Uniform corrector fields in CSBEND element?

[GPenn]

It would be convenient to include additional uniform dipole fields in bending elements such as CSBEND, to be used as correctors instead of affecting the nominal orbit. This is especially true for offset quadrupoles that are being labelled as CSBEND magnets (even though the CCBEND element would be a better fit in that case). If the bending kick winds up comparable to the orbit correction kick, that makes things complicated.

For a horizontal kick, this seems to be basically what FSE_DIPOLE does. But for a vertical kick, ETILT is modeling a physical rotation of the element and that model is not a good fit for corrector magnets. The alternative is to try to subtract out the effect of the shifted center of the multipoles, but this is just an approximation and doesn't work too well for long dipoles (sqrt(K)*L~1 or bigger). Or the dipole can be sliced and short kicker elements added. I am hoping it would be reasonably straightforward to implement these kicks intermittently at the same time as synchrotron radiation is applied, which would have the same effect as slicing the magnet but be a lot more manageable to implement in a lattice file. If this is something you would consider, I think one might as well have both vertical and horizontal fields done in the same way. This would be a smaller step compared to having the whole package of bending element + steering + steering multipoles.

On a slightly different topic, it appears to me that in CSBEND elements it's not possible to simultaneously use K1 and F1 to specify the quadrupole field, with the two terms adding together. Maybe this is done on purpose to keep the element description easier to read? I was thinking it would be nice to use K1 as the nominal value and stick errors in F1, but I can understand if the purpose of blocking that is to avoid confusion.

Thank you,
Gregg

[michael_borland]

Gregg,

You can combine Fn and Gn with Kn parameters. Perhaps you are thinking of the Bn parameters, which can only be used if USE_BN=1, in which case the Kn are ignored.

I like the suggestion of adding steering parameters to CSBEND and will have it in the next release.

--Michael

[GPenn]

Michael,

Having the steering available in CSBEND would help me a lot.

I was wrong about there being a problem combining Fn and Kn, I must have changed multiple things at once and blamed it on that particular aspect of the lattice when it didn't do what I expected.

Thanks,
Gregg