An approach to weld dissimilar polycarbonate and high-density polyethylene by friction stir spot welding

The test samples were prepared from commercial PC and HDPE sheets with thicknesses of 3 and 4 mm, respectively. The specimens were cut with a dimension of 101 × 24 mm, which matches the standard test specimen dimension specified in ASTM D3163.

A conventional drill press is modified and used to conduct the FSSW process, as shown in Fig. 1. The drill spins with four different rotational speeds of 800, 1400, 2300, and 3500 rpm, while the vertical feed (plunging) is controlled by means of a specially designed and controlled mechanism, including a DC motor, gearbox, and microcontroller. The DC motor and the gearbox give the vertical forward and backward motion. Additionally, the microcontroller is dedicated to controlling the input voltage to the motor and hence provides the specified plunging rate, in addition to adjusting the required dwell time.

Furthermore, the specimens are placed on a specially designed wooden fixture, which in turn impedes the specimens’ axial and rotation motion and attains the required overlap between the specimens of 24 mm. A groove is prepared in the wooden fixture to carry a thermocouple for temperature measurement during the welding process. The thermocouple is laying under the center of the overlap area.

The FSSW tool is fabricated from plain carbon steel with Rockwell hardness HRB 92.3. The tool has a tapered shape with pin length and diameter of 5.5 and 7.5 mm, respectively, and a shoulder diameter of 18 mm. The tool’s tapper and concavity angles are 15° and 6°, respectively, as shown in Fig. 2.

The welding process may be subdivided into three main phases, namely, plunging, stirring, and retracting. The plunging process is performed by moving the drill crosshead with the prespecified plunging rate of 15 mm/min downward toward the workpieces; the rotating tool keeps moving until the required plunging depth is achieved, which is 5.5 mm; then, the plunging motion stops, and the stirring phase starts for the designated dwell time. Three dwell times are selected in the present study, namely, 40, 80, and 120 s. Then, the retracting phase starts, moving the tool out of the workpiece. During the whole test, the resulting temperature at the center point of the overlap area at the bottom of the lower workpiece was recorded using a K-type thermocouple.

The test includes a variation of the two primary welding parameters: the rotational speed and the dwell time, while the plunging rate was kept constant during all tests. Table 1 summarizes the welding parameters.

Four replicates of each weld process were tested under lap shear configuration. The ultimate lap shear forces were determined and recorded using WDW 300 computer-controlled electronic universal testing machine, with a crosshead speed of 1.27 mm/min (0.05 inch/min) according to ASTM D3163.

To examine the microstructure and surface topology of selected specimens, TESCAN VEGA 3 scanning electron microscope (SEM) was used. The specimens were first prepared by coating with gold inside a turbomolecular pumped coater Quorum Q150T ES.

To identify the base material properties, a series of tests were conducted on both PC and HDPE sheets. Differential scanning calorimetry (DSC131 Evo SETARAM) was used to measure the actual glass transition temperature (T_g), melting temperature (T_m) for PC sheets, and T_m for HDPE. Additionally, the melt flow rate (MFR) and the melt density (ρ_m) were tested by Zwick Roell Mflow plastometer; the tests were conducted at 2.16 kg/190 °C and 2.16 kg/260 °C for HDPE and PC, respectively, according to ASTM D 1238. The base materials’ tensile strength was tested according to ASTM D638 using WDW 300 universal testing machine. Table 2 summarizes all the obtained material properties.

Figure 3 depicts the effect of the process parameters on the average values of the lab shear forces during the FSSW of similar PC sheets. The curves show that the lap shear force is inversely proportional to the tool rotational speed while slightly increasing with the increase of dwell time. This may be explained by increasing the dwell time, sufficient stirring, and mixing the materials takes place, which increases the friction heat transferred to the workpieces, and the weld strength enhances. While by increasing the rotational speed, high inertia force is generated and acts to increase the ejected material from the weld zone to outside the weld area, and thus, the material in the weld zone is reduced, and consequently, the weld area and weld strength deteriorate. The same findings were reported previously in [10].

