Since the 70’s it is known that superconducting magnets do not usually reach the design current at first powering: numerous quenches are needed to reach the ultimate current value. This phenomenon is the so-called training. It is general opinion that the training is due to mechanical origins, i. e., to successive settings of the coil in a more stable position under the Lorentz forces: the coil slips and sticks in a new position after each quench. In this work we give a numerical evidence of slip stick motion in the training data of the LHC dipoles obtained during the 5-year-long production.
The presentation starts with a reminder about the post-processing of the training data (SM18 test at 1.9 K in virgin conditions) of the LHC main magnets (dipoles and quadrupoles). Then, we present a simplified model of slip-stick motion for superconducting magnets based on the equilibrium between energy margin and frictional energy. The LHC main magnet training curve fits with the model over more than 60% of the training range, thus showing a strong quantitative evidence of slip-stick motion. A mathematical generalization of the model allows fitting the whole training curve!