[CO] To go further with losses

[BonneelP]

Hello all,

I'm about to merge PR #558, I'm opening this issue to keep track of the improvements that were suggested during the review:

• rework LossModelMagnet.comp_loss (Wmag/Hmag) => Require to add a new clean method for Hole, can be tricky for HoleUD (maybe we can assume that the magnets are rectangular)
• Move all Temperature in OP => Avoid to have differences between elec, mag and losses temperature + enable adding temperature to OP_matrix
• Remove duplicated code to compute the skin effect: LossModelWinding.comp_coeff and LossModelJoule.comp_coeff are identical and almost matches Conductor.comp_skin_effect_resistance
• Add a variable to clear meshsolution after run

More information about each improvements are available in the discussion around the PR.
Best regards,
Pierre