After some mechanical components have been found damaged as a consequence of the vessel sideways displacements, attention has been focused on the asymmetric behaviour of some disruptions.

Sideways vessel displacements are rare events: in the present magnetic configuration, they happen only during upward disruptions, if there is a kink, which locks long enough in the same toroidal position so that a net radial sideways impulse can build up. The radial sideways force is proportional to the product of the toroidal magnetic field with the difference of the vertical current moment in opposite cross sections. Experimental evidence shows that the impulse of the magnetic radial sideways force is proportional to the vessel sideways displacement and the upper envelope of the sideways displacements scales as the product of the plasma current with the toroidal field.

According to a kink model the plasma current should have a minimum where the plasma is closest to the wall; it is in disagreement with the measurements, but the presence of higher poloidal modes (m=2) could solve this discrepancy.

The force balance on the plasma is due to the eddy and the halo currents flowing in the vessel. Asymmetric eddy currents, which produce substantial asymmetric forces, have been computed with a FE model of the vessel.

The sideways force acting on the vessel in the worst asymmetric events is of the same order of magnitude of the vertical force. Even after the installation of a new lateral restraint system large vessel displacements have been observed (up to 7 mm). The cause of the small improvement gained from the new supports is in previously neglected damping action of toroidal field on the motion of the vessel.