TORSION STRENGTH OF ROUND BAR A6061 FRICTION WELD JOINT INFLUENCED BY FRICTION TIME, UPSET FORCE AND ONE-SIDE CONE GEOMETRY

The effects of friction time, upset force and one-side cone geometry on torsion strength of A6061 round bar friction weld joints were studied. Round bar commercial A6061 was friction welded with the initial compression force of 2.5 kN on stationary part and the rotated part had a revolution speed of 1600 RPM with the variations of friction time of 45, 50, and 55 minutes. The upset force variations of 5 kN, 7.5 kN and 10 kN with the same upset holding time of 110 seconds were applied. The stationary part of the specimen had friction area with the variation of cone geometry that represented by the ratio of upper diameter, D1 and lower diameter, D2, D1/D2. It was found that friction time and the ratio of D1/D2 affected torsion strength in the upset force below 10 kN. In the case of the higher upset force of 10 kN, the upset force more dominant to affect the torsion strength of the continuous drive friction weld (CDFW) joints. The specimen with maximum torsion strength has more precipitates in grains of microstructures compared to that of the specimen with lower torsion strength.


INTRODUCTION
Aluminum alloy A6061 is the aluminum alloy that contains alloys of magnesium (Mg) and silicon (Si). This kind of aluminum alloy is found in wide applications such as machine components, ships, aircraft structures, heavy vehicles and rail transportations (Bauccio, 2001). It is due to its properties that it has moderate tensile strength, good formability, good weldability and good corrosion resistance which is better than A2024 alloys (Budinski, 1996). However, there is difficulty in joining this alloy and other aluminum alloys using conventional welding technique. It occurs due to the existence of aluminum oxide that exists during melting in welding and the high thermal conductivity that makes heat difficult to be concentrated in the joint region to yield good weld joint (Charit et.al., 2002).
The difficulty to join Aluminum can be solved using a solid-state welding process such as friction welding. Friction welding is one of manufacturing methods to join metals or a non-metals by using heat generated from friction on the part that to be joined. The weld joint is the result of a coalescence of materials under compression state when the specimens rotate or relatively move each other (Zhang, et.al., 2006). A friction welding method that appropriates to join round bar of pipe metal is continuous drive friction welding (CDFW) which also known as rotary friction welding or spinning friction welding. This method can generate heat from friction between the rotated and stationary contact surface under compressive force.
There are parameters that affected the strength of CDFW joint. An appropriate set up on those parameters can increase the strength of CDFW joint, such as tensile, torsion strength. Many researchers studied the effect of parameters such as Sathiya et.al. (2007). They found that parameters such as friction pressure, friction time and upsetting pressure and upsetting time affected the tensile and impact toughness of ferritic stainless steel. The combination of those parameters that can give adequate heat input can produce maximum tensile strength and impact toughness of the weld joint. Irawan et.al.(2012) studied the effect of double chamfer angle on the tensile strength of spinning friction weld joint. They found that chamfer angle of 30 degrees gave maximum tensile strength of the joint and the chamfer angle become other parameters that can influence the strength of the CDFW joint Mohandas (2007) andIrawan et.al.(2016a) also confirm that surface roughness of the friction area can also affect the tensile strength of CDFW joint. Mohandas (2007) found that the increasing roughness Ra up to 5 µm can increase the notch tensile strength, but the strength becomes lower for surface roughness over 5 µm due to the banded microstructures. Irawan et.al. (2012) reported in the case of A6061 that using double chamfer on the friction area, the lower surface roughness of 0.6 µm can give the higher tensile strength of CDFW joint compared to the specimen with the higher lower surface roughness that has the lower tensile strength.
Besides the tensile strength of CDFW joint, torsion strength is one essential of mechanical properties that is important for the component that endures torsion load such as a shaft in an engine or a generator. It is static mechanical properties but it can be used to ensure the strength and the safety of the shaft. Irawan et.al (2016b) studied about torsion strength of CDFW joint. They used one-side cone geometry at the stationary friction area to increase torsion strength of CDFW joint. They found that cone geometry with the smaller ratio of higher diameter and lower diameter which formed almost complete cone geometry in the stationary part of the specimen can produce higher torsion strength of CDFW joint. However, the effects of friction time, upset force and one-side cone geometry on strength of CDFW joint especially torsion strength were not uncovered yet in order to improve torsion strength of A6061. This paper reveals the effects of friction time, upset force and oneside cone geometry on torsion strength of A6061 CDFW joint based on torsion strength test, macro and microstructure analysis, microhardness test and temperature measurement on formed flash.

