Preliminary study on friction stir spot welding of plastic sheets with non-metallic tools

In recent years, 3D printing technology of engineering plastics is often used in various fields to develop prototypes of parts. Friction stir spot welding of lap joints of different engineering plastics has considerable potential for industrial applications. This study used a self-developed non-metallic tool to successfully form the PLA joints on a modified conventional drilling machine. The results show the average downward force and average welding torque decreased with increasing the rotating speeds and dwell times. The stir zone width is slightly increased with increasing the rotating speed, and its signification increased with the dwell time. The increase in stir zone width on the different rotating speeds and the dwell times is similar to the increasing trend in average temperature. The average temperature multiplied by the dwell time defined as the bonding factors is similar to the diffusion bonding. The stir zone width is increased with increasing the bonding factor because there has still some distance between the shoulder to the lap surface, the stir action by the shoulder is not easy riches to the lap surface, so the stir zone only occurs within the FSP zone and the stir zone width never exceeds the shoulder diameter.


Introduction
The sizes of 3D printed products are often limited by the printable range of the 3D printer, so the larger products need to be disassembled and printed before joining.The joining strength affects the product function, which makes bonding technology more and more important under the development trend of 3D printing.Therefore, the development of FSW in the field of 3D printing to join thermoplastic engineering plastics has a production advantage.Review the bonding methods of 3D printed plastic products in recent years, except solvent bonding, there are many bonding technologies, such as using an electric soldering iron [1] to heat and melt the joint surface, and repeated application along the joint line to form a weld bead.Use a butane soldering iron [2] to heat the bonding wire of the two plastics, so that the materials are melting and bonded, but the structure will be deformed and the weld bead will be unsightly.Using melting 3D wire with a 3D printing pen [3], fill it into the pre-designed filler groove on the joint surface for filler welding.A plastic rod [4] is clamped on a hand-held engraving machine and pressed through rotation to form frictional heat so that the welding rod sticks to the joint surface to form a weld bead.This method is also applied to large-scale 3D printing and all-round welding of the parts [5].However, if the joint thickness is not enough, the joint strength of the products will be affected.
FSW is used to join thermoplastic materials, which is known as a low heat input method to plastic flow for materials.The material flow of PMMA material during FSW is significantly different from that of metal materials [6].The tool shapes, such as frustum, cube, and triangle, to explore the influence of probe profile on the flow and strength of PMMA material and found that the highest bonding strength can be obtained by using the frustum-shaped probe [7].The physical properties of different thermoplastic materials, such as high-density polyethylene (HDPE), polyamide 6 (PA6), polyvinyl chloride (PVC), etc. explore the measured temperature of the probe [8].Many studies have found that the tensile strength of the weld bead are related to the profile of the probe, the good weld quality cannot be achieved with conical probes, while square probes have the potential to obtain high-quality weld beads of PP material joints [9].It was found that when the tool is rotated counter-clockwise the heat-softened material is prevented from being carried out of the shoulder surface in the direction of thread rotation, resulting in a defect-free weld bead [10].The FSW of PLA materials with pin-shaped probes is more likely to produce high-quality weld beads than conical or threaded probe tools, and the welding temperature is in the range of 75-110°C at low tool feed travel speeds.Highstrength welding beads inside, can be effectively obtained [11].Based on the above literature, the key to successful welding is faster friction heating to reduce material softening time.However, conventional tools are often made of steel, which tends to warpage the specimen during tool penetration, as thermal conductivity makes it difficult for frictional heat to build up.Therefore, this study tries to use a non-metallic tool to FSSW of Polylactic Acid (PLA) sheets.The downward force, welding torque, and temperature are measured.Besides the effects of rotating speed, dwell time, and the average temperature on the weld formation are discussed.

Experimental apparatus
To study the welding properties during the FSSW process, a traditional drilling machine with a measuring system is used in this study shown in Figure 1.The tachometer and the torsion meter are installed on the rotating spindle, and the vice and the load cell are stacked and assembled on the X-Y moving table.The downward force and the welding torque are respectively measured by the load cells and torsion meter; the temperature during the welding process of the tool/specimen is recorded by the Infrared camera, and their signals are recorded by the data acquisition system and then feed into the personal computer for data analysis.

