intrabeam scattering with IBSCATTER

[tiansk]

Dear all,
I intend to track the electron bunch in storage ring with IBSCATTER element to simulate the IBS effect. I want to know that:How to bring in coupling(emit_y/emit_x) in the lattice ？I found the element SREFFECTS has a Parameter Name：COUPLING
if I set the COUPLING=0.01，is it that during the particle tracking,the parameter of electron bunch from WATCH element emit_y/emit_x keep 0.01? The attached files was the example I used for IBSCATTER.I think I should get the same result as

https://jacowfs.jlab.org/conf/y15/ipac1 ... PMA012.PDF
but the variation of the vertical emittance is strange,Is there any way I could do in the example?


Thanks in advance!
Tiansk

[tiansk]

Dear all,

The attached file is the result of particle tracking with IBSCATTER and SREFFECTS(COUPLING=0.01),the emittance of y direction is not growing so much! If using the ibsEmittance command,the final emittance_y/emittance_x will be equal with the coupling factor.Should I add some tilt bends or quads in the lattice？

Thanks in advance!
Tiansk

[simone.dimitri]

Dear all,
I am asking a very similar question of Tiansk: the agreement of ibsEmittance and tracking results on the horizontal emittance is fine. Instead, tracking does not reproduce the vertical emittance growth predicted by ibsEmittance. I also tried to include coupling sources in the lattice through quad roll errors, without success.
Any help to fix it? I am attaching elegant files and plots.
Thank you in advance,
Simone

p.s.: in order to enhance the IBS effect, the bunch charge was made huge, but this does not change the situation.

[tiansk]

Dear all:

When using IBSCATTER to simulate the IBS effect, there is a parameter called SMOOTH in IBSCATTER. I believe that setting SMOOTH to 1 or 0 may result in different paths for the variation of beam parameters, but the final equilibrium state should be relatively consistent. However, after comparing the results with different SMOOTH settings, I found that the beam equilibrium state parameters differ significantly, as shown in the attached files. How can this be explained?


Thank you in advance,
Tiansk

[tiansk]

The change in the vertical emittance is shown in the figure below.

[michael_borland]

Tiansk,

Could you provide your input files so I can check the results?

In general, I recommend using smooth=1 because otherwise you inject additional noise into the simulations.

--Michael

[tiansk]

Dear Michael,
Thank you for your reply.

After reviewing the IBS-calculation code, I noticed that setting smooth=0 modifies only xp, yp and dp. While this may be closer to the actual physical IBS process, it risks numerical issues from the program’s standpoint. The way the code transforms coordinates implicitly assumes α = 0 at the IBSCATTER element. Under that assumption the Twiss parameters behave as expected and the emittance grows at the predicted IBS rate. If α ≠ 0 (i.e. the beam distribution is tilted in phase space), the results could be wrong.
Is there a way to correct the calculation when α ≠ 0?

One possibility is to transform the particle coordinates so that the distribution becomes circular (as sketched in the attached figure). After this transformation the smooth=0 scheme can be applied to Px, Py, etc., and the coordinates can then be transformed back to the original (x, y, px, py) frame. However, I am not sure whether this procedure really makes sense, because it no longer corresponds to the physical picture of how IBS acts on the beam.

Thanks in advance!
Tiansk

[michael_borland]

Tiansk,

I don't think transforming to a round beam addresses the fundamental issue with the smooth=0 simulation, namely, that it attempts to model (x', y', delta) scattering from a section of a beamline with kicks at a particular location. The right way to address this is to accumulate a scattering matrix along the beamline. I've been thinking about how to implement this, but I think it may be very slow, equivalent to placing IBSCATTER elements more densely in the beamline. For now, if you want to use smooth=0, my suggestion is to put in more IBSCATTER elements, preferentially at locations with small beam size, but also so that the phase advance between IBSCATTERs is not too large (<<90 deg?).

On the other hand, because IBS growth rates in rings are moderate---comparable to the radiation damping rate---the smooth=1 approach seems reasonable.

I'm still curious to see your input files if you are able to share them.

--Michael

[tiansk]

Dear Michael,

I believe that adding IBSCATTER more densely does not solve the problem of smooth=1, but rather IBSCATTER should be placed at the position where alpha=0. Additionally, I would like to ask another question. Currently, the IBSCATTER simulation only performs slicing in the longitudinal direction and considers the density distribution. Why hasn't it been processed in the x and y directions? Is it unnecessary or more complicated to handle? Would it be feasible to perform a three-dimensional slicing simulation?


Regarding my input file, I have sent it to you via email.

Tiansk