Strange element positions RINGDEFINITION

I tried to run my simulation using the RINGDEFINITION and Sbend3D with OPAL/1.9.

Starting from the input file properly running with OPAL/1.6, I have

• update the version number
• replace DISTRIBUTION=FROMFILE with TYPE=FROMFILE

all the rest of the input file is unchanged.

I realized that there is a problem with the element position:

• OPAL1.6 (correct)

OPAL> * Added OFFSET to Ring
OPAL> * Start position (-1547.1, 9.47297e-14, 0) normal (6.12303e-17, 1, 0), phi 6.12303e-17
OPAL> * End position (-1547.1, 1800, 0) normal (6.12303e-17, 1, 0)
OPAL> * Added MAPS to Ring
OPAL> * Start position (-1547.1, 1800, 0) normal (6.12303e-17, 1, 0), phi 6.12303e-17
OPAL> * End position (1297.51, 2642.61, 0) normal (0.544639, -0.838671, 0)

• OPAL1.9

OPAL> * Added OFFSET to Ring
OPAL> * Start position (-1.5471e+06, 9.47297e-11, 0) normal (6.12303e-17, 1, 0), phi 6.12303e-17
OPAL> * End position (-1.5471e+06, 1800, 0) normal (6.12303e-17, 1, 0)
OPAL> * Added MAPS to Ring
OPAL> * Start position (-1.5471e+06, 1800, 0) normal (6.12303e-17, 1, 0), phi 6.12303e-17
OPAL> * End position (-1.54426e+06, 2642.61, 0) normal (0.544639, -0.838671, 0)

Are there modification also to RINGDEFINITION element between OPAL/1.6 and OPAL/1.9? I don't get any warning or error message.

Sorry, RingDefinition isn't a code I have worked with. It looks like there is a mm -> m convertion error somewhere since all values are factor 1000 bigger.

assigned to @ext-rogers_c and @adelmann and unassigned @kraus

unassigned @adelmann

Hi Valeria, we are trying to move the system of units of OPAL-CYCL to be metres not mm, degrees not radians. This was part of the move. Sorry if it caused confusion. The outputs should be in metres as well.

Thanks for the explanation. I will try to rerun with the correct units.

I made the units conversion in my simulation: mm -> m and rad -> deg

I have updated the units in the input file of the following elements: RINGDEFINITION, SBEND3D, LOCAL_CARTESIAN_OFFSET and the DumpFields function.

To get the same envelope as with OPAL/1.6, the end_position_x and end_position_y in the LCO have to be still in mm.

To perform this test I used the OPAL master compiled on 8th, November 2017. I don't know if the LCO units have been updated as well in the meantime.

How annoying. I think PROBE is the same. I will make a patch.

I just came across this issue. What is the status? Is the documentation for RINGDEFINITION in https://gitlab.psi.ch/OPAL/src/wikis/Manual/elements correct, or are some of the input units in meters?

@rizzoglio_v can you provide the input files please

Unfortunately I could not upload the field maps directly on gitlab because of size issue.

You can find the needed files on /afs/psi.ch/project/SCGantry/. In particular:

• input file: afs/psi.ch/project/SCGantry/RaceTrack/B1-LongDrift/MultiMaps/B1_MultiMaps.in
• field maps: /afs/psi.ch/project/SCGantry/nobackup/3D_Maps/Best/5Maps/
• particle distribution: /afs/psi.ch/project/SCGantry/nobackup/beam/Beam_185MeV.dat

The input file still refers to OPAL/1.6

@rizzoglio_v: Thanks.

From your comments (#195 (comment 4137)) and looking in the code RINGDEFINITION is in now in m. I changed the Wiki manual accordingly.

I think SBEND3D and LCO is still in mm (please correct me if wrong).

There is no documentation on the DUMPFIELDS command (I will make an issue).

As Chris mentioned above, PROBES units should be updated as well as CCOLLIMATOR (I guess).

At least these are the elements I used in my simulations with OPAL-Cyc.

Okay, I see. There are several other places where mm is used as input. Should these all become meter? And if so, what is the strategy for converting (which release / timeline, regression tests, etc.)? I propose to discuss this further in #45, and close this one (if all comments are understood).

Closing this as proposed in favour of #45.

closed

marked this issue as related to #45