Use of the function computeRBGGE

[michael_borland]

Ryutaro,

The xCenter and yCenter values are automatically computed by computeRBGGE to correspond to the center of the field data from the four planes. This is picked up automatically by BGGEXP, so you don't need to do anything. The expansion is done about the midpoint between the planes.

It is a good idea to center the x planes around the beam trajectory, but there will always be some difficulty here for dipoles. The fields have to extend well outside the magnet, until they fall close to zero. That means that the trajectory on one side is much further off axis. In the end, we need to rely on the "guaranteed" quality of the field expansion inside the bounding planes.

We've seen the issue you speak of wherein the expansion uses very few terms. In the latest version of computeRBGGE, there's a new option -varyDerivatives that seems to help by keeping the expansion order constant as the multipole order is varied. It's also helpful to ensure that the vertical extent of the fields is as large as possible, provide it is not too close to an iron surface. Another possible issue is if the 3D magnetic fields are not fully converged, so that the fields appear noisy; this can make it difficult for the program to find a good "fit" because the fields don't adequately conform to Maxwell's equations.

--Michael