Nội dung

Journal of Science & Technology 127 (2018) 006-010 Numerical Simulation of Temperature Field During Welding of K Type Pipe Joint Nguyen Hong Thanh1,2*, Nguyen Tien Duong2, Ha Xuan Hung3 1 Nam Dinh University of Technology Education, Loc Ha, Nam Dinh, Viet Nam Hanoi University of Science and Technology, No. 1, Dai Co Viet, Hai Ba Trung, Ha Noi, Viet Nam 3 University of Labour and Social Affairs, No. 43, Tran Duy Hung, Cau Giay, Ha Noi, Viet Nam Received: January 07, 2016; Accepted: May 25, 2018 2 Abstract This paper studies the finite element method (FEM) to simulate the temperature field during welding of K type pipe joint. Temperature variations at a point (node) in the heat source movement are examined. Using Gas Metal Arc Welding - GMAW, the welding method is now widely used in the fabrication of steel structure. The modeling and simulation of the temperature fie ld when welding K joint connection made on the basis of the model Solid (3D). In this paper we use VisualMessh / VisualWeld of ESI Group to model and simulate it. This is a useful simulation tool because it can solve the complex problems of thermomechanical and metallurgical in welding including nonlinear problems, pseudo steady state, kinetics, .. . Keywords: GMAW, Temperature field, Piping joint, K Joint, Welding numerical simulation. 1. Introduction distortion and phase transformation). The software allows modelling and simulating the welding process most sufficiently and practically. * Simulation allows shortening research time, reducing research cost and reducing the number of experiements. On that basis, the authors choose the software VisualWeld to simulate the structural carbon steel K pipe welding joint in order to study the heat transfer and the formation of residual stress and distortion during GMAW welding. Welding simulation is one of the most useful methods to predict the temparature field, residual stress and distortion using FEM. It is widely known that the temparature process directly affects the reaction of the structure in the welding process such as mechanism and microstructure. In the welding process, the welded object is partly heated at a high temparature. According to physical and chemical properties, metals and alloys expand when heated and shrink when cooled. This expansion is limited by areas with lower temperatures or with clamps, which leads to the temporary temparature stress in the welded object and the residual stress after the welded object is cooled. The temparature field identification in welding plays an important role in identifying the residual stress, welding distortion and microstructure of areas in the welding joint. 2.1 Heat source model The heat distribution in the welded object is the heat of the arc column, thermal conductivity of the base metal, the heat release into the environment and the material's thermal properties. With the arc welding heat source, the total of effective power is P=.Uh.Ih (W), in which Uh- is the arc voltage (V); Ihis the welding amperage (A) and  is the arc welding efficiency (0.6 ÷ 0.9). Goldak et al. [1] presented the heat source model in which the dobble ellipsoid distribution density was defined by the combination of two different semi-ellipsoids to make a heat source (Fig. 1). The heat density in each semi-ellipsoid is described in two separate equations. 2. Analysis and simulation tool Among current commercial welding simulation software such as ABAQUS, ANSYS, SYSWELD..., SYSWELD (VisualWeld) by ESI Group is the most powerful and sufficient specialized software in simulating welding and metal temparature treatment processes. The software can solve complex problems related to nonlinear analysis (in heat transfer, * Corresponding author: Tel: (+84) 942283232, E-mail: thanhspktnd.we@gmail.com Fig. 1. GMAW welding heat source model, [2] 6 Journal of Science & Technology 127 (2018) 006-010 With any point (x,y,z) inside the first semi-ellipsoid (in front of the welding arc), the heat source density is presented (Eq. 1). Table 2. Parameters of K joint, [5] Parameter Description Unit Value D The outer diameter of the main pipe mm 219 t The thickness of the main pipe mm 12.7 Dbtens, The diameter of the branch pipe mm 102 tb The thickness of the branch pipe mm 8 where Qf and Qr represent the maximum power density of the front and rear part of the model respectively and af , ar, b and c represent parameters which enable changing the model geometry and adjusting the shape of the heat source model to results obtained in real welding tests, see Fig. 1  Branch angle o 50 g gap mm 50 D/t Chord slenderness - 17.24 Goldak et al. [3] showed the relationship between the size of the heat source and the size of the welding pool, and supposed that the appropriate values for af, ar, b and c can be achieved by directly measure the geometric parameters of the welding pool. (Table 1). Db/tb Branch slender-ness - 17 Db/D Width ratio - 0.47 e Joint eccentricity mm 0  x2 y 2 z 2  QR ( x, y, z , t ) = Q f . exp  − 2 − 2 − 2   a b c  f  (1) With any point (x,y,z) inside the second semiellipsoid (behind the welding arc), the heat source density is presented (Eq. 2).  