Tool and specimens
The welding tool is made of abrasive material [12] with a simple appearance.Its geometry and size are shown in Figure 2 (a).The upper and lower specimens are the block (35×35×3 mm), a thermocouple hole at the joint interface near the pin of 1 mm.To facilitate observation of the bonding interface of the upper and lower specimens, the white and natural colour of Polylactic Acid (PLA) wires are used in their 3D printing process, respectively.The printing conditions are set in the specimen-slicing software as shown in Table 1.Table 1.3D printing conditions for specimens.

Experimental procedures
Because the high-speed extrusion of the tool into the PLA test piece is easy to cause warpage and damage to the specimen, this study divides this process into two stages.The first stage occurs at the probe is plunged into the upper surface of the specimen by 1 mm and dwells for 10S.Then the probe continues to plunge until the shoulder is pushed into the upper surface of the specimen by 1 mm, and starts to record the dwell time of FSSW.When the time reaches the set value, the third stage is the tool pull out of the test piece.The details of the welding conditions are shown in Table 2. Before the welding, the upper and lower specimens are cleaned with alcohol.During the welding process, the downward force, torque, and temperature are synchronously measured.After the welding, the top view and cross-section view of the weld is observed with an optical microscope.Table 2. Welding conditions.

Experimental results and discussions
The requirement of performed FSSW is faster friction heating to shorten the time of softening material, so the tool rotating speed and downward force are often the important operation condition.The welding torque, temperature, and dwell time, they often use to discuss the formation of the FSP zone of the weld.These are also associated with joint strength, such as the joint region, and joint interface.Therefore, this study discusses the effect of rotating speed, dwell time, and the average temperature on the weld formation as follows.

Dynamic responses of downward force, welding torque, and temperature
The time histories of the downward force, welding torque, and temperature under dwell time of 15s at three stages are shown in Figures 4(a), (b), and (c), respectively.To identify the influence of tool rotating speed, 600 rpm is represented by real line, 800 rpm is represented by dotted line, and 1000 rpm is represented by center line.The first stage (I) is similar to preheat of the probe, the probe touches the upper specimen until it continues to penetrate 1 mm below the surface of the upper specimen.The downward force and welding torque are increased from a to b caused of the material of the specimen resists the penetration of the probe, and it decreased from b to c due to the formation of the flash around the probe.The temperature does not change significantly because the measurement point is far from the probe.The temperature is gradually increased in the second stage (II) because the probe continues to penetrate to the lower specimen until the shoulder is penetrating 1 mm below the surface of the upper specimen.The downward force and welding torque are increased from c to d caused by the penetration of the probe and shoulder, and it decreased from d to e due to the softened material is squeezed out and flash building up around the tool.When the dwell time is rich to the 15s, the welding tool is pulled off on the third stage (III), and the downward force and welding torque are decreased rapidly.However, the temperature continues to rise for about 1~2s and then gradually decreases because of the low thermal conductivity of PLA. Figure 4 also shows the downward force and welding torque are the largest at a low speed of 600 rpm, and the temperature is lowest at a low speed of 600 rpm, which is related to the influence of temperature on material hardness.This is why the downward force and welding torque are faster decreased at the dwell time of stage II, 1000 rpm has the lowest downward force and welding torque, and largest temperature at point e before the tool pull-off.Figure 5 shows the time histories of the downward force, welding torque, and temperature under a dwell time of 45s.It is obvious that after the dwell time of 15s on stage II, the downward force and welding torque are continue to decrease, but the temperature is continues to increase.This means that with increasing dwell time at a high temperature, there is more softened material can be stirred by the welding tool, which indicates that temperature and dwell time are the important factors on the weld formation.

