Project

General

Profile

Definitions

  • APA -- Anode Plane Assembly. Contains readout wires in three planes, U, V, and Collection.
  • TpcSet -- In the dunetpc code, an APA is referred to as a TPC set, so that numbering between installation and offline numbering
    can be less confusing. This wiki refers to them as APA's in many places.
  • TPC -- The volume of liquid argon read out on one side of one APA. In the far detector geometry, the two TPC's read out by a TpcSet are numbered 2*TpcSet, and 2*TpcSet+1
  • CPA -- Cathode Plane Assembly. Flat panels made out of resistive material
  • FC -- Field Cage
  • PD -- Photon Detectors
  • FD -- Far Detector
  • Module -- One 10 kt (fiducial) piece of the Far Detector. Four are planned

More information can be had in the DUNE CDR, Volume 4, on Detectors. https://arxiv.org/abs/1601.02984

DUNE Geometries

There are two different 35t prototype geometries and several FD-relevant geometries in the repository, accessible through the dune35t_geo and dune10kt_geo art parameter sets, respectively. In short, the different files are:

  • dune35t.gdml -- The old 3-APA design for the 35t prototype
  • dune35t4apa.gdml -- The new 4-APA design for the 35t prototype
  • dune10kt.gdml -- standard FD design; 2 Cryostats side-by-side, 3x2x10 APAs in each, at the surface
  • dune34kt.gdml -- larger design; 2 Cryostats end-to-end, 3x2x18 APAs in each with a longer drift
  • dune10kt_APAoutside.gdml -- different FD design proposal fitting an extra anode plane in the same volume
  • dune4apa36deg.gdml -- FD-sized apas, but only 4 of them in a single cryostat with minimally sized world volume
  • dune4apa45deg.gdml -- Same as above but with U/V wires at about 45deg instead of 36deg.
  • protodune_v3.gdml -- Current (as of April 2017) ProtoDUNE-SP Geometry

DUNE 10kt v1 geometries

  • dune10kt_v1.gdml
  • dune10kt_v1_nowires.gdml
  • dune10kt_v1_workspace.gdml
  • dune10kt_v1_workspace_nowires.gdml
  • dune10kt_v1_1x2x6.gdml
  • dune10kt_v1_1x2x6_nowires.gdml
  • dune10kt_v1_3mmpitch.gdml
  • dune10kt_v1_3mmpitch_1x2x6.gdml
  • dune10kt_v1_3mmpitch_1x2x6_nowires.gdml
  • dune10kt_v1_3mmpitch_nowires.gdml
  • dune10kt_v1_3mmpitch_workspace.gdml
  • dune10kt_v1_3mmpitch_workspace_nowires.gdml
  • dune10kt_v1_45deg.gdml
  • dune10kt_v1_45deg_1x2x6.gdml
  • dune10kt_v1_45deg_1x2x6_nowires.gdml
  • dune10kt_v1_45deg_nowires.gdml
  • dune10kt_v1_45deg_workspace.gdml
  • dune10kt_v1_45deg_workspace_nowires.gdml

Tips on how to regenerate these

Selecting Specific Geometry Files

LArSoft has 2 DUNE geometry configurations throughout the all of the fcl files, defined in Geometry/geometry_dune.fcl. Here, parameter sets dunefd_geo and dune35t_geo are defined with a specific gdml file path, a geometry name, a distance from the surface, and a switch to use a file with no wires in Geant4. While the name should be left the same, the path to the gdml file can be changed to select from several options. For example, the code below has selected not the full geometry, but a mini-version for use in day-to-day software development. Note that the name doesn't really need to change unless the channel map, or some other detector ID specific code, cannot be general to kDUNE10kt and the needed functionality at the same time. So far, however, all of the dunefd_geo geometry files work exactly the same way in LArSoft.

dunefd_geo:
{
 Name:     "dune10kt" 

