Apar Agarwal authored 1 year ago and Administrator committed 1 year ago

Earlier the pads on one edge of the frame used to be under estimated on one
side while they were overestimated on the opposite side.
This problem should now be rectified by taking into account the special case
of module boundaries.
Add special case of pad size being smaller than the spot radius

Eoin Clerkin authored 1 year ago and Pierre-Alain Loizeau committed 1 year ago

A simulation geometry which converges to CAD. See info
file for more details.

Simon Neuhaus authored 1 year ago and Pierre-Alain Loizeau committed 1 year ago

- change outer dimension of RICH detector
  - change width and length of RICH
  - change mainframe
    - make shape in top view rectangular
    - make slope on front
    - reuse mainFrame3
      - change mainframe5 to trapezoid form
      - add new mainframe7 which replace some of the mainFrame2
    - cutoff shieldingbox
      - add new mainframe8 in x-direction
    - reduce some mainframes by a length of (2x) thickness from the mainframe to reduce overlapping
    - cutoff a cubic where mainframes cross each other
  - change the length of the pillars
- Rich entrance window
  - change width of the entrance window
  - add 10cm wide and 2cm thick Aluminum flange
  - reduce size of entrance window without flange a bit
- Rich exit window
  - add 10cm wide and 2cm thick Aluminum flange
  - change width to 569cm
  - change hight to 384cm
  - change material of exit window from kapton to Polycarbonat
  - increase thickness to 1cm
- Add Studs to each pillar
  - 10cm x 10cm 0.2cm thick Aluminum
- Add Connections between pillars
  - 6cm x 2.5cm 0.2cm thick Aluminum
- Make beampipe to beampipe tunnel
  - change material to carbon
  - change radius at entrance and exit
  - change pipe thickness to 0.3cm
  - increase the length by 0.5cm
  - change and increase cutoff at middle mirror parts
  - increase hole in sensitivity plane (move subtracted Cone)
  - remove belt types 4 and 5 because of overlap with the beampipe tunnel
- change CreateStuds
  - CreateStuds now need length and angle
  - selecting of length and angle with fix value if conditions is removed
  - remove some typo failurs where was schieldingbox written instead of shieldingbox.
- renamed Polycarbonat to RICH_Polycarbonat in media.geo
- change cutoff cone in exit, entrance window and senseplane to tubes to fix issue with a small structure hanging in the circle.
- add some comments about creating older version of geometry.
- make output reproduceble and change commet about producing older versions.

=> Also change geo setups check to be executed by CMAKE in geometry install step (also if already downloaded)
=> This is needed as the new way relies on the geometry git copy being already present

Eoin Clerkin authored 1 year ago and Pierre-Alain Loizeau committed 1 year ago

Correct TRD info files

Eoin Clerkin authored 1 year ago and Pierre-Alain Loizeau committed 1 year ago

Downgrades much in lmvm setup

When sampling overlaps are less than predefined amount, then
ignore for test success/failure purposes. Combine arguments
for flag and sampling size. Rename and comments. Clang format.

- Make overlap threshold optional input parameter (default = 0.0001 as before)
- Make sampling controllable by flag (disabled by default)
- Make number of samples in sampling mode controllable by input par (default = 0 leading to ROOT default of 1M)

Two problems from ROOT side lead to it being unrelyable for now
- Threshold is not working in this method
- Changing the number of samples lead to really small overlaps apearing and disappearing

See https://github.com/root-project/root/issues/14675 for reference

Two problems from ROOT side lead to it being unrelyable for now
- Threshold is not working in this method
- Changing the number of samples lead to really small overlaps apearing and disappearing

See https://github.com/root-project/root/issues/14675 for reference

Eoin Clerkin authored 1 year ago and Pierre-Alain Loizeau committed 1 year ago

Increase the number of sampling points to 2M

Administrator authored 1 year ago and Florian Uhlig committed 1 year ago

After the latest changes to flesnet it is possible to also build the tsclient
on arm64.