MATERIALS AND METHODS
The material used in this study was commercial round aluminum alloy A6061. This alloy has main alloys of magnesium and silicon. Table 1 shows chemical composition of A6061 used in this study. Round bar A6061 with 22.5 mm diameter was cut using a saw machine with cooling media of water as a coolant to prepare CDFW specimen. The geometry of CDFW specimen that machined by turning process can be seen in Figure 1. CDFW specimen has two parts, which are rotating part and the stationary part. In this case, rotating part is the left side which has flat friction area and stationary part is the right part that has one-side cone geometry on the friction area that represents the ratios of D1/D2=0; 0,25; 0,65; 0,8; 1. In CDFW process, the rotating part of the specimen was set in the chuck of the lathe machine. The stationary part of the specimen was attached to chuck that connected to the hydraulic cylinder with the capacity of 50 kN. Before started friction welding process, both friction surfaces were cleaned using acetone. The rotation speed of rotated specimen was 1600 rpm, then the stationary part of the specimen was engaged to the rotated specimen by applying compression force of 2.5 kN for friction time of 45, 50, and 55 seconds. After the friction time was reached, the lathe machine was shut down and the CDFW specimen that yields flash due to friction welding process was continued to endure the upset force variation of 5, 7.5 and 10 kN for 110 seconds and then cooled in the air. In the range of this study, the selection of maximum upset force of 10 kN was done to ensure the safety of the machine structure and the CDFW process.
Friction weld specimen was machined to produce torsion strength testing specimen according to Figure 2. Location of CDFW joint was in the center of the specimen. Torsion strength test was conducted using the torsion strength testing machine. Torsion loading during the test was controlled by giving the angle of twist to the specimen with speed of 1 degree/second until the specimen fractured. There were three replications of CDFW specimens for each variation of friction time, up set force and one side-cone geometry. Figure 2. Geometry of torsion strength test (ASTM, 2004). Observation was also performed on macro and microstructures of CDFW joint. Weld joint contains three zones of fully plasticized zone (Zpl) in the center and partly deformed zone (Zpd) (Ozdemir, 2005). In the Zpl zone, there are Zpl1 in the center and Zpl2 which is beside of Zpl1 (Irawan et.al., 2016). Area of Zpl1, Zpl2 zones and porosity zone were measured using ImageJ software. The hardness of Zpl1, Zpl2 and Zud were also measured using micro-Vickers hardness method with 50gf force indentation load for 6 seconds. Figure 3 illustrates the relationship of D1/D2 ratio and torsion strength of CDFW joint with the initial compression force of 2.5 kN and upset force of 5 kN. It shows that D1/D2 ratio influenced torsion strength of A6061 CDFW joint, where the lowest D1/D2 ratio of 0.02 with friction time of 50 seconds gave maximum torsion strength of 120.63 MPa. Friction time at each D1/D2 ratio has the different effect on torsion strength of CDFW joint. It shows that D1/D2 ratio has more significant effect on torsion strength of CDFW joint than friction time. In addition, friction weld joint has different properties to withstand torsion load compared to tensile load that perpendicular to CDFW joint. The same results are showed in Figure 4     In case of specimen with upset force of 10 kN, the effect of upset force is more dominant to give effect on torsion strength of CDFW joint that one side cone geometry, so that specimen with D1/D2 ratio of 1 yields maximum torsion strength of CDFW joint followed by specimen with D1/D2 ratio of 0.25 and friction time of 45 seconds that has torsion strength of 165.11 MPa. Figure 6 shows selected data for torsion strength of CDFW joint that has maximum torsion strength for each upset force. It is found that maximum torsion strength occurred in the specimen with the upset force of 10 kN and friction time of 55 seconds and ratio D1/D2 ratio =1. It is thought that in higher upset force, the effect of the upset force is more dominant than friction time and one-side cone geometry (D1/D2). It is supposed that with longer friction time and higher D1/D2 makes higher heat input to soften the weldment and more easily to more plastically deformed by high upset force (10kN). As the result of the higher portion of plastic deformation during the upset stage, the weldment has more slips and dislocations that supposed to have the higher hardness to yield higher