Add nested tube array geometry (optical fibers)

[WilfS]

Optical fibers (such as wavelength shifting fibers) are not explicitly defined in ratpac, despite their prevalence in physics experiments. One could create such a fiber in ratpac by constructing several concentric tubes with relevant materials, but when simulating experiments with thousands of fibers this can lead to undesirably large geometry files.

To address this issue, this update will add the nested tube array geometry, that allows construction of an array of three concentric tubes (a core, an inner tube and an outer tube). The length, position, and direction of each nested tube can be set individually. Materials and radii are set for the whole array.

Validation has been performed to ensure the nested tube arrays are correctly placed in the geometry and have the same optical properties as concentric tubes manually placed in the geometry. A validation macro and geometry file have been added that demonstrate how the nested tube arrays are placed.

The rat DS has been updated to include nested tube array info such as direction, length and material. This is similar to the PMTInfo class.

No extra materials have been added that would be useful when using nested tube arrays as optical fibers. We hope to add this later when we have more accurate properties.

The documentation has been updated to include the nested tube array geometry.

[JamesJieranShen]

This looks mostly ok with me on the geom construction perspective. I will wait on comments from others before merging.

[JamesJieranShen]

Argh.. please fix the formatting as well.

[WilfS]

Just pushed changes that are clang-formatted. Sorry, I missed that requirement.

[WilfS]

I've just added some commits relating to adding the ability to make the core of the nested tubes a sensitive detector. This is particularly useful when using the nested tubes as optical fibers. The hit information from the sensitive detector gets saved to the rat DS every event and can be accessed by processors.

[MarcFBergevin]

I have successfully compiled the PR with no issues. However when I run the following macro:

rat --vis macros/vis_nestedTubeArrays.mac -o ~/nested -l ~/nested.l

I obtain the following error:

GeoNestedTubeArrayFactory: Constructing volume fibers DB: Field GEO[fibers].outer_material lookup failure. Does not exist or has wrong type.

[MarcFBergevin]

I have switched values from material_outer to outer_material in the geo file and I was able to get an output (also for inner and core). I am not too certain I understand the following output:

Is this the same output on your end?

[WilfS]

Apologies i forgot to update the macro.

Yes that output is expected. It's demonstrating the construction of fibres. The different colors are for the different construction methods. One with the old method of defining individual tubes and the other with the new array method.

[MarcFBergevin]

Thank you, that makes sense!

I did a quick test to test out the features with for four fibers. The visual output looks very nice. However, I am having difficulties accessing time and hit location in the outroot file (The error might be on my side). Have you cross-checked that hits are recorded for a test case scenario?

[WilfS]

The fibers are sensitive detectors. Any hits that occur in them are saved as a hit collection in the ratds. I haven't added them to the outroot file, but they're easily accessed in processors so could be added

[MarcFBergevin]

Apologies, I should have made a clearer statement. I did notice that the ntuple root files were not filled out and I was going to propose to you to add that in (potentially as an optional output to the ntuple root files). I was referring to the /rat/proc outroot (instead of /rat/proc outntuple files) which records the ratds in a root format.

By modifying the MC you essentially have already filled out the outroot root files (or ds files). However when I query via the TBrowser I get empty values:

~71000 photon bombs event where generated in the scintillator material with 4 fibers for my test. I can forward the macro to you. The setup looked like this.

I expected to see some hits, but could not confirm any hits in the ratdb root files.

[MarcFBergevin]

I have tried to add the sensitive detector option:

{ name: "GEO", index: "fibers", enable: 1, valid_begin: [0, 0], valid_end: [0, 0], mother: "outer_tank", type: "nestedtubearray", core_r: 0.47, inner_r: 0.485, outer_r: 0.5, pos_table: "cable_pos", orientation: "manual", outer_material: "aluminum", inner_material: "glass", core_material: "mirror", sensitive_detector: "/mydet/fibers", #drawstyle: "solid", color: [0.8,0.0,0.0,0.8] }

However, I am still not seeing any registered hits at this point.

[WilfS]

Hmm. I've never looked at the outroot output, but I'm able to extract the fibre hits from the ratds when I use a custom processor, so it should be added to the outroot output... Looking into this now

[MarcFBergevin]

It would be nice to add an option so that the output could be save optionally to the outntuple files `src/io.

Maybe something like /rat/procset include_fibers 1 which by default would be 0. And then the macro could look like:

#/rat/proc outroot
/rat/proc outntuple
/rat/procset include_tracking 0
/rat/procset include_mcparticles 1
/rat/procset include_pmthits 0
/rat/procset include_fibers 1
/rat/procset include_untriggered_events 1

[WilfS]

We've changed the example macro to use some example WLS materials that we've added to ratdb. With these materials photons will actually get absorbed in the core of the nested tubes and be recorded by the sensitive detectors.

Looking at the outroot file, I see entries for the GetMCNestedTubeHitCount() method, so the hits are getting saved to the ratpac DS. Yet, if I look at the hit information using GetHitTime, I get errors saying

Error in <TTreeFormula::DefinedVariable>: Unknown method:photon.GetHitTime() in vector<RAT::DS::MCNestedTubeHit> 
Error in <TTreeFormula::Compile>:  Bad numerical expression : "ds.mc.nt.photon.GetHitTime()"

and if I try using the hitTime variable it says

Error in <TTreeFormula::DefinedVariable>: hitTime is not a datamember of vector<RAT::DS::MCNestedTubeHit>
Error in <TTreeFormula::Compile>:  Bad numerical expression : "ds.mc.nt.photon.hitTime"

I may just have things setup wrong, as I get the same errors when I try to retrieve hit information from the MCPMT class.

[JamesJieranShen]

We've changed the example macro to use some example WLS materials that we've added to ratdb. With these materials photons will actually get absorbed in the core of the nested tubes and be recorded by the sensitive detectors.

Looking at the outroot file, I see entries for the GetMCNestedTubeHitCount() method, so the hits are getting saved to the ratpac DS. Yet, if I look at the hit information using GetHitTime, I get errors saying

Error in <TTreeFormula::DefinedVariable>: Unknown method:photon.GetHitTime() in vector<RAT::DS::MCNestedTubeHit> 
Error in <TTreeFormula::Compile>:  Bad numerical expression : "ds.mc.nt.photon.GetHitTime()"

and if I try using the hitTime variable it says

Error in <TTreeFormula::DefinedVariable>: hitTime is not a datamember of vector<RAT::DS::MCNestedTubeHit>
Error in <TTreeFormula::Compile>:  Bad numerical expression : "ds.mc.nt.photon.hitTime"

I may just have things setup wrong, as I get the same errors when I try to retrieve hit information from the MCPMT class.

Hey Wilf, you might need vectored versions of your data structures added to LinkDef.hh. Something like #pragma link C++ class vector < RAT::DS::MCNestedTubeHit>;. I think this will let cint (therefore ROOT) know about how to deal with vectors of the new DS objects.

[WilfS]

Hi James. I believe I'm already doing that. And if I had forgotten to do it it wouldn't explain why the same thing is happening for the MCPhoton info. I must be doing something wrong with my root setup such that cint isn't getting the info from linkdef.

If someone else could test to confirm this is a "Wilf problem" rather than a general problem, that would help.

[WilfS]

I've just added the requested feature from Marc where the nested tube hits are written to the ntuple output if requested.

If /rat/procset include_nestedtubehits is included in the macro, then nested tube info will be saved to the meta TTree and nested tube hit info will be saved in the output TTree.