 # Choose GDML file
 #GDML:     "Geometry/gdml/dune10kt.gdml" 
 #ROOT:     "Geometry/gdml/dune10kt.gdml" 
 #GDML:     "Geometry/gdml/dune10kt_APAoutside.gdml" 
 #ROOT:     "Geometry/gdml/dune10kt_APAoutside.gdml" 
 GDML:     "Geometry/gdml/dune4apa36deg.gdml"   # *45deg.gdml or *36deg.gdml
 ROOT:     "Geometry/gdml/dune4apa36deg.gdml" 

 SurfaceY:           0.0e2               #in cm, vertical distance to the surface
 DisableWiresInG4:   true
}

Aiming Events In A Geometry

The GeometryTest module has a method named printVolBounds which is commented out by default. Uncomment this and run GeometryTest on whatever geometry you want to generate in -- this will print out cryostat and TPC boundaries in world coordinates. Note that the wire planes are slightly smaller in y-z than the TPC volume, though the TPCActive y-z boundaries are equal to the its corresponding wire plane y-z boundaries.

35t Prototype Geometry

The dune35t.gdml file is old and should only be used to look at files generated before the design change. The dune35t4apa.gdml file describes the new 35t geometry with 4 slimmer APAs; this is drawn in the image below without any wires. Drawn are the APA frames (dark grey), CPAs (burnt orange), vertical optical detectors (blue/purple), and 8 imaginary active volume bounds (light grey). The coordinate system is also overlaid in red, with the origin placed as follows:

  • x=0 is the +x face of the first wire plane (U, there is no grid plane in gdml) in the largest drift volume
  • y=0 halves the space between the top and bottom center APAs
  • z=0 is the front face of the active volume

The detector is oriented in the hall like so:

 Bird's eye view of trench, walkway is on the west side
    |          |
    |          |<-- trench wall
    |          |
    |   ___N__ |
    |  | ____ ||  <- short drift
    |  |      ||   <- long drift
    | W|______||E
    |  |______||     +z <---|  (+y out of screen)
    |      ^   |            |
    |     Neck |            v
    |          |           +x
    |          |
    |          | <-- PC4 entrance

Locations of the muon counters in the 35t

For a description as to where the counters are located in the 35t have a look at this wiki page.
https://cdcvs.fnal.gov/redmine/projects/35ton/wiki/TSU_Counter_Locations

Far Detector Geometry

The standard full file for Far Detector geometry is dune10kt or dune34kt. At the time of writing, the only differences are that 34kt has a longer drift distance, 18 sets of APAs adjacent in the z direction, and cryostats end-to-end instead of side-by-side. Another proposed geometry, with 4 planes of APAs per cryostat has also been checked in with the name dune10kt_APAoutside.gdml. In all of these geometries, the origin is placed as follows:

  • x=0 the center of the cryostat for dune34kt, and in the middle of the 2 cryostats in dune10kt and dune10kt_APAoutside
  • y=0 halves the space between the top and bottom APAs
  • z=0 is the front face of the active volume.

Here are images of the 34kt, 10kt, and 10kt_APAoutside geometries, respectively. Drawn are many CPAs (burnt orange), optical detectors (blue/purple), the cryostat shells (dark grey), and the TPCActive volumes (light grey). Wires are not drawn, but APAs can be imagined to hold the drawn optical detectors.

34kt: 216 APAs, 288 CPAs, 432 volTPCs

10kt: 120 APAs, 160 CPAs, 240 volTPCs -- note that this is also available with U/V wires at 36 degrees from vertical as well (dune10kt36.gdml)

lbne10kt_APAoutside.gdml: 160 APAs, 120-200 CPAs (??), 320 volTPCs, though the outer ones would be treated differently.

dune10kt_v1.gdml: 150 APAs, 120-200 CPAs (??), 300 volTPCs, though the outer ones (there are 100 of them) are treated differently. Electrons do not drift in the outer TPC's, though scintillation photons are created.

