I want to triger KH instability in the shear layer spatially, I mean I want the perturbation grows with the distance to inlet. How should do this? I think off two methods, I don't know whether they are feasible? Can someone give some advice? 1)impose a sinusoidal in the shear layer at the inlet in the transverse direction. for example: at inlet x=0, impose v=0.003*sin(omega*t)at the interface. 2)A fellow student suggests me add the perturbation in the whole shear layer with the mathematical form v=0.003*sin(omega*t-k*x), not just the inlet.But in this case, the position of the interface is changed persistently, how to track the interface and add the perturbation to it becomes a problem. Anyone can help me? My ultimate aim is to model the perturbation growing process spatially.

Best Regards!

I presume that you are unable to trigger the K-H instability without a perturbation i.e. if you do not see the vortical structures based on a simple (constant or parabolic) inlet conditions.

Try this first

1) Refine the grid at the interface. That might help. 2) Try increasing Reynolds number i.e. reduce molecular viscosity (if it is not an issue) 3) If 1 and 2 do not work, try the sinusoidal function only at the inlet (as you suggested)