Available online at www.sciencedirect.

com

Procedia Engineering 36 (2012) 253 260

IUMRS-ICA 2011

Mesh Reconstruction Technology of Welding Process in 3D FEM Simulation of Porthole Extrusion and Its Application
Zhi-hao Zhanga,b, Wen-rong Houa, Dong-nan Huangc, Jian-xin Xiea,b,
a b

Institute for Advanced Materials and Technology, University of Science and Technology Beijing, Beijing 100083, PR China State Key Lab for Advanced Metals and Materials, University of Science and Technology Beijing, Beijing 100083, PR China c School of Materials Science and Engineering, Inner Mongolia University of Technology, Hob. hot 010051, PR China

Abstract For extrusion of hollow profile with asymmetric cross-section, existing FEM software cannot simulate the influence of divergent and welding on the metal flowing behaviors, because it is inability to deal with meshes separation and penetration during welding. A mesh-reconstruction technology based on the reverse engineering is proposed in this paper, and extrusions of Al-alloy hollow profiles with various cross-sections are simulated. The results show that the mesh-reconstruction technology can solve the problem of simulation termination caused by the meshes separation and penetration during porthole die extrusion. It is effective to analyze the metal flowing behaviors, temperature distribution, hydrostatic stress in weld-chamber, and stress on dies, which provides a practicable way for die structure design and optimization of porthole extrusion as well as quality forecast of extrusion products.

2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of MRS-Taiwan
Keywords: Porthole extrusion; mesh reconstruction; FEM; Al-alloy

1. Introduction During porthole extrusion of Al-alloy hollow profile, as the metal deforming and flowing in closed container, die portholes and weld-chamber, it is difficulty to obtain correct and comprehensive metal flow behaviors by the physical modeling methods (such as visual plasticity, etc.). Numerical simulation

* Corresponding author. Tel.: +86 1062332254; Fax: +86 1062332253. E-mail address: jxxie@mater.ustb.edu.cn.

1877-7058 2012 Published by Elsevier Ltd. doi:10.1016/j.proeng.2012.03.038

254

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

become an important method for analyzing metal flow, designing and optimizing die structure, forecasting temperature distribution and quality of product [1,2]. However, for extrusion of hollow profile with cross-section that welding locations are inconsistent with geometry symmetry axis, the existing 3D FEM software can only simulate the diversion or steady extrusion process [3-5]. When simulating the welding process, mesh penetration or separation usually leads to the termination of the calculation. So, it is inability to analyze the effects of the metal flow behaviors on the quality of extrusion product. Finite volume method can simulated the welding process with no need of re-meshing. But the metals on two sides of the rigid weld-plane are not welded together in fact, thus some extrusion defects such as bending and twisting caused by uneven flow velocity in both sides of weld-plane can not be forecasted [6]. For solving the mesh penetration or separation problem during the simulation of porthole extrusion, a mesh-reconstruction technology based on the reverse engineering is proposed in this paper, and applied in the simulations of the extrusions of Al-alloy hollow profiles with various cross-sections. The metal flowing features and its effect on the homogeneity of metal flow are analyzed. The temperature distribution, mandrel stress, and quality of extrusion product are forecasted. 2. Mesh reconstruction technology Fig. 1 shows the model of a porthole extrusion for tube when meshes began to penetrate. The extrusion direction is along the positive y-axis in the graph.
z x yExtrusion
Bdirection
Symmetry plane Side surface of mandrel Bottom of porthole bridge

(a)

(b)
Side surface of weld-chamber Outline of mandrel

(c)

Fig. 1 Mesh penetration of a porthole extrusion: (a) three-dimensional view; (b) the view parallel to the extrusion direction; (c) the view along the negative z-axis.

