Traditional homogenization surface treatment overlooks the morphological mechanical matching requirements between hard and soft metals, thereby limiting the potential for interfacial synergy in laminated composites. This study proposes a differentiated surface pretreatment strategy. Sharp periodic grooves are constructed on the hard side, while a relatively smooth surface is maintained on the soft side. This structure, characterized as a hard tooth/soft matrix interface, achieves synergistic enhancement of interfacial properties. Copper and aluminum were used as model materials. Periodic grooves with a maximum profile height ( R t ) of 28.4 μm were fabricated on the copper side, and a smooth surface with an R t of 7.9 μm was prepared on the aluminum side. This combination yielded an interfacial bonding strength of 83.5 MPa, which is 21.89% to 39.63% higher than that of uniformly treated combinations. Furthermore, this configuration achieved the highest interfacial hardness gradient (54.1 HV) and the widest interdiffusion layer (~4.46 μm). The thickness of the interdiffusion layer increased by 7.99% to 17.06% compared to uniform treatments. Localized severe plastic deformation, induced by the enhanced mechanical interlocking from the hard tooth/soft matrix structure, led to the formation of a grain refinement layer near the interface. Grain sizes were refined from ~30 μm in the matrix to less than 2 μm, accompanied by the accumulation of high-density dislocations. These defect-rich regions served as “short-circuit” paths for atomic diffusion, significantly accelerating atomic migration. This study provides a new approach and paradigm for precisely controlling the interface structure of composites through material property-based differentiated surface morphology design. • Differential pretreatment creates "hard-tooth/soft-matrix" interface for Cu/Al composites. • Optimal combination achieves 83.5 MPa bonding strength, 21.9-39.6% higher than uniform treatments. • Mechanical interlock-induced defects serve as short-circuit paths for atomic diffusion.
Tian et al. (Wed,) studied this question.