We have performed a number of simulations for the case in the question. The goal is the drag Force(X) in our settings.

The fact is that this is a fairly simple external simulation case. Fluid is water, the hull is stream lined, no appendages have been included .......

Conditions are the same and only the initial mesh is changed. Initial mesh level changes from 4 to 8, and a new simulation is started for each case.

We get no consistent data. I mean each case study gives a different result that shows no relationship or sign of convergence.

May I have any clue about what could be happening and how to solve it ?

I include a results table.

I would recommend that you check how many cells your model actually has, rather than just relying on the automatic meshing level. After that, maybe trying using manual meshing techniques to refine your model even further and see how that affects your results.