Observation of the top view and cross-section view of weld
Further understanding the relationship between rotating speed and dwell time on the weld formation, Figures 6 and 7 respectively show the top view and of the cross-section view of the weld corresponding to Figures 4 and 5. On the top view of the specimen that is shown in Figures 6 and 7, two circles about the contour of the welding tool can be observed on the surface of the upper specimen, and the flash accumulates around this tool contour, which is called the weld of FSSW.The colour of the large circle is white, but the colour inside the small circle is natural because the penetration depth of the probe is set at 0.7 mm below the surface of the lower specimen.At the cross-section of the weld, a stirred zone, as known as friction stir processing (FSP) zone that is associated with the width of the stir zone width (W SZ ) can be found in the profile of the weld.In this zone, the heat-softened material will be stirred by the tool, so the interface between the sheets will be deformed, such as zigzag, or hooking.As increasing the rotating speed of the welding tool, Figures 6 (a) to (c) show the flash accumulation around the large circle is increased, and some flash materials are separated on the surface of the upper specimens.The stir zone width (W SZ ) is slightly increased with increasing the rotational speed.It may be that the dwell time of 15s is not enough, and the material of the joint surface cannot fully stir.With increasing the dwell time to 45s, Figures 7 (a) to (c) show the flash accumulates around the large circle and separate flash on the upper surface of the specimen are both increased, especially the accumulate high that can be observed on the cross-section of the weld.The difference of tir zone width (W SZ ) between the rotating speeds also increases slightly, but it is significantly increased in each rotating speed.This indicates that the dwell time is more important than the rotating speed of the tool, it affecting sufficient mixing of the interface and from the FSSW joint at high temperatures.

The stir factor on the bonding width
The above results indicate the effect of rotating speed and dwell time on the appearance of weld formation.However, the influence of parameters in the process related to the joint results is not clear.Therefore, the average downward force, average welding torque, and average temperature during the dwell time of different rotation speeds will compare with the stir zone width (W SZ ), as shown in Figures 8 (a), (b), (c), and (d), respectively.Figure 8 (a) shows the average downward force (Fd AVG ) decreased with increasing the rotating speed and the dwell time.This trend is inversely proportional to the average temperature (T AVG ) shown in Figure 8 (c).Because the material softened by friction heat has high fluidity, it is easy to squeeze out by the rotating tool and extrusion from the Tool/PLA friction interface.Therefore, as the tool speed increases or the dwell time increases, the flash formed around the tool becomes obvious, this can be observed in the weld surface shown in Figures 6 and 7, which could affect the material being stirred in the FSP zone.Moreover, the welding torque often correlates with friction behaviour of the friction interface between the tool and the material, which is related to the heat generated by sliding friction and the sticking friction inside the softened material.Figure 8 (c) shows the average welding torque (τ AVG ) is decreased with increasing the rotating speed and the dwell time.It has been known the softening temperature of PLA is about 50 o C to 60 o C [13], it can be judgement that the friction behaviour on the dwell stage is mainly sticking due to the softened PLA material at high temperatures.Figure 8(d) clearly shows the increase in stir zone width (W SZ ) on the different rotating speeds and the dwell times, which is similar to the increasing trend in average temperature (T AVG ) shown in Figure 8(c).This could be interpreted when the tool gives a certain pressure on the overlapping surface; the interface material of the upper and lower specimens takes temperature and time to form a good joint.Furthermore, the average temperature multiplied by the dwell time is defined as bonding factors, and the relationship between the bonding factors and the stir zone width (W SZ ) is shown in Figure 9.  Figure 9 illustrates that the stir zone width (W SZ ) is increased with increasing the bonding factor, this trend is similar to diffusion bonding.Because there has still some distance between the shoulder to the

Figure 1 .
Figure 1.Photo image of the traditional drilling machine with the measuring systems.

Figure 2 .
Figure 2. Schematic plots of sizes of (a) the welding tool and (b) the specimens with thermocouple hole.

Figure 3 .
Figure 3.The schematic plots of FSSW stage on this study.

Figure 4 .Figure 5 .
Figure 4. Time histories of the (a) downward force, (b) welding torque, and (c) temperature under dwell time of 15s and different rotating speed of 600, 800 and 1000 rpm.

Figure 6 .
Figure 6.Top view and cross section view of the specimen under different rotating speed of (a) 600, (b) 800, (c) 1000 rpm, and dwell time of 15s.

Figure 7 .
Figure 7. Top view and cross section view of the specimen under different rotating speed of (a) 600, (b) 800, (c) 1000 rpm, and dwell time of 45s.

Figure 8 .
Figure 8.The (a) average downward force, (b) average welding torque, (c) average temperature, and (d) stir zone width under different rotating speeds and dwell times.

Figure 9 .
Figure 9. Relationship between the bonding factors and the stir zone width.