• No results found

A Feasibility Study on the Three Dimensional Friction Stir Welding of Aluminum 5083 O Thin Plate

N/A
N/A
Protected

Academic year: 2020

Share "A Feasibility Study on the Three Dimensional Friction Stir Welding of Aluminum 5083 O Thin Plate"

Copied!
7
0
0

Loading.... (view fulltext now)

Full text

(1)

A Feasibility Study on the Three-Dimensional Friction Stir Welding of Aluminum

5083-O Thin Plate

Young-Gon Kim

1

, In-Ju Kim

1

, Young-Pyo Kim

2

and Sung-Min Joo

3,* 1Green Manufacturing Process R&D Group, KITECH, Gwangju 500–460, Korea

2Hwacheon Machine Tool Co., Ltd, Gwangju 506–733, Korea

3Dept. of welding & joining science engineering, Chosun University, Gwangju 501–759, Korea

Aluminum ships are increasingly being built these days according to the international marine environment contamination regulations. However, aluminum alloys are associated with many problems such as welding deformation. A basic study was conducted to address the prob-lems using friction stir welding (FSW), a solid-state joining method. The AA 5083-O specimen was 2 mm-thick, and a device that provided five-axis processing control was used for the curvature welding to determine the feasibility of the process. X-ray was used to test the FSW joints quality and to examine inner defects. The characteristics of the FSW joints were evaluated based on the deformation after the welding, tensile strength, and Vickers hardness of each test specimen, according to the welding speed. The welding results were visually satisfactory for the approximate joint length of 200 mm. The joints strength was almost constant regardless of the welding speed. The joint efficiency was about 90% of that of the base material, while the elongation of the joints was about 50% of that of the base material. The hardness distribution of the joints was almost the same as that of the base material regardless of the heat input, which corresponds to the tensile test results.

[doi:10.2320/matertrans.M2015475]

(Received December 24, 2015; Accepted March 23, 2016; Published April 28, 2016)

Keywords: three-dimensional friction stir welding (FSW), curvature welding, five-axis control, deformation, joint efficiency

1.  Introduction

Due to the recent international marine environment con-tamination regulations, more aluminum ships are being built these days than fiber-reinforced plastic (FRP) ships. Alumi-num ships are over 50% lighter than normal steel ships, have a high specific strength, and can be recycled when they be-come obsolete. Thus, studies are being conducted on alumi-num shipbuilding methods.1) The most important

technolo-gies in aluminum shipbuilding are cutting, curvature forming, and welding. In terms of fusion welding, the aluminum alloy has many problems, including welding defects such as poros-ities and hot cracks caused by a low melting point and a high thermal conductivity, aside from welding deformation and spatters during the welding.2–4)

Accordingly, friction stir welding (FSW) is drawing atten-tion as a method that addresses the problems associated with aluminum welding. Developed by TWI in the UK in 1991, the solid-state welding uses the mutual friction heat between the tool and the material. It is an eco-friendly welding tech-nology that hardly involves coarsening or welding deforma-tion, unlike in the existing fusion welding technologies, such as arc welding. It also has no hazardous factors such as fumes or spatters.5) Since its development, it has been practically

applied in diverse industries such as in railway vehicles, cars, ships, and civil engineering structures, and its application range is widening.6–8) In light of this situation, the

develop-ment of the FSW equipdevelop-ment technology is expected to facili-tate applications that have not been possible due to the lack of hardness of the welding tools and the lack of FSW joints shaping control techniques. It is particularly important to choose the right material and shape of the welding tool be-cause of the FSW mechanism. To weld Al and Mg alloys, which are the representative non-ferrous alloys, SKD (Steel

Kogu Dice) alloy tool steel and a probe as thick as or slightly thinner than the material to be processed are generally chosen for the tool. To ensure the quality of the FSW joints, not only the use of a proper tool but also the optimization of the weld-ing process parameters are required. The weldweld-ing process pa-rameters basically include the rotation speed of the tool and welding speed, which have the greatest influence on the heat input. The tilting angle and inserting depth should also be considered in detail because they influence the metal flow of the material.

