Hi,
I noticed the divide_element changes the output results in some instances (i my case when tracking a particle distribution using a previously fitted lattice). I have attached a targz file that reproduces my problem. In brief, I do some fitting and in the last pass I track an incoming distribution in the fitted lattice.
If I run the ele file prob_divide.ele and then analyze the output distribution I found, e.g. for betax 7.675589825605547e+02 . If I now comment the divide_elements block starting line 389, run elegant again I find the output betax is changed to 2.484119249735075e+01 . See the readme file to follow the command I use.
I am not sure if this is a problem with elegant or an issue with the way I am running my file. I have tried to strip my file to a minimalist test case (though the ele file is still long). Thank you for any help. -- Philippe.
issue with divide_elements
Moderators: michael_borland, soliday
-
- Posts: 83
- Joined: 20 Aug 2008, 13:18
- Location: Northern Illinois University & Argonne National Laboratory
- Contact:
issue with divide_elements
- Attachments
-
- prob.tar.gz
- set of FAST file to check the divide elem issue
- (578.73 KiB) Downloaded 430 times
-
- Posts: 1945
- Joined: 19 May 2008, 09:33
- Location: Argonne National Laboratory
- Contact:
Re: issue with divide_elements
Philippe,
I found two things that contributed to the problem.
and the load_parameters command to
--Michael
I found two things that contributed to the problem.
- Your element-division command was acting on all elements, including RFCA elements. However, the RFCA elements have to be divided carefully otherwise the pieces will not be physical. Usually you should divide these into cells and no further, then use a PHASE_REFERENCE to ensure the simulation is correct.
- When you divide elements, you can't load the length values. Elegant isn't smart enough to know that the elements were divided when it is loading parameters. You'll end up with a beamline with the wrong length.
Code: Select all
÷_elements
name = *
exclude = CAV*
maximum_length = 0.01
&end
Code: Select all
&load_parameters
filename = injector.param,
change_defined_values = 1
allow_missing_parameters = 1
exclude_item_pattern = L
&end
-
- Posts: 83
- Joined: 20 Aug 2008, 13:18
- Location: Northern Illinois University & Argonne National Laboratory
- Contact:
Re: issue with divide_elements
MIchael,
Thank you very much for the suggestion! -- Philippe.
Thank you very much for the suggestion! -- Philippe.