The meshes penetration degree increases with the extrusion stroke, as shown in Fig. 2. In Fig. 2 (b), the hatched part (bdfe area) is the penetration region, non-hatched parts (abc and fgh areas) is the unfilled region in weld-chamber. To make sure the billet volume is same before and after the mesh-reconstruction, reconstruction should be performed when the volume of the penetration region is equal to the unfilled region. When reconstructing, the volume of penetration region and unfilled region are approximated as the multiply of the weld-chamber height with the hatched part and non-hatched parts showed in Fig. 2 (b). As the arc curvature and radius of the hatched part and non-hatched part (with boundary of solid line in Fig. 2 (b)) are hardly to measured, and the calculation of the area is complicate, the area of hatched part can be simplified as the areas of bde and fed, while the area of non-hatched part can be simplified as the areas of abc and fgh (with boundary of dotted line in Fig. 2 (b)).The length of ac, de, gh, and the height of kb, bo, fo, fm is easily measured in Deform-3D software. When acbk + ghfm is equal to

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

255

debo + defo, the area of bde +fed andabc +fgh is the same. Then, mesh reconstruction is performed.
Symmetry plane

a
Side surface of mandrel

k c b d o e f (b) h

(a)
Penetration meshes Side surface of weld-chamber

Fig. 2 Simplified model of mesh penetration: (a) the finite element mode; (b) simplified geometry diagram.

Reconstruction procedure is as follows: (1) Transform the penetrated meshes model into a 3-D model with small triangle as the basic unit, which is STL (Stereolithography) model. (2) Remove the penetrated and distortion triangles by small plane feature technology in Pro / ENGINEER, and select three adjacent vertices to reconstruct a triangle, making the original penetration region reform a surface constituted by triangles. Meanwhile, fill the surface of the unfilled region with triangle units. Then, a 3D STL model with triangle units of the original penetration region and unfilled region are re-obtained. (3) Add element node data to the mesh reconstructed model in Deform-3D and calculate again, the welding and forming simulation of porthole die extrusion will continue. 3. Feasibility verifying For verifying the feasibility and accuracy, taking the extrusion process of square tube as an example, the mesh reconstruction technology and the method assuming the weld-plane as rigid plane are adopted respectively. Weld-plane location of the square tube and the extrusion die structure are shown in Fig. 3. The diameter of the extrusion container is 95 mm, extrusion ratio 31.6, and diversion ratio 12.6. Set the initial temperature of the billet (A6005 Al-alloy) as 480, container 400, die (H13) 450, dummy block 30, and shaft speed of 4 mm/s. Shear friction model with friction factor m=1 between billet and die is selected.
Porthole Weld-plane

Top die Weld-chamber

(a)

Bottom die

(b)

Fig. 3 Extrusion profile and die model: (a) cross-section; (b) die model.

256

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

Assuming the weld--planes are same as the diagonals of the square tube, as shown in Fig. 3(b). When taking the 1/4 model (hatched part in Fig. 3(a)) for simulation, it is necessary to use mesh reconstruction technology because the weld-plane is included in the model. The geometric model and meshes are shown in Fig. 4(a). When taking the 1/8 model (half of the hatched part in Fig. 3(a)) for simulation, it is unnecessary to use mesh reconstruction technology because the weld-plane can be simplified as rigid plane. The geometric model and meshes are shown in Fig. 4 (b).

Dummy block

Container

Billet Top die Weld-chamber Mandrel Bottom die

(a)
Fig. 4 Geometry model and meshes: (a) 1/4 model; (b) 1/8 model.

(b)

The welding process of the square tube extrusion is shown in Fig. 5. The simulation results of 1/4 and 1/8 model are almost same in the diversion process, filling weld-chamber process and the weld-plane contacting process, as shown in Fig. 5(a) and (b).

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 5 Welding process of the square tube extrusion: (a) model and meshes; (b) weld-plane begin to contact (1/4 and1/8 model, stroke 30.90mm); (c) weld-plane before reconstructing (1/4 model, stroke 30.95mm); (d) weld-plane contact increase (1/8 model, stroke 30.95mm); (e) after reconstructing (1/4 model, stroke 30.95mm); (f) welded completely (1/8 model, stroke 31.05mm).

