*4.3. Discussion on Disassembly Method and Its Limitations*

In this study, it was observed that disassembly temperature, controlled via resistance heating, has an impact on the shear stress of ultrasonically welded joints and their HAZ. This section compares the mechanical behavior, microstructure and HAZ of specimens disassembled at every temperature. The lap shear stress was strongly influenced by the disassembly temperature with a drop up to 74% at 110 ◦C, 93% at 130 ◦C and 94% at 150 ◦C, compared to room temperature (Figure 9). Unlike fracture surfaces at room temperature, as shown in Figure 11, disassembled joints displayed a less uniform surface, indicating the interface reached the melting point, affecting both the nanocomposite film at the interface and the adherends' upper plies. The MWCNT/PP films melted at the bond line, with the most visible examples marked by the red circled areas in Figure 11b,c.

This non-uniformity was confirmed through SEM micrographs, where significant matrix softening and drawing was noted (Figure 12), exhibiting ductile failure. The melting of the nanocomposite film was mostly observed at higher temperatures (Figure 12c,d), with the presence of porosity under temperature and strain increase [55]. Similar matrix drawing was observed for carbon fiber (CF)/PPS joints tested at temperatures above Tg, 120 ◦C and 150 ◦C [56,57]. The joints were manufactured through ultrasonic or resistance welding. In the former case, substantial matrix drawing and ductile fracture was confirmed through SEM micrographs, leading to a decrease in lap shear strength.

**Figure 11.** Representative photographic images of welded GF/PP adherends after disassembly procedure: (**a**) 15 wt.% MWCNT/PP film; (**b**) 20 wt.% MWCNT/PP film, and; (**c**) 25 wt.% MWCNT/PP film. The dashed red lines show the location of the overlap for the upper adherend. The circled areas indicate melted nanocomposite films. The arrow in (**c**) shows the location of a crack in the GF/PP adherend, damaged during the disassembly process. Room temperature images reproduced with permission from [36].

**Figure 12.** Representative fracture surfaces and SEM micrographs of samples welded with 20 wt.% MWCNT/PP films after disassembly process: (**a**) comparison with room temperature fracture surface images, reproduced and modified from [36] with permission; (**b**) disassembly at 110 ◦C; (**c**) disassembly at 130 ◦C; (**d**) disassembly at 150 ◦C. All scale bars are 100 µm.

**Figure 13.** Estimated heat-affected zone area in GF/PP adherends after disassembly procedure at difference surface temperatures, based on adherends' surface color images presented in Figure 11. Two examples of delineated areas are shown in insets.

The HAZ reported in Figure 11, then quantified in Figure 13, is consistent with the temperature curves in Figure 10b,c, where a temperature above 150 ◦C was reached on the adherend's surface during the disassembly phase. Nonetheless, the experiments confirm resistance heating can facilitate disassembly of ultrasonically welded TPC joints through a manual process, especially at higher weight fractions (20 wt.% and 25 wt.% MWCNT) and surface temperatures (130 ◦C and 150 ◦C). Under the parameters investigated in this study, welds disassembled at a surface temperature of 130 ◦C with 20 wt.% MWCNT present the best balance between required shear stress and heat-affected zone.

Given the results presented in this study, a discussion on the limitations of this disassembly method and future work is warranted. It was demonstrated that resistance heating through an electrically conductive nanocomposite film at the welded interface can facilitate joint disassembly by lowering the required shear stress by more than 90%. However, as the process is relatively slow (<120 s heat-up phase) and the total interface/adherends thickness is low (<4 mm), the heat-affected zone extended through the thickness, mostly at higher temperatures (130 ◦C and 150 ◦C). Consequently, disassembly was not uniquely concentrated at the bond line where the MWCNT/PP film was placed, but affected the GF/PP adherends at the overlap as well. Thus, the method might be better suited for recycling at end-of-life or reuse of components by cutting off the damaged overlap section.

Finally, as the interface was structurally compromised during disassembly, it partially affected the efficiency of resistance heating. It is expected that a faster cross-head speed during disassembly, use of highest MWCNT weight fractions (such as 20 wt.% or 25 wt.%) and control of the applied voltage during the process could mitigate this limitation, as well as the extent of the HAZ. A faster cross-head speed would reduce the time between the beginning of the disassembly phase and the peak in the stress curves (Figure 10), as well as the overall duration of the disassembly phase. Therefore, the temperature when failure initiates and propagates at maximum stress would be lower, potentially limiting the HAZ in the adherends. Further, if failure were to occur at a faster rate, the interface might not have time to cool down due to lower heating efficiency. Some future research directions include (1) investigation of disassembly parameters (e.g., crosshead speed, voltage regulation through constant power output [39]); (2) use of thicker adherends to investigate HAZ; (3) healing of bond-line defects/damage through resistance heating.
