flux to the anode; jz is the component of the current density in z direction)
Fig.8 Shear stresses in x direction (a, c) and y direction (b, d) at the weld pool surface with pure Ar shielding (a, b) and Ar+O2 shielding (c, d) about 2 s later after arc ignition (t—shear stress, tp—plasma drag force, tM—Marangoni shear stress, tp+tM—total shear stress)
Fig.9 Weld pool flows at xz sections (a, c, e, g, i) and yz sections (b, d, f, h, j) driven by buoyance (a, b), Lorentz force (c, d), surface tension ∂g /∂T<0 (e, f), surface tension ∂g /∂T>0 (g, h) and plasma drag force (i, j)
Fig.10 Weld pool geometry (a) and maximum velocity driven (b) by various forces (vm—maximum velocity in the weld pool, D—weld depth, W—weld width, Wx and Wy —weld widths in x and y directions)
Fig.11 Variations of weld pool geometry (a) and maximas of heat flux and current density at the anode (b) with time (t)
Fig.12 Photographs of double electrodes TIG arc in xz section (a) and yz section (b)
Fig.13 Comparison between simulated and experimental weld pool profiles at xz section (a, c) and yz section (b, d) for pure Ar shielding (a, b) and Ar+O2 shielding (c, d)
Fig.14 Comparison between simulated and experimental weld pool geometries
[1]
Kobayashi K, Nishimura Y, Lijima T, Ushio M, Tanaka M, Shimamura J, Ueno Y, Yamashita M. Weld World, 2004; 48(7/8): 35
[2]
Fan D, Lin T, Huang Y. Trans China Weld Inst, 2008; 29(12): 1
(樊丁, 林涛, 黄勇. 焊接学报, 2008; 29(12): 1
[3]
Wang X X, Huang Y, Fan D, Yang L, Yan L Q. J Lanzhou Univ Technol, 2013; 39(1): 14
(王新鑫, 黄勇, 樊丁, 杨磊, 晏丽琴. 兰州理工大学学报, 2013; 39(1): 14)
[4]
Leng X, Zhang G, Wu L. J Phys, 2006; 39D: 1120
[5]
Huang Y, Qu H Y, Fan D, Liu R L, Kang Z X, Wang X X. Trans China Weld Inst, 2013; 34(3): 33