x2 y 2 z 2  QR ( x, y, z, t ) = Qr . exp  − 2 − 2 − 2  c   ar b Dbcomp (2) The size of K joint is designed according to the standard of AISC (Fig. 2). The parameters are described in Table 2. Table 1. Parameters of moving heat source, [4] Layers/Pass Layer 1 (1, 2) Layer 2 (3, 4) Layer 2 (5, 6) b(mm) 7 6 6 c(mm) 4 4 4 af mm) 3 3 3 wall ar (mm) 5 5 5 2.3 Discrete model In order to properly simulate, it is necessary to describe on FEM model exactly the same as factual welding process, including: welding line trajectory, reference line, start point and end point of the welds. The branch pipe thickness is 8 mm, bevelled edge, we weld 2 layers with 3 passes, see Fig. 3. 2.2 Material parameters The material used in K type pipe joint is the structural carbon steel and the material used in simulation is S355J2G3 steel, equivalent to the material used to manufacture K type pipe joint. The chemical composition of S355J2G3 steel are: C (0.18%), Mn (1.6%), Si (0.55%), S (0.035%), P (0.035%) and it has the melting temparature of 1500 o C. S355J2G3 steel has the yield strength of 355 MPa, elastic modulus E = 210 GPa (at 20 oC), poisson coefficient of 0.33. In the welding process simulation, the authors study the temparature variations from 20 oC to the melting status of the material (1500 oC). Fig. 3. Welding layers arrangements The K type pipe joint is discreted model with 145126 elements and 115495 nodes. There are 94288 elements that using in K type pipe joint are hexa, penta, tetra (Fig. 4). In order to increase the accuracy of analyses, the mesh was consentrated in the joint area as well as in the area adjacent to it (HAZ); in other areas, the distance between nodes is sparser to decrease the calculating time and the number of elements and nodes of the model (Fig. 5). Fig. 2. K type pipe welding joint 7 Journal of Science & Technology 127 (2018) 006-010 In which: Vh: Welding speed [mm/s]; Dd: The diameter of wire [mm]. Table 4. Simulation parameters for GMAW. Layers/ Passes Layer 1 (1, 2) Layer 2 (3, 4) Layer 2 (5, 6) Fig. 4. K type pipe joint modelling Linear (J/mm) 520 610 570 energy Welding speed (mm/s) 5 5 5 Fig. 7. The heat source calibration In Fig. 7, the welding mode in Table 4 shows that the fusion depth between simulation and experiment is almost the same and acceptable. Therefore, we can use this heat mode to simulate the temparature of the welding process. Fig. 5. Type of mesh and mesh dividing area 2.4 Welding process With K type pipe joint, the welding process and welding layers are arranged as in Fig. 3. To finish this welding joint, we have 12 passes (half a pipe). 2.6 The clamp condition The K type pipe joint is clamped as in Fig. 8, in which the main pipe is tightly clamped in three directions (x,y,z) and the two branch pipes are only clamped in the direction z. The clamping time remains during the simulation (4500 s). Fig. 6. Welding sequences In this paper, the author choose the start point of welding line in the narrow angle and the end point in the wide angle. The welding order is respectively 1,2,3,4 as in Fig. 6. The latter welding layers (passes) apply the same as above. The start point and end point of welding passes are alternate. Fig. 8. Clamping position in K type pipe joint 2.7 Calculation and simulation To get the accurate simulation results, we have to describe parameters exactly the same as the factual welding process, including: welding line trajectory, reference line, start point and end point of the welds (Fig. 9). This process is undergone on the software VisualWeld [6]. 2.5 The heat source model modification This is an important step because according to the cross section of the welding joint compared with the temparature field when simulation, we can modify some welding parameters (linear energy, torch angle, parameters af, ar, b and c) suitably to the fusion depth and the width of HAZ area of the weld beads. Table 3. Welding parameters for GMAW. Layers/ Pass Layer 1 (1, 2) Layer 2 (3, 4) Layer 2 (5, 6) Ih (A) 120 130 125 Uh (V) 22 24 23 Vh (mm/s) 5 5 5 Dd (mm) 1,0 1,0 1,0 Fig. 9. Weld operation description 8 Journal of Science & Technology 127 (2018) 006-010 3. Simulation results and discussion After modelling the welding joint [7], we declare the properties of materials, the heat source power and set the calculating conditions as described above. Solving the problem, we get the following results: 3.1 The temparature field When the welding heat source moves along the weld, there appears a temporary temparature distribution field and heat stress in the joint. In Fig. 10 is the temparature distribution results of the welding pool and HAZ area when welding the first line. Fig. 11 is the temparature distribution results of the welding pool and HAZ area when