Transcript ITS_Alignment_Plus

ITS Alignment Status
Plus other things
By
Bjørn S. Nilsen
The Ohio State University
page 1 of 18
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Outline
page 2 of 18
●
ITS internal Alignment
●
Open questions about calibration
●
Changes in Tgeant3 and the Virtual MC interface
●
Speeding up of SPD simulation
●
planned Changes to AliITSgeom
●
Conclusions
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
ITS Alignment
●
●
page 3 of 18
2198 Si detector wafers
–
240 SPD wafers
–
260 SDD wafers
–
1698 SSD wafers
–
X 6 DoF = 13188 DoF
All of the other Radiation
thick Materials of ITS.
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Alignment view
●
Parts of “Detector response Analysis and Alignment”
may be done before “Reconstruction” and may use
some of the results from the Tracking and PID.
page 4 of 18
Offline wants new ITS
geometry to have
“movable” parts.
–
In principle do-able, but...
–
Moving part of the ITS
geometry leads to overlaps..
–
To make remove the
modified geometries
overlaps, shapes must change
(e.g. box->Trap)
–
This is a problem difficult to
solve even in principle.
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
ITS view (my opinion)
●
●
We align using AliITSgeom (the geometry of our
sensitive volumes)
–
This isn't quite good enough for tracking, the need to
know about the material budget between layers
–
Don't know quite how to solve this.
Using reconstructed virtual photons converted
within the ITS, we can know about all of the
material within the ITS
–
page 5 of 18
This, along with the corrected AliITSgeom, will be
used to rewrite a new ITS geometry (up to some
possible limits to be determined)
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Alignment Status
Work just starting up again (see ITS TDR for original results)
●
●
●
page 6 of 18
new AliITSAlignRecPoint (derived from AliITSRecpoint)
–
Adding data members needed for Alignment (Event#, Det#, Global
coordinate,...)
–
Adding functionality, not found in AliITSRecPoint
new memory resident array classes for AliITSAlignRecPoint
(by detector, track, ...)
More still to come
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Calibration
●
We had a long and active meeting about
calibration at the last Offline week.
–
Offline want to know, from us(?)
●
●
●
–
What data will be stored in the different data bases
What data “we” need for the Offline
And where all of this data can be found.
Offline presented a prototype “framework” to get
access to information kept in ALICE data bases
●
●
●
Prototype only worked with DCS data base
Most ITS information “we” need is not in DCS data base
“User” interface very poor
–
–
–
●
Seems as cumbersome as accessing data bases directly
–
page 7 of 18
must give an ascii input file with a list of all data possibly wanted
ascii file must also contain address of data base
format of ascii file dependent on which data base accessing
–
Aren't there supposed to be API's for all Data Bases?
Isn't there ROOT interfaces to all of these Data Bases?
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Process to Get Calibration Data
Peter Chochula
page 8 of 18
Boyko Yordanov
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Calibration Information
Whole ITS
Simulation
Reconstruction
●
page 9 of 18
Cooling: Air (flow rate, humidity, Temperature), Water
Cooling {1/half-Barrel (Flow meter, Waist Oxygen Levels), 26 each side
(Cooing circuit on/off, Temperature, Pressure)}, 2 Phase (?)
●
ITS Alignment monitoring System:
●
Racks: ?
●
Inter Lock: ?
●
?
●
Geometry, Survey (not in any data base, Alice Note?)
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Calibration Information
SPD
●
Simulation
Reconstruction
●
●
●
page 10 of 18
10 Sectors, 6 Staves/sector, 4 detectors/Stave, 5
readout chips/detector, 256×32 pixels/readout
chip
?
Dead pixels/readout chips/Staves, Noisy pixels,
Geometry
Noise levels, Thresholds, Temperature, Bias
voltage, Geometry
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Calibration Information
SDD
●
●
●
Simulation
Reconstruction
●
●
●
●
page 11 of 18
22 [14] ladders/layer, 2 sets of ladder end boards/ladder , 8
[6] detectors/ladder, 2 hybrids/detector 3
Ambra+Pascal/hybrid, 256 Anodes/hybrid
Power supply, low voltage, medium voltage, high voltage
Hybrid: Temperature, over current, voltage setting, readout
frequency,
CARLOS: low & high thresholds, enable/disable zero
suppression
?
High & low thresholds, baseline, noise level, temperature
(injectors), Geometry
Correction maps (?)
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Calibration Information
SSD
●
Simulation
Reconstruction
●
●
page 12 of 18
2 Layers, 33 [33] ladders/layer, 25 [22]
detectors/ladder, 2 hybrids/detector, 768
strips/hybrid
?
Threshold, Baseline, noise level, temperature,
bias voltage, Geometry
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Calibration Sources (?)
●
Detector Construction Data base
–
●
Special data and none data runs (after some
analysis)
–
●
temperature (injector events), ...
Data runs, real data (after some analysis)
–
page 13 of 18
noise, temperature,...
Regular data runs, calibration event (after some
analysis)
–
●
Detector & other positions,
noise (zero non-suppressed data), alignment,...
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
TGeant3
and
AliITSgeom initialization
Changes to TGeant3, TGeant3TGeo...,
and TVirtualMC.
●
●
●
Added function, GetTransformation, GetShape,
GetMaterial, & GetMedium
Now we can remove the use of TGeant3 functions
This will now allow all of our geometries to be use
with Geant3 (native/TGeo), Geant4, and Fluka
(except for the use of MANY)
●
page 14 of 18
Waiting for, Update of Geant3, ROOT, and OK from Peter
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
SPD Simulationdubna
Updates, to come
Under Development
Presently
1A
1
Trapezoidal Integration
along
track path between hits.
-
2
Easy to code & debug
Inefficient
Gaussian Integration along
track path between hits.
-
Very efficient
Many fewer steps (~10 compared to
~200)
functions
Require 4 evaluations of Erfc
Step size is of the order 1µm 1B
And/Or Analytic Integration
-
-
Much smaller than needed
Guarantees accuracy
Must be done via an expansion
Includes evaluation of Erfc functions
Loop over ±3 (x), ±2 (z) pixels
4 evaluations of Tmath::Erfc 2
~exp(8th order polynomial)
Max
4
Loop over ±3 (x), ±2 (z)
Only as needed
3
More general algorithm
pixels
3
page 15 of 18
Applicable to SPD and SSD
Bjorn S. Nilsen, ALICE November-16 ITS Software
Maymeeting
be expendable to SDD
SPD Simulationdubna
Updates, to come
●
All parameters will be unchanged
–
Work on optimizing parameters can use existing code
–
Only algorithm will change
●
Code ready by End of November
●
Code tested by End of December
page 16 of 18
–
Comparison of SDigits between versions
–
Require same numbers up to ±1 electron or better
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Changes to AliITSgeom
●
●
page 17 of 18
Allow it to work with the “new” tracking standard
detector local coordinate system (see presentation
by Juira)
Give AliITSgeom direct access, if not complete
ownership of, the segmentation class
–
AliITSgeom would then do transformation from
module index coordinates through ALICE global
coordinates.
–
With one call/function one can go from any set of
coordinates to any other.
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting
Conclusions
●
ITS internal Alignment work started
●
Open questions about calibration ????????? help!
●
Changes in TGeant3 and the Virtual MC interface
ready to be put to use
●
Speeding up of SPD simulation coming soon
●
planned Changes to AliITSgeom coming soon
page 18 of 18
Bjorn S. Nilsen, ALICE November 16 ITS Software meeting