The recent studies on the FSW of aluminum alloy were mostly on the metal flow of the heterogeneous material9) and

its microstructure change,10) and for the materials, on the butt

joint and lap joint for plates with AA 5xxx11,12) and 6xxx.13)

There are a few studies on the FSW joint for the 3D curvature shape.14,15)

In this study, 3D curvature FSW was performed on the AA 5083-O, which is recently drawing attention as the best mate-rial for aluminum ships among non-heated aluminum alloys because of its relatively good mechanical and corrosion char-acteristics. The main purpose of 3D curvature FSW is to con-centrate on the core technology using an aluminum sheet for small ships such as a yacht. Its application can improve the working accuracy and reduce production time for aluminum ships. It is also possible to develop the automatic welding technology for high value-added leisure boats with free curved surfaces. The key process parameters were widely changed to derive the optimal conditions for the curvature welding through the FSW joints quality inspection. Deforma-tion after the welding, the tensile characteristic, and the hard-ness distribution of each specimen were examined according to different traveling speeds to evaluate the basic characteris-tics of the FSW joints.

*

(2)

2.  Experiment Procedure

In the experiment, 2 mm-thick AA 5083-O specimens were used, and 300 mm (L) ×  100 mm (W) specimens were first prepared for the bead-on-plate welding. Table 1 shows the chemical composition and mechanical characteristics of the material. SKD61 was used as the material for the Al alloy FSW tool. As shown in Fig. 1, the shoulder was mechanically cut in a convex form, and then thermally processed to achieve the desired hardness. The tool shape was determined by the material thickness. The probe length was 2 mm; the outer probe diameter, 1.6 mm; the inner probe diameter, 2 mm; and the shoulder diameter, 6 mm. For the thermal treatment, it was heated to 1,000–1,050 C and nitrogen-quenched, and then re-heated to 550–600 C, after which it was air-cooled and tempered.

[image:2.595.308.548.90.468.2]

Figure 2 shows FSW system for the experiments. This sys-tem can use the position information of the tool insertion depth through a load feedback function as input data even with a curved surface from the spindle head by the numerical control programs. The three-axis (X/Y/Z) machining center with 22 kW spindle (maximum rotation speed 12,000 rpm) and feed rate 1–24,000 mm/min used for bead-on-plate ing is shown in Fig. 2(a). In order to test for curvature weld-ing, a five-axis (X/Y/Z/A/C) machining center was used as shown in Fig. 2 (b). This machine has the same spindle and feed rate with the three-axis machine, and also has a circular table that can tilt (+30~−120 degrees) to the A direction in the X-axis and rotate to the C direction (360 degrees) in the Z-ax-is. In addition, the fixing jig for curvature welding was made with 250 mm radius of curvature as shown in Fig. 3. The steel jig firmly fixed the backing plate that prevented the deforma-tion with the tool pressure and supported the test specimen during the welding process with the upper cover and uniform-ly spaced bolts.

For the basic visual inspection and internal defect exam-ination of the bead-on-plate FSW joints under diverse weld-ing conditions, X-ray inspection and cross-sectional observa-tion were conducted to evaluate the welding quality. Mechanical characteristic tests such as the room-temperature tensile test and the Vickers hardness test were conducted on the curvature FSW joints, in addition to the measurement of the deformation after the welding. The tensile test was con-ducted in the direction perpendicular to the welding line. Fig-ure 4 shows the shape and dimension of the test specimen. Each specimen based on the ASTM-E8 standard was cut out from the curvature welds by the wire cutting machining. The hardness distribution was measured at 0.4 mm intervals hori-zontally from the base material through the FSW joints.

3. Results and Discussions

3.1  Bead-on-plate test results

Table 2 shows the conditions for the bead-on-plate weld-ing. The tilting angle of the tool was 0 degree, and the initial holding time after the insertion of the probe was fixed at 3 seconds to ensure tool heating stability. For the preparatory

0.35 0.29 0.08 0.67 4.32 0.12 0.18 0.07 - Bal. 145 290 20

Fig. 1 FSW tool dimension and whole shape.

Fig. 2 Appearance of (a) the two-dimensional and (b) three-dimensional FSW system for the experiments.

