I've done as Mats suggested and it works. But, although the analysis goes well, in the sense it goes to the end, there is a trouble: since the area where I put the "displacement load" is not symmetric, the spring bends sideways.
How can I simulate the real/good movement of the spring?