torsion strength. It is thought that the effect of one-side cone geometry will take more effect in lower friction time compared to the range in this study due to its lower heat input during CDFW process. Figure 7 shows macrostructure in longitudinal section of 13 mm diameter torsion specimen of CDFW joints for the specimen with D1/D2 = 1, friction time of 55 seconds, 10 kN upset force that has maximum torsion strength. It has three zones of Zpl1, Zpl2 and Zpr (porosity zone). Figure 8 shows of that for CDFW specimen with D1/D2 = 0.02, friction  time of 45 seconds that has low torsion strength. The area for each (Zpl1, Zpl2, Zpr) was measured by ImageJ software and the results were shown in Table 2. It can be seen that specimen with higher torsion strength D1/D2 = 1 has the smaller area of Zpl1 and Zpl2 that has porosity, Zpr of 0.279 mm 2 , compared to the area of Zpl1 and Zpl2 for the specimen with lower torsion strength which has D1/D2 = 0.02 and friction time of 45 seconds. It has the correlation with the mean hardness at those zones as shown in Table 2. It is found that mean hardness in Zpl1 for the specimen with maximum torsion strength is higher than that of the specimen with minimum torsion strength and D1/D2 = 0.02. Higher hardness in ZPl1 zone contributes to higher torsion strength of CDFW joint, even there is a small portion of porosity in the weldment. Figure 9 is thermal cycles for two specimens with maximum and minimum torsion strength. The temperature was measured on the formed flash of CDFW from beginning to the end of CDFW process using an Infra-Red Thermogun. It can be seen that thermal cycle for the specimen with D1/D2 = 1 is higher with the maximum temperature of 201.6 o C. Meanwhile, maximum temperature for the specimen with D1/D2 = 0.02 is 184.4 o C with different value of 16 o C. In this state, it is thought that heat input for the specimen with D 1 /D 2 = 1 is little higher than the specimen with D1/D2 = 0.02. The higher heat input and the effect of higher upset force of 10 kN makes precipitates more dispersed in to the aluminium matrix grains as shown in Figure 10. More gray color in the aluminum grains contains more precipitates of Mg2Si contributes to yield higher hardness in grains (Irawan et.al.,2016b). According to Figure 10, even there is no significant difference of grain size in two specimen microstructures. However, due to more precipitates exist in the grains of specimen with D1/D2 =1, the higher hardness occurred in Zpl1 zone as shown in Table 3. Therefore, the specimen with D1/D2=1 and higher hardness has maximum torsion strength. Figure 11 and 12 show fractured torsion specimen with has maximum torsion strength (D1/D2 =1) and minimum torsion strength ((D1/D2 = 0.02). Both specimens were fractured in shear mode because the fracture surface is perpendicular to longitudinal direction due to torsion loading. The specimen with maximum torsion strength fractured beside the center line of weldment of in Zpl2 zone due to lower hardness in the zone. Meanwhile, specimen with lower torsion strength fractured in the center of weldment because the lower hardness in Zpl1 zone as confirmed in Table 2. In addition, fracture surface of specimen with higher torsion strength had more flat fracture surface (Fig.11a) than that of specimen with lower torsion strength (Fig.11b). It is thought that fracture occurred in the weakest zone that has lower hardness and affected by the contour of Zpl1 and Zpl2 zone as seen in Figure 7 & 8. Namely, specimen with higher torsion strength fractured in Zpl2 that has narrower wide of zone that supposed to yield more flat morphology of fracture surface of torsion strength test specimen. Meanwhile, specimen with lower torsion strength fractured by shear stress in Zpl1 with broader zone of Zpl1 that may give less flat fracture surface. It is thought that formed zones in the weldment of CDFW also affect the morphology of fracture surface.

CONCLUSION
One-side cone geometry at friction surface of stationary part of CDFW specimen affected on torsion strength of CDFW joint of Aluminum alloys A6061. Smaller ratio of friction area diameter and specimen diameter (D1/D2) and shorter friction time yields higher torsion strength of CDFW joint in low upset force. However, higher upset force gave more effect in increasing torsion strength of CDFW joint and decreasing the effect of one-side cone geometry due to longer friction time. Higher upset force contributes to make bigger portion of plastic deformation to produce more dispersed precipitates in aluminum grains that is thought to have correlation to yield higher torsion strength of CDFW joint.