The far detector geometry's wire dumps are available here: DUNEFD10KTWireDumps

Far Detector Workspace

Events generated/reconstructed in the above FD geometry files, however, take significant amounts of time and memory when using LArSoft, potentially making day-to-day development using these files impractical. If working with this, be sure to try out the optimized version of LArSoft when setting up your environment . The size of a simulated root file will be scaled by the number of channels (and number of photons if running optical simulation), so a geometry with many APAs is going to be unnecessarily cumbersome. For trying out new algorithms and simulations, it would be most practical to start out in a geometry with as few APAs as possible that still had vertical and horizontal gaps in active drift volume -- so there is a geometry with 2 APAs high by 2 APAs long (beam direction). Also, given the continued discussion of altering wire angle to optimize disambiguation, and given the aspect ratio of the current FD APA design (2.5x7), this 4 APA geometry is available with wrapped wires at 36 degrees from vertical (dune4apa36deg). There are also 10 optical detectors per APA.

This is sufficient for development until we start to link up muon tracks across many APAs, most likely beginning with solid 3D information.

ProtoDUNE-SP Geometry

Details on the ProtoDUNE-SP detector are available in the Technical Design Report https://arxiv.org/abs/1706.07081

See also this wiki page for more details on the wire ordering:

https://wiki.dunescience.org/wiki/ProtoDUNE_geometry

The origin is located at x=0 at the cathode, y=0 on the bottom of the APA's, and z=0 on the upstream edge of the most upstream APA, so that z coordinates are all positive.

There are six APA's. In the offline geometry code, these are called "TPC Sets". The offline numbering is as follows

Plan view from the top.  TPC Set numbers are shown.  The beam enters from the left and travels to the right.

     1     3     5
   ===== ===== =====

   -----------------  CPA

   ===== ===== =====
     0     2     4

And here are the offline TPC numbers -- shown inside the TPC volumes.  The beam enters from the left.

     3     7    11
   ===== ===== =====

     2     6    10
   -----------------  CPA

     1     5     9
   ===== ===== =====
     0     4     8

TPC's 0, 3, 4, 7, 8, and 11 do not have a field which drifts electrons towards the 
APA's, though light flashes can be made there and particles depositing charge between
the wire planes will cause detected pulses on the wires.

ProtoDUNE-SP Wire Dumps

ProtoDUNE-SP Online nomenclature

The APA's are referred to by the DAQ group and online in their installation order. Here is a map of the APA numbers referred to by the online:

Plan view from the top.  APA numbers are shown.  The beam enters from the left and travels to the right.

     5     6     4
   ===== ===== =====

   -----------------  CPA

   ===== ===== =====
     3     2     1
TPC Plane Channel Range
0 0 0 - 799
0 1 800 - 1599
0 2 1600 - 2079
1 0 0 - 799
1 1 800 - 1599
1 2 2080 - 2559
2 0 2560 - 3359
2 1 3360 - 4159
2 2 4160 - 4639
3 0 2560 - 3359
3 1 3360 - 4159
3 2 4640 - 5119
4 0 5120 - 5919
4 1 5920 - 6719
4 2 6720 - 7199
5 0 5120 - 5919
5 1 5920 - 6719
5 2 7200 - 7679
6 0 7680 - 8479
6 1 8480 - 9279
6 2 9280 - 9759
7 0 7680 - 8479
7 1 8480 - 9279
7 2 9760 - 10239
8 0 10240 - 11039
8 1 11040 - 11839
8 2 11840 - 12319
9 0 10240 - 11039
9 1 11040 - 11839
9 2 12320 - 12799
10 0 12800 - 13599
10 1 13600 - 14399
10 2 14400 - 14879
11 0 12800 - 13599
11 1 13600 - 14399
11 2 14880 - 15359

Far Detector Workspace 1x2x6 Geometry

In order to contain events better than the previous workspace geometries that only had two APA's along the beam direction, a new workspace geometry has been made that has six APA's along the beam direction. There is one plane of APA's, two tall as in the Far Detector (12 meters), and six along the beam (13.9 meters in z). It is two APA's high, like the far detector. There are two cathode planes, situated one drift length away on either side of the APA's in x. The ordering of APA and TPC numbers advances with X (fastest), Y, and Z (slowest).

