Bonding metals at elevated temperature

(1) Bonding by elastic adhesive contact: This community here is perhaps more familiar with bonding elastic solids by adhesive contact, such as gecko adhesion, or more generally rough surface adhesion. But what about bonding two pieces of metals? 

(2) Counter diffusion for bonding: High school physics may teach us that when perfectly polished, metallic sufaces bond naturally. Or the small gaps may be closed by diffusion of elements in the perpendular direction to the interface. The former is too ideal to be applicable to reality. The latter is not correct, in our view. 

(3) Interfacial diffusion for bonding: If the small gaps or pores or cavities are closed by diffusion, the diffusion is not perpendicular. Instead, it is the lateral diffusion flux (or called interfacial diffusion) that drives the closure. This is similar to the Hull-Rimmer mechanism. 

(4) So what governs solid-state bonding at high temperature: In this paper of ours (https://www.sciencedirect.com/science/article/pii/S235243162030078X?via%3Dihub), we humbly propose our view on this important issue (particularly useful for welding/joining industry). We suggest that bonding is by closing the interfacial cavities, which can be accomplished by the Hull-Rimmer mechanism (or more refined Chung-Rice model), and more importantly by creep deformation. The comparison between the diffusional Hull-Rimmer mechanism and the creep-induced mechanism dictates a length scale, as Neeldman-Rice model suggests for their high temperature fracture model. In most solid-state-bonding experiments (again welding and joining industry), this length scale is too small, so that the dominant mechanism is creep-induced closure. 

This paper can be found in the above link, or in the attachment below. We sincerely welcome your comments and critics.