Thanks for your answer.

I understand your point about the solver, and your proposal for fixing the problem using imposed displacements is quite ingenious. However, how do I know the importance of the displacements and where to put them? I fear the results I would get wouldn't be rigurous enough. 

isn't there another way of modelling the pontoons bridge behaviour that wouldn't require that? I just need to have central pressure on the vertical axis (passage of vehicles) and hydrostatic pressure counteracting, while the perimeter is not fixed (because in reality it's the case).

Best regards,