There are 12 APA's, and therefore 24 offline TPC numbers. The offline numbering is as follows

Plan view, from the top, of the lower 12 TPCs, shown inside the TPC volumes.  The beam enters from the left.

   ----------------- -----------------  CPA

     1     5     9    13    17    21
   ===== ===== ===== ===== ===== =====  APA
     0     4     8    12    16    20

   ------------------ ----------------  CPA

Plan view, from the top, of the upper 12 TPCs, shown inside the TPC volumes.  The beam enters from the left.

   ----------------- -----------------  CPA

     3     7    11    15    19    23
   ===== ===== ===== ===== ===== =====  APA
     2     6    10    14    18    22

   ------------------ ----------------  CPA

Here's a cutaway view of the lower half of the 1x2x6 geometry. Only the bottom APA's are shown along with the argon volumes they read out, as well as a cutaway of the cryostat. The origin is on the side. Y=0 between the lower APA's and the upper APA's. All the APA's have Z>0. X=0 in the plane of the APA's.

Iceberg

Documentation in progress. Iceberg consists of just one APA, which is half the size on top of a ProtoDUNE-SP APA. It has 10 FEMB's, five on each side. There are 480 total collection-plane wires, 400 total U-plane wires and 400 total V-plane wires, divided evenly among the two sides. A wire mapping spreadsheet is available in DUNE DocDB 11164. The cryostat layers and the world is borrowed from the 10kt Far Detector and thus does not represent Iceberg's situation. There was also an issue with the GDML-making Perl script and its interaction with the APA object sorter in the offline code that requires there to be at least one wire in each induction plane that goes from one vertical side to another without ending on the top or bottom edges. As an expedient, in order not to rewrite code, the length of the iceberg APA has been artificially extended in the offline geometry, making the wires wrap more than they really do in the actual APA. The top anchor points are in the right place, and the wires that are physically present are represented where they should be, they just extend further down in the simlation.

Here's an ascii art cartoon of how the channels are numbered, looking down from the top. X points along the drift direction.


   FEMB   709 704 716 726 706
                                           x
        200 .-----------> 399            
            |                              ^
        799 |  <--------.  600             |
            |           |                  | 
       1040 |  z2 ----> | 1279             '----> z
            |           |
            |           |
            |           |
        800 |  z1 ----> | 1039
            |           |                  
        400 |  v -------'  599
            |
        199 '---------- u    0   

   FEMB   703 727 715 723 721

Here is a table of the WIB and FEMB number on the wib assignments for each FEMB, as well as channel number ranges. Many thanks to Ivan Terrazas for verifying this in the data. This table refers to data up to and including Run 1332, taken on June 7, 2019. After this, two cables were swapped in WIBs 2 and 3. Table updated May 13, 2020 to match the notation in the JPG images in DUNE-Doc-11164-v3. In this table and in the JPGs in DUNE-Doc-11164-v3, WIBs are numbered from 1-3 and FEMBs are numbered from 0-3. The DAQ however reports "slot" numbers from 0-2 and "fiber" numbers from 1-4. So "slot"=WIB-1 and "fiber"=FEMB+1.

FEMB ID WIB FEMB on WIB DAQ slot DAQ fiber Low U High U Low V High V Low Col High Col
703 1 0 0 1 160 199 400 439 800 847
727 3 1 2 2 120 159 440 479 848 895
715 2 1 1 2 80 119 480 519 896 943
723 1 1 0 2 40 79 520 559 944 991
721 3 2 2 3 0 39 560 599 992 1039
706 1 3 0 4 360 399 600 639 1232 1279
726 2 3 1 4 320 359 640 679 1184 1231
716 3 3 2 4 280 319 680 719 1136 1183
704 1 2 0 3 240 279 720 759 1088 1135
709 2 2 1 3 200 239 760 799 1040 1087

WIB 2, FEMB #1 and WIB 3, FEMB #1 are "fake" FEMBs. The data for the "fake" FEMBs is a counting pattern produced by the WIB firmware. Because there are not channels in the geometry to associate with the fake FEMB's, these data are not copied from the raw fragments into raw::RawDigits.

Iceberg Wire Dumps