Accordingly, the optimum weld strength for PC sheets was obtained at a low rotational speed (800 rpm) and high dwell time (120 s).

Figure 4 reports the effect of tool rotational speed and dwell time on the lap shear strength of similarly welded HDPE sheets. During the weld of similar HDPE at low rotational speeds (800 and 1400 rpm), the frictional heat generated increases with increasing the rotational speed and dwell time up to 80 s, and hence, good mixing and stirring occur between both workpieces, leading to proportional weld strength, whereas the weld strength slightly decreases at higher dwell time, which may be ascribed to the increase of the quantity of extruded material from the weld zone leading to a decrease in the weld area and thus the weld strength slightly decrease.

However, at higher rotational speeds (2300 and 3500 rpm), and due to the increase of inertia force and the quantity of extruded material, the material separates from the workpieces and sticks to the tool pin, producing a plastic ring, as shown in Fig. 5; the ring functions as a thermal isolator between the rotating tool and the HDPE workpieces; and thus, a reduction in the frictional heat takes places, and subsequently, the weld strength fails.

Now, it could be observed from the above results of the similar weld of PC and HDPE that the highest lap shear force during welding of PC and HDPE were 1.6 kN obtained at 800 rpm and 120 s dwell time and 0.915 kN obtained at 1400 rpm and 80 s dwell time, respectively.

PC and HDPE were welded for the first time by FSSW to investigate the feasibility of the welding technique on both materials. The investigation includes the weld of both dissimilar materials, first when PC was on top and HDPE on the bottom (PC to HDPE) and then when HDPE was on top and PC on the bottom (HDPE to PC). The variation of induced temperature during the process and the weld’s lap shear strength were recorded and illustrated in Figs. 6 and 7 for PC to HDPE and in Fig. 13 for HDPE to PC, respectively.

The resultant strength is lower than that of the similarly welded base material, especially in comparison with PC. This is due to the fact that each material has a different range of T_g and T_m, in which the HDPE reaches its complete melting, while PC is still in the solid form, and thus, an incomplete bonding takes place between molten HDPE and solid PC material.

The above results conclude that the weld strength is highly dependent on both process parameters and the resulting process temperature. By correlating the temperature curves, in Fig. 7, with the obtained values of lap shear forces in Fig. 6, it could be observed that the higher strength is obtained during the solid-state welding process, or in other words, at temperature ranges below the melting temperatures of both materials. When the temperature increases and exceeds the T_m of HDPE (the lower workpiece), a molten material exists in the weld region, and during the material’s cooling and solidification, the material shrinks, leaving cavities and pores. This porous structure results in high residual stresses in the material, resulting in a reduction in the weld strength.

The results may also be investigated using macroscale observation. At low rotational speed, e.g., 800 rpm, the quantity of frictional heat generated leads to low process temperature. At this temperature range, both materials are entirely in a rigid solid form, and thus, the plunging and stirring process acts to press the upper sheet toward the lower sheet, as shown in Fig. 8, without any mixing or bonding between both materials, and thus, weak weld joints are produced.

At rotational speed of 1400 rpm and 40 s dwell time, a small but homogenous weld area was formed due to the better stirring in consequence of the higher speed. By increasing the dwell time to 80 s, the temperature increases reaching the T_g of PC and exceeding the T_m of HDPE, leading to the presence of molten material in the weld zone. Because of these reasons, some irregularities and cavities were observed in the weld zone. These irregularities function as residual stress risers in the weld zone. By increasing the dwell time to 120 s, the HDPE and PC workpiece exceeds their T_m and T_g, respectively. Thus, a soft–molten region occurs below the tool. A significant quantity of the molten material sticks to the sample fixture, resulting in excessive material loss and separation from the weld zone. For these reasons, the lap shear force decreases with increasing dwell time. Figure 9 depicts the weld area shape’s development with dwell time at 1400 rpm.