[image:2.595.312.542.511.647.2]
(3)

test, the insertion depth was determined within the range of 2.05–2.1 mm, which was slightly longer than the probe, to maximize the contact area of the shoulder. To avoid contact-ing between the probe end and the backcontact-ing plate, an alumi-num plate with the same thickness was placed under the test specimen. The rotation speed was changed within the range of 2,500–3,000 rpm, and the traveling speed, within the range of 27–50 mm/min.

In all the conditions, no significant defects such as surface cracks existed, as shown in Fig. 5, but burrs appeared in most of the FSW joints with a deep insertion depth of 2.1 mm. The burrs decreased in the FSW joints that had a shallow insertion depth of less than 2.07 mm. As shown in the X-ray inspection

and the cross-section examination, the voids, which are inter-nal defects along the line in the direction of the welding, showed a similar trend. When the insertion depth was shal-lower, the amount of burrs and the void size were smaller. Voids are mostly formed when the rotation speed is low or when the traveling speed is fast. With the convex shoulder, the insertion depth is proportional to the amount of burrs, and the amount of burrs is closely related to the void size.

In the additional test, the insertion depth was fixed at 2.07 mm, whereas the rotation speed was changed within the range of 1,000–2,500 rpm and the traveling speed, within the range of 50–250 mm/min. Table 3 and Fig. 6 show the test results. When the conditions were 1,000 rpm and 250 mm/

min, the probe was damaged during the FSW, seemingly due to the significant lack of heat input. The effects of heat input

[image:3.595.126.465.75.161.2]

Fig. 4 Dimensions of the tensile test specimen and cutting position.

Table 2 1st of FSW condition for the bead-on-plate experiment.

No.Insertion depth(mm) Holding time(sec) Rotation speed(Rt, rpm) Traveling speed(V, mm /min) Rt/V

1 2.1 3 3500 27 129.6

2 2.1 3 3000 27 111.1

3 2.1 3 3000 50 60

4 2.07 3 3000 50 60

5 2.07 3 2500 50 50

6 2.05 3 3000 50 60

[image:3.595.113.484.202.446.2]

Fig. 5 1st of bead-on-plate experiment results for comparison of the joint qualities, such as the bead appearance, X-ray radiography, and cross-section.

Table 3 2nd of FSW condition for the bead-on-plate experiment.

No.Insertion depth(mm) Holding time(sec) Rotation speed(Rt, rpm) Traveling speed(V, mm /min) Rt/V

1 2.07 3 2500 50 50

2 2.07 3 1500 50 30

3 2.07 3 1000 50 20

(4)

on the FSW joints will be further addressed in the next sec-tion.

Internal defects also appeared in the other conditions, but it was verified that the amount of burrs and the internal defect size decreased when the traveling speed was constant at 50 mm/min and when the rotation speed was lower. The burrs were discharged more toward the retreating side, where the tool rotation direction was opposite the welding direction, than toward the advancing side, where the tool rotation direc-tion was the same as the welding direcdirec-tion. It seems that at the retreating side, near the surface, the extrusion of the rotation tool and the metal flow in the material first transported and accumulated the heated material in the advancing side. In ad-dition, the internal defects were all located in the advancing side. It seemed that the metal flow in the advancing side was relatively not smooth because of the relative shear reaction between the rotating tool and the material.

3.2 Curvature welding test results

In the case of the welding of a structure, heat is generally applied to melt the welding area, which results in the welding deformation and residual stress. The welding deformation ad-versely affects the appearance and assembly quality of the structure, and may degrade its strength if the effect is signifi-cant. Therefore, close attention should be paid to the welding deformation particularly when low-melting-point thin plates are welded.16)