welding the first and the second line. Fig. 13. The quantity of heat to the weld 3.2 The heat process Fig. 14. The total of heat in three welds Fig. 14 is the location of the nodes defining the temperature field. Fig. 10. The total of heat in the first weld Fig. 11. The total of heat in the first and second weld Fig. 15. The heat process at node 341 Fig. 15 is the heat distribution at node 341 in the 1 of the K type pipe welding joint with the useful heat power and the respective welding speed. At 33.1946 seconds, the temperature is 2125.14 oC, this is the time that the melting welding pool pass. After about 317 seconds, the temperature of the joint decrease to 38 oC, at that time, the second bead will be welded. Fig. 12. The total of heat in three welds Fig. 12 is the temparature distribution results of the welding pool and HAZ area when welding three lines with the heat parameters as in Table 4. Because the heat transfers to the main pipe in 2 dimensions and to the branch pipes in 1 dimension, the heat affected area in the main pipe is narrower than in the branch pipes. At the second number 1435.65, the temperature at this node is 685.3 oC; the reason for this increase is affected by the second welding line. However, the second welding line is below this node; therefore, at this time node 341 is affected by the heat as in HAZ area. Fig. 13 is the process of heat transfer to the weld, which shows that the heat parameters used in the simulation is appropriate; the joint is not damaged, the heat energy at the beginning and finish of the weld is well treated. It can be seen in Fig. 15 that with the chosen heat source power, the welding joint is not damaged. 9 Journal of Science & Technology 127 (2018) 006-010 field in the K type pipe welding joint with 2 layers and 3 passes using GMAW process. The first pass is the innermost layer, so besides being heat affected itself, it is directly affected by the heat transfer when welding the second and third lines. Therefore, the mechanism and microstructure of the weld happens complicatedly, directly affecting the welding quality. By changing the input parameters such as the effective heat source power P (or the linear energy) and the welding speed Vh using the software Sysweld/ VisualWeld, we can find out the appropriate welding parameter set in order to provide the weld the sufficient quantity of heat to melt the steel but not damage the steel. Fig. 16. The heat process at node 5034 According to the temparature field simulation in K type pipe welding joint, we can know: + The transfer of the melting welding pool of each welding line and the whole process. + HAZ area when welding each line and the whole process. + The heat distribution in the chord pipe and the branch pipe. Fig. 17. The heat process at node 48870 Numerical simulation is a useful and reliable tool in calculating and analyzing the temperature variation at any time in the welding joint. It can be predicted that the possibility to form the weld and detects can appear during the welding. Simulation gives rapid results and optimizes the welding parameters to control parameters directly affecting the quality of the weld. We can change the welding parameters rapidly without experiments. Fig. 18. The heat process along the chord pipe References Fig. 16 is the temparature field at node 5034, this is the node at HAZ area of the branch pipe. This node is heat affected by three welding lines. However, because of the far distance with the second weld bead, the temparature only increases to 300 oC. This temparature does not affect the mechanism and the microstructure of the welding joint and the HAZ area. [1]. Nguyễn Tiến Dương, Mô phỏng quá trình truyền nhiệt khi hàn, Hà Nội (2008). [2]. Nguyễn Hồng Thanh, Hà Xuân Hùng, Nguyễn Tiến Dương, Mô phỏng số trường nhiệt độ khi hàn liên kết chữ T thép A36 bằng phương pháp hàn GMAW, Tạp chí cơ khí Việt Nam, số 6 (2014). [3]. Farid Vakili-Tahami. Finite Element Analysis of the Inservice-Welding of T Joint Pipe Connections. [4]. Eager, T. W., Tsai N. S., Temperature fields produced by traveling distributed heat sources, Welding Journal, 62(12). (1983) 346-355. [5]. ESI Group, 99 Rue Des, Solets Silic 112 94513 Rungis Cedex FRANCE [6]. Goldak, J., Chakravarti, A., Bibby, M., A new finite model for welding heat sources, Metallurgical Transaction, 15B (1984) 299-305. [7]. Goldak, J., and Akhlaghi, M. (2005) Computational Welding mechanics, SpringerScience Bussiness Media,Inc. [8]. Jeffrey Packer, Steel Design Guide (2010) 96-97. Fig. 17 is the heat process at node 48870 in HAZ area and the main pipe is 15 mm away from the weld bead. When the second bead is welded, this node is heat affected (125 oC). This temparature does not affect the microstructure and mechanism in this area. Fig. 18 shows the temparature variations along the main pipe at step 435 with the highest temparature about 150 oC. 4. Conclusion This paper focuses to simulate the temparature 10