Figure 10 depicts the weld zone shape during welding at 3500 rpm, which is almost similar to that of 2300 rpm. At these rotational speeds and 40 s dwell time, higher heat is generated than that reached with 800 and 1400 rpm, by which the process temperature is slightly below the T_m of HDPE and T_g of PC, so both materials are in the softened state. Because of the material softening, big and homogeneous weld rings are produced, resulting in a strong joint between the upper and lower workpieces, and thus, high lap shear forces were obtained. When the dwell time increases to 80 and 120 s, the process temperature exceeds the T_m and T_g of HDPE and PC, resulting in the occurrence of irregularities, cavities, and pours in the weld area, and then, the molten material sticks to the sample fixture leading to the material’s deterioration in the welding zone, and material separation occurs. For this reason, the weld strength fails at higher dwell times.

Furthermore, the microscale investigation shows the same findings; samples of 1400 rpm and 40 s dwell time and 3500 rpm and 40 and 120 s dwell times were scanned using SEM. Figure 11a shows the microstructure at 1400 rpm; both materials are clear, and the weld morphology seems to be very homogeneous, whereas the weld junction’s thickness is very thin and almost not seen. By increasing the magnification, it could be noted that the structure is clear of any voids or cavities, as shown in Fig. 11b.

Figure 12 shows the microstructure of welded samples at 3500 rpm; in the case of 3500 rpm at 40 s, the sample structure is also clear from any cavities or irregularities, and the weld junction thickness is also higher than that obtained at 1400 rpm, with a thickness of approximately 500 µm. Conversely, at 3500 rpm and 120 s, due to the excessive heating and the complete melting of the HDPE, and the PC workpiece comes to a region very near to its complete melting, an irregular structure was noticed, with excess cavities and voids. For this reason, the lap shear strength deteriorates and reaches less than a quarter of the values obtained at 40 s.

Generally, it may be concluded that the efficient welding condition may be obtained at moderate process temperatures slightly below the T_m or T_g of both welded materials. Also, the excessive dwell time and rotational speed lead to highly deformed material and hence low weld strength.

Welding of HDPE to PC (HDPE workpiece at the top and PC at the bottom side) was unsuccessful except with 2300 rpm and 120 s and 3500 rpm and 80 and 120 s with average maximum lap shear force of 0.045, 0.15, and 0.115 kN, respectively. The obtained weld strength was very weak compared with the weld strength of both base materials and with that of dissimilar weld of PC to HDPE, whereon the maximum weld strength is approximately one-third of the peak weld strength obtained in welding PC to HDPE and approximately 10% and 15% of the peak weld strength of similarly welded PC and HDPE, respectively. Otherwise, the specimens directly break during the removal from the wooden fixture. Analyzing the process temperature in Fig. 13, it was evident that when the HDPE workpiece was on top, the temperature was extremely low, below the T_g and T_m of PC and HDPE, respectively. Since almost all the area of the tool pin and shoulder are in contact with the upper workpiece, HDPE, and due to the fact that the dynamic friction coefficient between steel and HDPE is very low [17], the process temperature is much lower than that when the PC sheet was on top. This indicates that inadequate frictional heat is generated and transferred to the materials, and consequently, no softening takes place, and hence, a weak bond is obtained.

At high rotational speed and higher dwell time, 2300 and 3500 rpm, the PC workpiece experiences severe friction with the rotating tool, leading to a tip formation at the interface region, as shown in Fig. 14. Additionally, the figure shows that there is no any softened or pasty material found during the weld of HDPE to PC at 3500 rpm, and the formed joint is due to the presence of the tip in the PC workpiece, which interlocks with the upper HDPE workpiece without any bonding.

The tip was only formed at 2300 rpm and 3500 rpm. The load-bearing of the interlocking tip is of very low order, leading to a very low weld strength.