Draft: Correct material in gibbet bars and orientation of TRD(1D) modules in the set-up

The changes in the v22c geometry are copied from the v21d one, because it was overlooked when producing the v21c geo that there is already a v22c one.

@d.emschermann Can you review the changes I made in the file?

added 2 commits

• 3d98e989 - 1 commit from branch computing:master
• 7c6cb23a - Correct material in gibbet bars and orientation of TRD(1D) modules in the set-up

Compare with previous version

@apuntke, @d.emschermann,

I canceled the pipeline since a rebase was needed.

resolved all threads

requested review from @d.emschermann

assigned to @f.uhlig

@apuntke The modifications are correct. However I do not understand why we try to patch the v22c here for the 2021 setup?!

because it was overlooked when producing the v21c geo that there is already a v22c one.

That was not overlooked!

We are patching the TRD v21b geometry for the mCBM 2021_07_surveyed setup. That TRD geometry will be v21d - to be updated in the 2021_07 setup. I do not see why a v22? geometry should end up in a 2021 setup, we do not backport into the past.

Independently the gibbet material patch should be applied to all TRD v22? geometries, not only to v22c.

Hello David,

I don´t know why the v22 setup tag is used in the mCBM 2021 setup. The change was introduced in CbmSoft/cbmroot_geometry@ed45ae5b by @p.-a.loizeau, maybe he can comment on it. From my side it would be no problem to use v21c or v21d in the setup file.

I changed the v22c geometry only because it is just one week old and was hopefully not already used for any simulations.

We simply followed the naming discussed in #2417 to avoid overwriting existing geometries. Not sure whom picked the name, so probably 22 was used to indicate it was generated in 2022, not to indicate it was assigned to a 2022 setup.

@d.emschermann @apuntke Are you sure that overwritting v22c is the correct approach. Hasn't v22c geometries been used for generation of presentation at last weeks simulation meeting? I would fear there may be a danger that this may lead to confusion.

@apuntke @p.-a.loizeau @e.clerkin As conclusion of the above comments, we should use TRD v21d in the 2021_07_surveyed setup, as soon as it is confirmed that the material is now correct in the gibbets. Who can check this, Eoin?

@d.emschermann,

the correctness of the materials you can check yourself.

Please run a simulation for the setup you are interested in

root -l 'mcbm_transport.C(1,"${setup}","data/${setup}_test","",kTRUE)'

and open the created geometry file

root -l 'data/${setup}_test.geo.root'

In the TBrowser navigate from the "Master Volume" to the node you want to check and right click on it. From the menu choose "InspectMaterial". If the result printed in the terminal is

Material dummy A=0 Z=0 rho=0 radlen=0 intlen=0 index=3

the material is still incorrect. If you see the expected material everything is fine.

I did a quick check with the macro from this MR and ther are no TRD support structure at all. Could you please check.

Sorry. This is wrong. By accident I looked in layer1 which doesn't have the support. Checking in layer3 the support structure has

Material KANYAProfile10x10Normal A=26.98 Z=13 rho=0.81 radlen=2.95836 intlen=12.9541 index=10

which is probably the expected material.

OK, this means that TRD v21d can be the new default for mCBM 2021_07_surveyed. @f.uhlig @axel

@f.uhlig and @d.emschermann

Oh my, we really need to double check each time. Although the gibbet bar materials are no longer defined as dummy which is good, please notice that the radiation length of the material Florian has posted is 10 times smaller than expected. I assume this is in error is in Florian's current setup. Please investigate and eliminate this error from happening again. You are welcome to use the rad_len checker script.

I did not post anything. I only checked if the material definition is okay. Nothing what I did locally was commited.

But you are right. In the creation macro one should add a line which loads the FairRunSim manager such that the units are set properly.

marked this merge request as draft

As another conclusion from Eoins comment I should also change this MR such that there are no changes made on v22c, but instead a new geometry v22d is made with the updates from v21d applied.

@apuntke I do not agree. v22c is of no use as that is not the TRD configuration in the mCBM 2022 setups. We should therefore not generate a v22d, as that would be a replica of v21d. We should simply delete the v22c, as it is of no use. Please speak up if you disagree.

Hello David, okay, I thought because of the naming of this parameter files and their usage in the 2021_07_surveyed setup they would correspond to the mCBM 2022 setup. I agree then to drop this MR. Sorry for the confusion.

@d.emschermann @apuntke

I would strongly suggest that a new tag is made for a geometry which has changed it's material budget. The dilemma is not, that a geometry is no longer of use, nor is it that it wasn't used for anything official, future users of the geometry may not have updated their geometry repos, may be using a old build, and continue to use a old version of the geometry well into the future without any reasonable way to check which version of the geometry it is.

The version tag is used to allow confirmation that a specific geometry, if we subverts this, version tag confirmation will become useless.

Non-creation of new tags were only done under special conditions which is not fulfilled here. It doesn't cost anything to use a new tag, I would strongly suggest that we do so here.

mentioned in merge request CbmSoft/cbmroot_geometry!151 (merged)

Dear @f.uhlig, @p.-a.loizeau,

you have been identified as code owner of at least one file which was changed with this merge request.

Please check the changes and approve them or request changes.

added CodeOwners label

@f.uhlig @apuntke This MR can be closed as TRD v22c will be deleted anyway.

closed

Could you please add the following lines her to create the file needed for the media check macros. The created file needs to be put in the directory "geometry_check" of the input repository

  // create medialist for this geometry
  TString createmedialist = gSystem->Getenv("VMCWORKDIR");
  createmedialist += "/macro/geometry/create_medialist.C";
  std::cout << "Loading macro " << createmedialist << std::endl;
  gROOT->LoadMacro(createmedialist);
  gROOT->ProcessLine("create_medialist(\"\", false)");

@apuntke,

sorry, the requested changes are not needed in this macro but in the macros to create the geometries trd_v21c_mcbm and trd_v21d_mcbm.

Could you please add the following lines here

// Load FairRunSim to ensure the correct unit system FairRunSim* sim = new FairRunSim();