Skip to content

Important: Bug fixes affecting tracking resolution + Shape of luminous region for resolution studies#405

Merged
alkemyst merged 11 commits intotkLayout:dev_gabiefrom
ghugo83:luminous2
Dec 8, 2017
Merged

Important: Bug fixes affecting tracking resolution + Shape of luminous region for resolution studies#405
alkemyst merged 11 commits intotkLayout:dev_gabiefrom
ghugo83:luminous2

Conversation

@ghugo83
Copy link
Contributor

@ghugo83 ghugo83 commented Nov 29, 2017

In this PR:

  • A) Instead of IP = (0,0,0), a flat luminous region from Z = -70 mm to Z = 70 mm is chosen for resolution studies.
  • B) BUG FIX: One should not select the Z > 0 inactive volumes only (assuming eta >= 0). Indeed, the IP can be at Z < 0, hence inactive volumes on (-Z) side can be hit.
  • C) BUG FIX: One should not select the Z > 0 sensors only (assuming eta >= 0). Indeed, the IP can be at Z < 0, hence sensors on (-Z) side can be hit.
  • D) BUG FIX: Sensors tilt angle was not taken into account for crossed MB calculation, in tracking resolution studies.
  • E) Improved code on hits on inactive elements. This is used to calculate hit position and crossed MB. (though, on this point, issues were rather in devLite than in master).

Results (on OT613_200_IT404 geometry) :

The effects are significant on resolution (pixel), resolution (tracker) and resolution (trigger).
Example: Longitudinal impact parameter resolution vs. η (const Pt across η):

  • Before this PR:
    before

  • After merging this PR:
    after

NB: Details:
A detail of the impact of the fixes can be found here:

NB: TO DO list (coming in further PR):

  • Cleaner handling of shape of luminous region (in Simparms).
  • Remove old duplicated code on hits on Materials (tested, will have no effect on MB) + remove InactiveElement::getEtaMinMax().
  • Report issue in InactiveElement::checkTrackHits in devLite.

…e in the case of an inactive element crossing the Z=0 plane. 1) Should take the negative value of eta on the left side, otherwise comparisions with track eta does not make any sense!!! 2) Should be innerRadius + rWidth instead of innerRadius. Fortunately, was unused since all inactive surfaces are cut in 2 by the Z=0 plane.
…ctive element, one should not select the Z > 0 volumes only (assuming eta >= 0). Indeed, the IP can be at Z < 0, hence volumes on (-Z) side can be hit.
…ement::checkTrackHits has the following issues in devLite: some hits positions are wrongly computed and are actually not inside the inactive volume, hugely increasing the total crossed MB (and hence worsening the resolution). + Case of a tube or disk crossed in the non-trivial sense not always taken into account. Created new and fully debugged InactiveElement::checkTrackHits method (should be backported to devLitegit log).
@ghugo83
Copy link
Contributor Author

ghugo83 commented Nov 29, 2017

Fully tested :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants