ECAL status update. DUNE ND General Meeting Eldwan Brianne DESY - - PowerPoint PPT Presentation

ECAL status update.

Eldwan Brianne DESY Hamburg, 28th November 2018

DUNE ND General Meeting

New Geometry Design.

Eldwan Brianne | ECAL Status update | 28/11/2018 2 Page

Adapted to HPgTPC design

• Requirements:
• Contains the HPgTPC fiducial volume
• R = 2.7 m, L = 5.5 m
• Fits planar geometry of a high granular sampling calorimeter and matches cylindrical

design of the TPC and PV → octagonal shape

𝛏

• Rough dimensions
• Side length ~ 2.25 m
• Surface: ~ 100 m2 (Barrel), ~ 50 m2 (Endcap)
• Magnetic field
• Parallel to TPC axis (drift direction in x)
• Perpendicular to beam direction

New Geometry Design.

Eldwan Brianne | ECAL Status update | 28/11/2018 3 Page

Taking into account the physics

• Mostly physics is going forward (~ fixed target experiment)
• May not need same depth/granularity everywhere
• However, requires some understanding and optimisation
• First thought:
• Downstream region with High resolution/granularity
• Upstream region + Endcaps with Low resolution/granularity
• Variable longitudinal segmentation:
• Thin layers in front for good Eres for low-E photons
• Thicker layers in the back for containment while keeping

compact detector

Readout considerations.

Eldwan Brianne | ECAL Status update | 28/11/2018 4 Page

Need balance between granularity and cost

• High granularity readout planes
• → Individual scintillator tiles, one SiPM/tile
• Low granularity planes
• Layers of crossed strips readout on

both sides (strips have full length of segment ~ up to 2.5 m). Light guided with WLS

• Modules can be optimised for the strip

length

Calorimeter structure.

Eldwan Brianne | ECAL Status update | 28/11/2018 5 Page

A mix between high and low granularity

• Modules consist of a mix with a HG region and

LG granularity region

• HG layers interleaved with LG layers, need

some optimisation

• Typical EM shower see ~ 4 HG layers
• Transition region between HG/LG regions?
• Number of HG layers can differ for the DS/US

regions

• Use of different absorber thicknesses to enhance

the angular resolution / shower containment

A starting point.

Eldwan Brianne | ECAL Status update | 28/11/2018 6 Page

Considering world constraints

• Taking previous studies into account:
• Granularity is important in the first layers
• Thinner absorber (in the first layers) is beneficial for low-energy photons + give better lever arm on the shower direction
• Calorimeter depth relevant for higher energies (not to neglect)

DS Segments (3) US Segments (7) HG Layers (1 cm of Sc) 8 6 HG Tile size 2.5 x 2.5 cm2 2.5 x 2.5 cm2 HG Absorber thickness (Cu) 2 mm 2 mm Number of front layers 20 20 LG Layers 47 39 LG strip width (1 cm of Sc, crossed) 4 cm 4 cm LG Absorber thickness (Cu) 4 mm 4 mm Number of rear layers 35 25 Total thickness 15.6 X0 12.8 X0 Number of channels ~ 1.5 - 2 M (ignoring module widening, overlaps…)

First steps.

Eldwan Brianne | ECAL Status update | 28/11/2018 7 Page

Creating the new geometry

• Changed to new software (DD4hep) to simplify the optimisation

(https://github.com/ebrianne/DUNE_DD4hep_Geo)

• Global/local variables used to define the geometry
• Can be changed easily at run-time
• Several options for the optimisation
• Conversion to gdml then use GArSoft to run the analysis/

reconstruction; main trouble will be in the segmentation…

• Run directly simulation with DD4hep (direct conversion from

TGeo -> G4 geometry) and write plugin to get the same

• utput as GArSoft; advantage segmentation plugins are

already provided; however would need plugin to use GENIE as input for the simulation

• New geometry is almost finished
• Design will be optimised based on results (hopefully soon)

Next steps.

Eldwan Brianne | ECAL Status update | 28/11/2018 8 Page

Understanding performance and limitations

• Detector geometry is almost implemented
• Implement segmentation in GArSoft for reconstruction… (studies can be started at the simulation level)
• Understand photon/neutron distribution of neutrino events, is the detector geometry sensible?
• Understand detector performance:
• Need reconstruction for tiles and strips (can be rudimentary, maybe not straight forward)
• Performance scaling with parameter variation:
• Absorber thickness, scintillator thickness
• Cell size, strips width
• Arrangement of HG/LG layers: optimal placement of HG layers? Benefits?
• What are the limitations of such design? In term of energy resolution, angular resolution, non-uniformity response and

material in HG layers…