It follows from the relevant litersature sources that the dominant wear mechanism in FSW will be adhesion and diffusion of FSWed alloy to the FSW tool made of HSS of W+rare earth metal oxides. However the shape of the FSW tool has not so big effect on the stirring efficiency because of strong adhesion of aluminum alloy.

there are some difference between the dissimilar welding of Al alloy to Mg alloy and the similar welding of these alloys. for example because of difference between mechanical properties of 7075 and az31, the tool is under different condition in advancing side and retreating side ( different of strain, force, strength of two materials)

my question is about dominant mechanism of tool failure in dissimilar fsw.

In any case adhesive wear will be the dominant mechanism unless some of the materials contains abrasive particles. The difference between mechanical strength of materials in dissimilar welding is not so great to interfere with because of plasticizing the hardest one. In other words the higher melting point material has to be plasticized enough to be welded with the softer one and in this case no mechanism other than adhesive will be important. I can not imagine a situation when the hardest material is worn by oxidative, fatigue or abrasive mechanism.

I have a SEM image of worn steel FSW tool used to welding aluminum alloy sheets. Look at the wear particles formation by grain boundary diffusion

Oh, sorry here it is

I really appreciate for your attention. Its helpful. can I ask about SEM image?

the light region is Al and the gray region is steel?

I did not understand correctly.

Thank you. The light region is quenched steel of FSW tool. Large gray area is aluminum alloy stuck. Thin gray spike layer is the FeAl3 intermetallic compound. This is called reaction diffusion when iron atoms diffuse into aluminum alloy and form the brittle intermetallics. You can see here how diffusion is by grain boundaries and then large wear particles pull out of the FSW tool by brittle fracture mechanism.