Placement of IONEFFECTS elements

Moderators: cyao, michael_borland

Post Reply
Teresia
Posts: 30
Joined: 04 Oct 2018, 08:42

Placement of IONEFFECTS elements

Post by Teresia » 08 Jan 2021, 05:42

Hi,

I am trying to simulate ion instabilities for different fill patterns. Thank you so much for providing the ionEffects1 example and a lot of comments in the manual for how to it. Those are very helpful.

I however have a question about the reasoning behind where to place the IONEFFECTS elements. Looking at the example the lattice seems to have been divided in equal sections with IONEFFECTS in between. That makes sense, but is it of importance how this division is done? For example, that it is done such that the IONEFFECTS elements end up at positions in the lattice where we expect most ions to be trapped?

Best regards,

Teresia

jcalvey
Posts: 19
Joined: 08 Jan 2021, 12:20

Re: Placement of IONEFFECTS elements

Post by jcalvey » 08 Jan 2021, 12:49

Hi Teresia,

In general, the placement of the IONEFFECTS elements can be important.

By default, the ionEffects1 example uses the RING_ILM10 beamline, which divides the ring into 10 ILMATRIX elements, with ion IONEFFECTS elements in between. In order to put the elements at specific locations, you would need to use the RING beamline (in the run_setup command), which will do element-by-element tracking. You can then manually insert the IONEFFECTS elements wherever you want in the in the .lte file. Another option is to use the insert_elements command to automatically place them at certain intervals. For example, to place an IONEFFECTS element after every 10th quadrupole, you would use:

&insert_elements
type = "QUAD",
skip = 10,
element_def = "IE1: ionEffects"
&end

Let me know if you have any more questions.

-Joe

Teresia
Posts: 30
Joined: 04 Oct 2018, 08:42

Re: Placement of IONEFFECTS elements

Post by Teresia » 11 Jan 2021, 14:29

Hi Joe,

Thank you. Then I guess I need to do some tests and see which positions I need to get convergence of the results.

I also have another question now. I am trying to use the bigaussian fit and I haven't given any other values than the default. It seems to kind of work (I get histograms and fits), but I get the error "error: can't find valid initial simplex in simplexMin()". What does that mean?

Best regards,

Teresia

jcalvey
Posts: 19
Joined: 08 Jan 2021, 12:20

Re: Placement of IONEFFECTS elements

Post by jcalvey » 12 Jan 2021, 18:23

Hi Teresia,

That error just means that one of the attempted fits failed (the code tries multiple starting points in parameter space). If the final fits in the .ionHist file look ok, then you can ignore the error.

That being said, I would recommend that you start by using the gaussian kick method. It should be much faster than the bigaussian method, and has fewer numerical parameters to worry about. I've had success modeling various effects with the gaussian method, except for very strong instabilities.

-Joe

Teresia
Posts: 30
Joined: 04 Oct 2018, 08:42

Re: Placement of IONEFFECTS elements

Post by Teresia » 13 Jan 2021, 12:05

Hi,

Okay, thank you. Then I won't worry about that.

I also thought I better start with the gaussian method, but the results looked so strange that I thought perhaps it wasn't good enough. At the moment I get similar results independent of gap length. Even with a single very long gap I don't see any significant effect compared to having all buckets filled. I tried to increase the number of macroions, but that didn't make any significant difference either. So perhaps it's my placement of the ioneffects that is the problem. At the moment I'm only having two of them in the lattice for testing purposes, but perhaps that is way too little?

Best regards,

Teresia

jcalvey
Posts: 19
Joined: 08 Jan 2021, 12:20

Re: Placement of IONEFFECTS elements

Post by jcalvey » 15 Jan 2021, 17:27

Yes, I would recommend trying more ion interaction points.

If that doesn't work, you can send some plots of your results to me (jcalvey@anl.gov) and I'll see if I can help figure out what's wrong.

-Joe

Teresia
Posts: 30
Joined: 04 Oct 2018, 08:42

Re: Placement of IONEFFECTS elements

Post by Teresia » 22 Jan 2021, 10:21

Thank you. Then I will try the gaussian method with more interaction points and see if I manage to get something I think make sense. I will let you know how I get on. It is very nice of you to offer to look at the results :)

tiansk
Posts: 39
Joined: 16 Mar 2012, 22:01

Re: Placement of IONEFFECTS elements

Post by tiansk » 03 Nov 2021, 10:12

Teresia wrote:
11 Jan 2021, 14:29
Hi Joe,

Thank you. Then I guess I need to do some tests and see which positions I need to get convergence of the results.

I also have another question now. I am trying to use the bigaussian fit and I haven't given any other values than the default. It seems to kind of work (I get histograms and fits), but I get the error "error: can't find valid initial simplex in simplexMin()". What does that mean?

Best regards,

Teresia
At present, I have encountered the same error reporting as you in my simulation( bigaussian fit). Have you solved this problem?

Post Reply