In this test, as mentioned in Section 2, a precision location control welding device was used for the AA 5083-O FSW for the curvature welding as shown in Fig. 7. Unlike in the bead-on-plate welding for the plates, the tilting angle was changed from the right angle to 2 degrees to ensure smooth metal flow in the material. Figure 8 shows the cross-section of the welds with and without tilting angle in curvature welding. It does seem to be smaller than that of without tilting angle and the size of void will be slightly decreased under applying the tilt-ing angle. It is considered that the tool tilttilt-ing is more applica-ble to decrease the internal void in the stirred zone (SZ) by activating the metal flow against without tilting condition. The insertion depth was determined as 1.97 mm, which was thinner than the depth of the material, to minimize the amount of burrs. Figure 9 shows the appearance of the welded joints after the FSW at a constant 1,000 rpm rotation speed and at a traveling speed within the range of 25–100 mm/min. In all the processing conditions, visually good appearance was

ob-served with the friction-stir-welded length of 200 mm. As shown in Fig. 10, the deformations of the welding specimens were measured, and the maximum heights before and after welding were compared. The test specimen of 2 mm-thick-ness could be fabricated using a three-roll bending machine which has one top-roll and two bottom-roll, and that speci-men has a curved surface by press forming from the top-roll. In this case, the AA 5083-O specimen has a total length (L) of 260 mm and the 200 mm-long bead-on-plate welded.

No significant change was observed in all the conditions, but the total length decreased to a maximum of 2 mm, and the maximum height (H) increased by as much as the decrease in the length. The clamping jig is removed after the peak tem-perature is cooled down to room temtem-perature. All the speci-mens are showing convex out-of-plane deformation and max-imum out-of-plane deformation occurred at a traveling speed of 100 mm/min. There has been some study on the deforma-tion of curved plates, including fricdeforma-tion-stir-welded plates, but it was reported that the longitudinal deformation in the FSW of the AA 6061 was lower, about 1/10–1/18 of those in the MIG arc welding and YAG laser welding.17) This is

be-cause the welding is achieved at about 70% of the melting point of the metal in the FSW, and the transverse and longitu-dinal shrinkages occur due to the compressive load, but no angular deformation basically occurs. In addition, the test specimen is firmly fixed by the clamping jig during the weld-ing process, so local deformation is also small.

Figure 11 shows the overall deformation distribution for

Fig. 6 2nd of bead-on-plate experiment results for the selection of the sound conditions.

[image:4.595.127.469.72.207.2] [image:4.595.313.542.248.399.2]
(5)

the measurements at 10 mm intervals for each welding condi-tion. The deformation distribution was almost the same re-gardless of the welding condition, but there was a slight dif-ference in the deformation near the maximum height that had the largest curvature. The FSW joints are formed in close re-lation to the heat input per unit welding length. In the FSW that uses friction heating, the heat input is simply

proportion-al to the tool rotation speed and inversely proportionproportion-al to the traveling speed, when the friction factor, tool pressing load, and shoulder shape are the same, as shown in the following formula.18)

Q= 4

3 π2µPNR3

V ∞

N V =

Rotation speed (Rt) Traveling speed (V) (1) In this formula, Q is the heat input per unit length, μ is the friction factor, P is the load per unit area, N is the number of rotations, R is the shoulder radius, and V is the traveling speed. Friction-stir-welded specimens show more out-of-plane deformation than that of not welded specimens due to the heat input. Compared to other welding processes, like arc welding etc., the amount of welding deformation is very small.

3.3  Mechanical test results of curvature welding

Figure 12 shows the results of the tensile test of the curva-ture welding at a constant rotation speed and a changing trav-eling speed. The test specimen for the curvature welds was made by flattening using a press after the wire cutting ma-chining. The graph shows that the strength of the FSW joints was almost constant regardless of the traveling speed. The joint efficiency was about 90% of that of the base material. The elongation of the FSW joints was about 50% of that of the base material. As shown in Fig. 11, the fracture after the tensile test occurred at all the FSW joints regardless of the welding condition. As suggested by the existing test results, in the FSW of the aluminum alloy, the FSW joints that have no defects are obtained by properly choosing the rotation and traveling speed when the tool plunge downforce is constant.18)

In this case, the FSW joint had a high strength according to the Hall-Petch equation, which represents the effects of the grain size on the strength in the grain refinement due to the dynamic recrystallization.19) However, the existence of

inter-nal defects was verified through the welding section, which served as the starting point of the fracture. Moreover, Fig. 13 shows the cross-section and SEM images for the typical frac-ture surface of the curvafrac-ture welding. In here, a fine ductile fracture is clearly observed in stirred zone after curvature welding. Based on this result, it is clear that the internal de-fects should significantly affect the tensile property.

