CMS Tracker Visualization

What we have now:

• CMSIM Events
• GIOD Project Image Gallery obtained with JavaCMS applet ?

Tracker Atlas

• Complete detector:Full-scale mock-up
• Pixel: global 3D view,Barrel detail
• Sensors for Silicon Strip Tracker:15 different masks
• TIB:R-Phi and R-Z cross-sections,single sided and double-sided module.
• TID:disk layout,disk cross-section
• TEC:Numbering and names of the 16 wedges and 4 sides,sensor position on the wedges

1. Pixel Detector based on silicon pixels.
2. Inner Tracker based on silicon microstrips.
3. Outer Tracker based on silicon microstrips.

Silicon pixels are diodes implanted on semiconductor wafers that offer high spatial resolution in a plane giving the position where the particle has passed to within few microns. Silicon microstrip modules use instead strips as detecting elements. Concerning these, we have thin and thick modules, single and double sided modules . Each module consists of 1 or 2 6'' silicon wafers and is read by many readout chips called APV. Single sided modules measure only one coordinate , double sided modules give instead stereo coordinates. Double sided modules consist of two single sided modules mounted back to back sligtly tilted.

Tracker Barrel

Layer #	Layer name	Avg. radius	Modules in z	Modules in phi	Total # of modules	Annotations
1	PIX barrel	41.05	8	18	128	cylinder divided in half with half modules at the borders
2	PIX barrel	70.16	8	30	224	''
3	PIX barrel	98.88	8	42	320	''
4	TIB1	255.	12	26-30	336+336=672	tilted double sided modules
5	TIB2	340.	12	34-38	432	' '
6	TIB3	430.	12	44-46	540	tilted modules
7	TIB4	520.	12	52-56	648	' '
8	TOB1	610	12	42	504+504=1008	double sided modules
9	TOB2	696	12	48	576+576=1152	' '
10	TOB3	782	12	54	648
11	TOB4	868	12	60	720
12	TOB5	965	12	66	792
13	TOB6	1080	12	74	888

Tracker Endcaps

Disc #	Disk name	Avg. z position	Modules in phi	N of rings	Total # of modules	Annotations
1	FPIX	34	24	7		In fact the disk is built like a "turbine disk" with 24 "blades"
2	FPIX	46	24	7		' '
3	TID		24-24-40	3		Rings 1 and 2 double sided
4	TID		24-24-40	3		Rings 1 and 2 double sided
5	TID		24-24-40	3		Rings 1 and 2 double sided
6	TEC		24-24-40-56-40-56-80	7		Rings 1, 2 and 5 double sided
7	TEC		24-24-40-56-40-56-80	7		Rings 1,2 and 5 double sided
8	TEC		24-24-40-56-40-56-80	7		Rings 1,2 and 5 double sided
9	TEC		24-40-56-40-56-80	6		Rings 2 and 5 double sided
10	TEC		24-40-56-40-56-80	6		Rings 2 and 5 double sided
11	TEC		24-40-56-40-56-80	6		Rings 2 and 5 double sided
12	TEC		40-56-40-56-80	5		Ring 5 double sided
13	TEC		40-56-40-56-80	5		Ring 5 double sided
14	TEC		56-40-56-80	4		Ring 5 double sided

TOB

The TOB system consists of the following main components:

• The detectors: 4128 Rf -detectors, 1800 Stereo-detectors, total 5928 detectors.
• The detectors are collected on independent units, called rods. On the 'single-sided layers' a rod carries 6 Rf -detectors. On the 'double-sided layers' a rod carries 6 Rf - and 6 stereo-detectors mounted back to back. Each rod contains all the necessary cabling, electronics and cooling, needed for its modules. At the end of the rod there are connectors for all the services. In all there are 688 rods, 344 on the Z+ side and 344 on the Z- side.
• The rods are carried by a wheel structure, which consists of four vertical disks, which are joined together by three inner cylinders and three outer cylinders.

TIB

The TIB is made-out of two independent units (Forward and Backward TIB) each one containing four layers of silicon modules arranged in a tilted, cylindrical geometry.Single-(double)-sided detector modules carry one(two) silicon sensor(s) with an active area of about 61.5x116.9mm 2 . In each layer the modules are organized in longitudinal strings of 3 elements on both sides of the supporting structure. The position of the modules in such strings is optimized to assure full coverage from tracks coming from the interaction point. The supporting structure for each layer is made by two half cylinders.

TEC

Modules are mounted on a disk organized in petals. We have 8 inner petals and 8 outer petals for each disk. Sensors are mounted on both sides of a petal: so we have 4 positions of the sensors in z:these are named A,B,C,D in order of increasing z.Ring 1,3,5,7 are mounted on sides A and C; the other rings on sides B and D.

Detector Description Database

DDD is the CMS project that is in charge of the detector database. Waiting for the final database (in XML) to be ready we have now at least two descriptions of the tracker:one as tz files for CMSIM and ORCA and the other for OSCAR. There is also this first approximation of XML description.
The OSCAR description uses Geant4 and can be visualized with Iguana.

Tracker construction database

This is the reference page for tracker construction. There is a database ready to collect all information about the detector pieces. It is a Oracle relational database in Lyon and can be accessed through a Java program. It is necessary to install a stand-alone program on your computer. There is also the Tracker EDMS : a site that collects all mechanical drawings.

Tracker Detector Control System(DCS) Tracker DCS is already operational on the test beam: it controls classical "slow control"operating conditions(power supplies voltages, temperatures, humidities,leakage currents) and also Front End electronics configuration data and laser calibration data. It is based on a Oracle Database plus a control system called SCADA.
The Tracker is self-calibrating and regular electronic calibration should take place in dedicated runs.