For the 1/4 model simulation, when the extrusion stroke is 30.95mm, the area of penetration region and unfilled region are the same based on reconstruction rules, as shown in Fig. 5(d). The reconstructed weld-plane meshes are shown in Fig. 5(f). For the 1/8 model simulation, when the extrusion stroke is 30.95mm, the welding condition is shown in Fig. 5(d). When the stroke is 31.05mm, welding is completed as shown in Fig. 5(f). Figure 5 indicated that when suppose weld-plane as rigid, the extrusion stroke was only 0.15mm from the beginning of the weld-planes contact to the welding process finished. That is to say, when using the reconstruction method, metal flow behavior only slightly changed in 0.15mm extrusion stroke compared

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

257

with the rigid plane method, which can hardly affect the metal flow simulation results of the whole process of extrusion. 4. Application 4.1. Analyzing metal flow behavior and optimizing die structure of porthole extrusion Adopting the mesh reconstruction technology, the metal flow behaviors during extrusion of simple and complicated cross-sections of hollow profiles are obtained, as shown in Fig. 6 and Fig. 7. In diversion process, the billets are divided into several streams and flow into the portholes. With increasing the extrusion stroke, the streams of the metal contact with the bottom of the weld-chamber successively. Surrounding the mandrels, the streams flow radially and fill the welding chamber. Finally, the metal streams are welded and form the extrusion profile. In the welding process, the eight streams of the metal contacted with the bottom of the welding chamber successively.
(a)

(b)

(c)

(d)

(e)

Fig. 6 Metal flowing behaviors of simple cross-section profile: (a) Cross-section; (b) Diversion; (c) Filling and welding; (d) Forming; (e) Steady extrusion

(a)
Deviation of weld-plane

Q4 Q5

Q3 Q6

Q2 Q7

Q1 Q8

(b)

(c)

(d)

(e)

Fig. 7 Metal flowing behaviors of complicate cross-section profile: (a) Cross-section of a profile; (b) Diversion process; (c) Filling process; (d) Welding process; (e) Forming process [7].

Figure 7 indicates that, the welding of the streams flow from Q1 and Q8, Q3 and Q4, Q4 and Q5 is lagging behind other streams. Meanwhile, because a part of metal is extruded out of the die orifice before welding completely and the velocities of the central streams are higher, which lead to the weld-plane of Q3 and Q4 deviating from the symmetry plane of the porthole bridge(as shown in Fig. 6(b)). In the

258

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

forming process, the asymmetry flowing of the metal process is continued, which leads to the wavelike defect near the central region of the cross-section. The metal flow behaviors indicate that, the wavelike defect is mainly caused by the difference of the velocities of the metal streams in the diversion and weld process. So the die structure can be optimized by decreasing the velocities of the streams in the four central portholes, and increasing the velocities of the steams in the four side portholes. The purpose of that is ensured the weld-planes welding simultaneously. After optimizing of the die structure, the flowing velocities and volumes of the metal streams are matching reasonably. The wavelike defect of the profile is avoided. The shapes of the extruded section before and after optimization are shown in Fig. 8.

(a)

(b)

Fig. 8 The shapes of the extruded section: (a) Shape forming process (before optimization); (b) Shape forming process (after optimization) [7].

In order to design and optimize the porthole die reasonably, it is important to determine the metal flow behaviors as well as the influence of die structure and dimensions on it. For the extrusion of hollow profile with cross-section that welding locations are inconsistent with geometry symmetry axis, the mesh reconstruction technology of the paper is effective to obtain and analyze the metal flow behaviors during diversion and weld processes, which is impossible when using the existing 3D FEM software. 4.2. The temperature distribution during porthole extrusion Applying the mesh reconstruction technology, the temperature distribution of the billet and profile in the steady extrusion process of the profile shown as in Fig. 6(a) is shown in Fig. 9. It indicates that, the temperature increases gradually from the billet to extruded profile along the extrusion direction. Because of the thermal conduction with dummy block and without deformation, the temperature of the local of the billet around contour A is lower. Around the die orifice, the temperature is high as 525 because of the severe plastic deformation.
A B C D E E E E F F Temperature A=400 B=425 C=450 D=475 E=500 F=525

Fig. 9 Temperature distribution.