Figure 14 shows the hardness distribution of each connec-tion in the same welding condiconnec-tion. Prior to welding, the hardness of the base material of the curvature was 100 Hv, which was slightly higher than that of the flat plate. This was attributed to the work hardening, and the average hardness of the friction stir area was approximately 97 Hv. This was al-most equal to the hardness of the base material of the curva-ture. In addition, the hardness of the stirred zone did not sig-nificantly differ regardless of the traveling speed, while the

Fig. 8 Comparison of the test result with and without tilting angle in curvature welding. (FSW condition: 1,000 rpm-100 mm/min)

[image:5.595.126.468.71.120.2]

Fig. 9 Curvature welded specimen at various traveling speeds at 1,000 rpm: (a) 100 mm/min; (b) 50 mm/min; and (c) 25 mm/min.

[image:5.595.70.269.167.287.2]

Fig. 10 Method of measuring the distortion of the curvature welded speci-men.

[image:5.595.67.269.349.439.2] [image:5.595.55.283.490.663.2]
(6)

hardness of the heat-affected area near the thermally and me-chanically affected zone (TMAZ) was slightly less than that of the base material. It seems that a higher traveling speed suppresses the grain growth by reducing the heat input and accelerating the cooling. In this study, however, the hardness only slightly increased due to the grain refinement in the stirred zone compared with that of the heat affected zone (HAZ).

4.  Conclusion

This study focused on lightweight shipbuilding and the

new material, AA 5083-O. The FSW welding device with a five-axis control and the welding jig with a 250 mm radius curvature were fabricated to try the 3D curvature FSW.

[image:6.595.100.496.75.305.2]

The key process parameters in deriving the optimal condi-tions for the 3D curvature welding were a rotation speed of 2,500–3,000 rpm and a traveling speed of 27–50 mm/min. In all the conditions, no significant defects such as surface cracks existed, but burrs appeared in most FSW joints with an insertion depth of 2.1 mm. The burrs decreased in the FSW joints that had a shallow insertion depth of 2.07 mm. This indicates that control of the insertion depth greatly influences the amount of burrs and internal defects.

[image:6.595.112.484.335.602.2]
(7)

With the curvature welding conditions derived from the preparatory test, good welding results were obtained at a con-stant 1,000 rpm rotation speed and at a traveling speed within the range of 25–100 mm/min with a welded length of 200 mm. For each welding condition, the total deformation was measured at 10 mm intervals. The deformation distribu-tion was almost the same regardless of the welding condidistribu-tion, but the deformation near the maximum height that had the largest curvature slightly changed. This was because the welding was a solid-state form of welding with a heat input that was significantly lower than that of fusion welding, and because a firm fixing jig was used in the FSW.

The strength of the FSW joints was almost constant regard-less of the traveling speed. The joint efficiency was about 90% of that of the base material, while the elongation of the FSW joints was about 50% of that of the base material. The hardness distribution of the FSW joints was almost the same as that of the base material regardless of the heat input, indi-cating a trend similar to what was obtained from the tensile test results. To improve the mechanical characteristics of the curvature FSW joints, there is a need for further studies on improving the tool shapes, such as the shoulder and the threads of the probe in the convex tool, as well as on the

con-trol of the key process parameters to increase the metal flow in the material.

Acknowledgments

I would like to thank Dr. Y.P. Kim for the various discus-sions and helpful experiments in this work. Financial support was provided by the Korea Ministry of Planning and Budget Foundation through the Korea Institute of Industrial Technol-ogy (KITECH).

REFERENCES

1) J.H. Lee, J.S. Yoon, C.H. Ryu and S.H. Kim: Advanced Materials Re-search 28 (2007) 981–984.

2) E. Cicala, G. Duffet, H. Andrzejewski, D. Grevey and S. Ignat: Mater. Sci. Eng. A 395 (2005) 1–9.

3) G.A. Moraitis and G.N. Labeas: J. Mater. Process. Technol. 1 (2008) 260–269.

