How to create a AGCCT magnet model

[1129759189]

Dear professor ,

‌ I was trying to build a Alternating gradient canted-cosine-theta (AGCCT) magnet model ，which are applied to implement combined functions of beam bending and focusing. I expect to build by code without a real magnetic field map. A confusing problem is that I don't know if the relevant code has been developed, and is it modeled by "multipole"in the manual? I am looking forward to your guidance. Thank you.

‌

‌

[Laurie Nevay]

Hello,

Can I confirm exactly which field components you need? You want a magnet that would be a torus shape (some part of one) with a constant radius of curvature; then multiple components on top of this including a dipolar field?

Does your field vary with curvilinear S along the magnet length?

Currently, the multipole component is only straight (i.e. no dipole component, only k1,k2,k3.. ). The field does not vary with S along its length.

If you don’t need variation in S, then we can add the angled part to this component. Otherwise, if it does vary with S, we should think about a new component for BDSIM (not too difficult) but we would need think of the best set of parameters to describe the field.

Cheers,
Laurie

[1129759189]

Thanks for your reply. I want to build a curved magnetic field with alternating gradients with S, which is actually a collection of dipole and quadruple or higher poles. For example, a 40°dipole provides a field of 2T to bend the beam, while along the bending path it has a quadrupole field gradient of 20T/m between 0°-20°, and then a quadrupole field gradient of -17 T/m between 20°-40°. I hope I have explained my intention clearly.

‌

[Laurie Nevay]

Hello,

Ok, I understand now. How would you specify the transitions between the field components? Do you have an equation or parameterisation that describes the field you see? Of course, I assume you can have a solution / field map from the winding of all of the coils, but is there an expression for the field you have with some multipole expansion? For constant in S, I can see this can easily be done with multipole components, but for variation along S there must be some other information / parameters that define how the strength changes in S.

We would also have to think about the end regions. Normally, we have hard edged components where the field abruptly stops. For certain elements, like a dipole, we have a thin ‘fringe’ element that represents the integral of this transition region outside the magnet. If the magnet is longer though we can specify the field as it tapers to 0.

I’m assuming it wouldn’t be acceptable to have a hard-edged section for 0 to 20 degrees with one set of multipoles, followed by (as a separate hard-edge component) another magnet with different multipoles at 20-40?

[1129759189]

Thanks for your professional help. I designed the magnetic field with the formula

The first term on the right side of the equation is the dipolar field component, the second term is the quadrupolar component, and the third term is the sextupolar magnetic field component. As you mentioned, magnet has SCOFF in a ideal case. An example of a AGCCT field distribution including both real filed and hard edge models is as follows. A magnet with a bend of 67.5 degrees includes both dipole and quadrupole components.

(cite: Design and optimization of beam optics for a superconducting gantry applied to proton therapy ,Physica Medica 73 (2020) 158–163 )
So I want to create a magnetic field distribution as shown in the figure by code.

[Laurie Nevay]

Please can you email me at laurie . nevay @ cern . ch (without spaces). This would require some additions to the code and we should discuss who will do this.