This is most likely caused by the roughness of the coating and not by the chemistry.  The two components probably do not mix, so the 50% mixture may well be the roughest.

Thank you for you answer , but what do you mean by the two components probably do not mix, so the 50% mixture may well be the roughest ? Is that mean that 50% has better mixture ?  

Surface roughness is a key but you also need to identify the surface composition to answer; is it a physical mix between FA and Ti or a new phase has been produced.

Also it depends on the temperature of processing. 

I suggest XRD  to identify composition and SEM to see the surface morphology.