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

259

Determine the temperature distribution is significant for evaluate the structure and mechanical properties of extrusion products. Meanwhile, it has special importance for iso-thermal extrusion controlling. Only the temperature distribution in steady extrusion process (the effects of diversion and welding process on the temperature being neglected) can be obtained when using existing FEM software. So, the mesh reconstruction technology provides an effective way for forecasting and controlling the temperature of complicated hollow extrusion products. 4.3. Forecasting of welding quality and size accuracy of hollow profile During porthole extrusion, the welding quality of the hollow profile mainly depends on the magnitude of hydrostatic stress in weld-chamber. The welding quality increases with increasing the hydrostatic stress. The hydrostatic stress in weld-chamber of an irregular cross-section A6005 profile extrusion is shown in Fig. 10. It indicates that the minimal hydrostatic stress is 212MPa. It is about 5 times than the yield stress (45MPa) of A6005 at present temperature (500), which is enough for meeting the demand of welding.

(a)

Side surface of weld-chamber E D C B A

Mandrel

Hydrostatic stress A=212MPa B=242 MPa C=272MPa D=3022Mpa E=332Mpa F=362Mpa

(b)

Fig. 10 Hydrostatic stress distribution in welding chamber: (a) Cross section of a extrusion profile, (b) Hydrostatic stress.

The elastic offset of mandrel is one of the main factors leading to wall thickness deviation of hollow extrusion profile. The elastic offset of the mandrel of the profile in Fig. 10(a) can be calculated, as shown in Fig. 11 (the arrow direction is the elastic deformation direction of the mandrels). The offset is 0.083mm, which means the wall thickness deviation is about 0.08mm. Using existing FEM software, it is impossible to evaluate the die stress at weld process of complicated hollow profile extrusion. The mesh reconstruction technology is effective to simulate the hydrostatic stress, die stress, mandrel offset, etc., and which is beneficial to forecast welding quality and size deviation as well as verify the die strength.

260

Zhi-hao Zhang et al. / Procedia Engineering 36 (2012) 253 260

B A A

A B B C DE

Mandrel offset A=0.018 mm B=0.033 mm C=0.050 mm D=0.067 mm E=0.083 mm

B A

Fig. 11 Elastic offset of the mandrel.

5. Conclusions A mesh-reconstruction technology based on the reverse engineering is proposed. It is effective to analyze the metal flowing behaviors, temperature distribution, hydrostatic stress in weld-chamber, and stress on dies, which is impossible when using the existing 3D FEM software. It provides a practicable way for die structure design and optimization of porthole extrusion as well as quality forecast of extrusion products.

Acknowledgements This work was supported by State Key Lab of Advanced Metals Materials and the research project of Guangdong Province, PR China.

References
[1] Xie JX, Murakami T, Ikeda K. Experimental simulation of metal flow in porthole-die extrusion. J Mater Process Technol 1995; 49:1-11. [2] Jo HH, Lee SK, Jung SC. A non-steady state FE analysis of Al tubes hot extrusion by a porthole die. J Mater Process Technol 2006; 173:223-31. [3] Liu G, Zhou J, Duszczyk J. FE analysis of metal flow and weld seam formation in a porthole die during the extrusion of a magnesium alloy into a square tube and the effect of ram speed on weld strength. J Mater Process Technol 2008; 200:85-98 . [4] Jung ML, Byung MK, Chung GK. Effects of chamber shapes of porthole die on elastic deformationand extrusion process in condenser tube extrusion. Mater Des 2005; 26:327-36. [5] Li L, Zhang H, Zhou J. Numerical and experimental study on the extrusion through a porthole die to produce a hollow magnesium profile with longitudinal weld seams. Mater Des 2007; 5:2-9. [6] Yang DY, Kim KJ. Design of processes and products through simulation of three-dimensional extrusion. J Mater Process Technol 2007; 191:2-6. [7] Huang DN, Zhang ZH, Li JY, Xie JX. FEM analysis of metal flowing behaviors in porthole die extrusion based on the mesh reconstruction technology of the welding process. Int J Miner Metall Mater 2010; 17:763-9.