Hi,
It is “possible” in theory, however there are far better workflows available.
In short, you would have to use a table input for the velocity, changing the V_y after a number of iterations. And here is where the problem lies: you would have to run enough iterations for the simulation to converge for each one of the V_y of interest.
Just as an example, using random numbers: you run 1000 iterations to converge for V_y= 1 m/s, then you change V_y to 2 m/s and run an additional 1000 iterations to allow the simulation to converge, and so on.
In short, this is not a good option. A better option would be to set multiple runs in parallel to run, one for each value of interest for V_y. This would require less time to run, and would be a more straightforward approach.
Cheers