4) Y. Liu, W. Wang, J. Xie, S. Sun, L. Wang, Y. Qian, Y. Meng and Y. Wei: Mater. Sci. Eng. A 549 (2012) 7–13.

5) W.M. Thomas, E.D. Nicholas, J.C. Needhan, M.G. Murch, P. Tem-ple-Smith, C.J. Dawes: International Patent Application PCT/ GB92/02203 and GB Patent Application 9125978.8. UK Patent Office, London (1991)

6) R.S. Mishra and Z.Y. Ma: Mater. Sci. Eng. Rep. 50 (2005) 1–78. 7) R. Nandan, T. Debroy and H.K.D.H. Bhadeshia: Prog. Mater. Sci. 53

(2008) 980–1023.

8) J.H. Cho, M.H. Kim and J.W. Choi: J. Korean Weld. Join. Soc. 2 (2012) 135–140 (in Korean).

9) Y. Li, L.E. Murr and J.C. McClure: Scr. Mater. 9 (1999) 1041–1046. 10) Y.C. Chen and K. Nakata: Mater. Des. 30 (2009) 469–474.

11) S.W. Song, B.C. Kim, T.J. Yoon, N.K. Kim, I.B. Kim and C.Y. Kang: Trans. JIM 7 (2010) 1319–1325.

12) J.K. Paik: Int. J. Nav. Archit. Oc. Engng. 1 (2009) 39–49.

13) Y.S. Sato and H. Kokawa: Metall. Mater. Trans., A Phys. Metall. Mater. Sci. 32 (2001) 3023–3031.

14) J.S. Noh, J.H. kim, G.H. Go, M.C. Kang: J. KSMPE 12 (2013) 37–44 (in Korean).

15) A. Amini, P. Asadi, P. Zolghadr: Advances in friction stir welding and processing (2014) 671–722.

16) J.M. Kim, C.S. Park, Y.P. Kim and H.S. Bang: J. Korean Weld. Join. Soc. 1 (2003) 93–98 (in Korean).

17) H. Okamura: J. Japan Weld. Soc. 7 (2000) 565–571 (in Japanese). 18) Y.G. Kim, H. Fujii, T. Tsumura, T. Komazaki and K. Nakata: Mater.

[image:7.595.54.284.69.242.2]

Sci. Eng. A 415 (2006) 250–254. 19) N. Hansen: Acta Metall. 8 (1977) 863–869. Fig. 14 Vickers hardness distributions of different traveling speeds at

Figure

Fig. 2 Appearance of (a) the two-dimensional and (b) three-dimensional FSW system for the experiments.
Table 3 2nd of FSW condition for the bead-on-plate experiment.
Fig. 6 2nd of bead-on-plate experiment results for the selection of the sound conditions.
Fig. 9 Curvature welded specimen at various traveling speeds at 1,000 rpm: (a) 100 mm/min; (b) 50 mm/min; and (c) 25 mm/min.
+3

References

Related documents

This research demonstrated the corrosion behavior of 6061 welds obtained by shielded metal arc welding (SMAW), friction stir welding (FSW), gas tungsten arc

The rotation rate (rpm) and traversing speed (mm/min) used for various samples is shown in table 1. As can be seen, the direction of welding and pin rotation is

Recapitulation, during the welding phase the plasticized material is exposed to displacement, extrusion and shearing mechanisms eased by the tool rotation, thrust and transverse

This paper looks at the review, on friction stir welding process, various welding variables like tool rotation, transverse speed, tool tilt, plunge depth and

1) The welding speed has a significant effect on the strength improvement as well as defect-free joint where the highest speed of 80 mm/min achieved good strength improvement at

In order to optimize the process parameters such as tool rotational speed, weld travel speed and tool tilt angle on tensile strength of friction stir welded dissimilar AA5083

From the table 6 and figure 3 it can see that the optimum machining condition to get high tensile strength is tool rotation speed of 1400 r.p.m, welding speed of 75 mm/min and

Experimental result reveals that the tool rotation speed, welding speed and axial force are the significant process parameters affecting the welding performance.. The predicted