The tools

• Iguana
• OpenInventor
• OpenGL
• Qt
• SoQt
• VRML
• Web interface :Example with clickable images created using this Java program
• Geant3 visualization
• Geant4 visualization
• ROOT
• Lizard : the new Paw ?
• Java:Wired,Jas,Java3D,...
• Paw

Objectives To build a set of graphics objects to help the visualisation of :

• The detector
• Raw event data
• Reconstructed event data(tracks)
• Information contained in the Construction database
• Other ... (Calibration,etc..)

1. Inside the reconstruction program Orca (integrated in Iguana).
2. In a stand-alone program (possibly integrated in Iguana) that will process events created by Orca and ready to be displayed.
3. In a Web interface

3D vs 2D : up to the level of the module, the structure is 3D. So the idea is to use 3D up to this level and then use 2D for further detail. Since single modules are planar in structure, this should not be a problem.

Not a replacement of what we still have but an addition to make what we still have really useful. Now the main problems inside Orca/Iguana are:

1. Too many 3D details:the display of everything is impossible unless you have a powerful video card. (For example the full display of the tracker seen in the previous picture requires around 500M of memory in OpenInventor!). On the other end, if you can select what you need , it can be used without problems on any computer.
2. There are no events "ready to be displayed", you have to wait for Orca to process the event and this means a lot of time for "normal" events.
3. You can't do a detailed selection (at the level of the single module) of elements to be displayed.

Why and when use a Web interface : in some cases you don't need to do complex manipulation (for example rotating a part of the detector in space), instead you just want to visualize some quantity. In this case a simple Web interface would be easier to develop and use. For example : display of data in the Construction Database. prototype

Why a stand-alone program? I some cases, it is absolutely necessary to work inside Orca, although difficult and slow this can be. Instead there is also the possibility that you want to process a lot of events (for example to sum the signal from many events) and you don't want to run Orca one week to get a single picture. In this case the idea is to run Orca in batch for one week, writing events ready to be displayed on disk and then use a fast stand-alone visualisation program to do the visual analysis. This program will get the detector and event description via XML files. DDD already has XML description of the detector; a similar description will be implemented/used to allow program indipendent transmission of events data between ORCA/Oscar and this visualisation stand-alone program. This format will be used for the "ready to be displayed events".

Tentative list of graphical objects to be developed:

• 2D map to make a fast selection of tracker parts like subdetectors,layers, rings,disks,endcap/barrel,etc.. prototype
• 2D map to allow selection of single modules: 41 different pictures one for each barrel layer and end cap disk. (Can we possibly reduce their number?) This page can give an idea of the level of detail required
• 2D module map to display information about single modules coded in various way: for example color to indicate state of construction. This page can give an idea of the level of detail required
• 2D single module representation

Todo

• Make Tracker display in Orca faster!
• Experiment with new 3D and 2D representation of the detector
  • Exploded 3D images(for example like a russian doll with all internal dolls shown outside)
  • Smart use of color
  • Animation
  • Use of transparency.
  • Try to create less nodes in the scene graph by grouping toghether nearby modules.
  • Forget about displaying all modules and use instead simpler shapes like cilinders.(For example start with 13 open cylinders for the barrel ; add 14 + 14 disks for the endcap ; then try to represent the modules on the surface perhaps as an image?)
  • Make a more clear image cutting part of the detector.
  • Make a more clear image taking away and rotating one of the endcap
  • See how fast is OpenInventor on a PC with a videocard that implements Opengl.
• Build Web interface that allows users to select a single modules of the around 20000 present. This should be a graphical interface that would display every module. Clicking on the module you can have access to some information about it. Start with a general map + 41 detailed maps (one for each barrel layer and forward disk).
• Try to connect this interface to the tracker construction data base.
• Representation of the signal in the tracker.
• Representation of the tracks
• Representation of detector+signal
• Representation of detector+tracks+signal
• It is possible to have a map of all 16632 modules in a single planar image?
  Here is the distribution of the centers of all modules:x vs y, x vs z and y vs z. A possible way can be to represent endcap -z, barrel, endcap + z side by side as exploded view of 14+13+14 layer. To represet a barrel layer we cut it along z and represent a phi vs z projection. This representation can be zoomed allowing a detailed view of the signal in thesingle modules. Instead the unzoomed view can easily represent quantities pertaining to single modules coded in a suitable way (color,vertical bars). In such a way it should be possible to see in a single picture the state of all 17000 modules.
• Find a smart way to select/deselect any part of the detector
• Add to the 3D representation a 2D one
• Build a purely 2D representation
• Show only hit modules
• Create a stand-alone display for the tracker? What can be the use of such a program? To monitor detector construction? As event display? As a online monitoring display?
• What about a theta-phi representation of the signal and the modules ?

Some open problems

• How do you use the Ig... objects in a OpenInventor program outside Orca
• How do you use the CMsscan viewer in a OpenInventor program outside Orca
• How do you read events to be displayed outside Orca
• Find a Java3D program that displays OpenInventor .iv files and experiment with it

Bibliography

• Making tracks for the LHC from the Cern Courier
• Atlantis:Event display in Java for Atlas.
• Atlantis presentation by Hans Drevermann
• Visualization for Atlas
• Atlas Event Display Gallery
• The CMS Tracker:production control
• The Silicon Strip Tracker for CMS
• Demonstration of Iguana
• Mechanical Engineering Division - CMS Pisa
• SST Tracker Links
• btau tasks 2004
• Addendum to TDR
• CMS theses
• HEPVIS 96
• HEPVIS 98
• HEPVIS 99
• HEPVIS 2001
• The Atlas and CMS trackers
• Tridas TDR
• DCS - SCADA tutorial
• Alcuni tutorial sul software CMS
• Il paradigma Model-View-Controller
• bozza Tesi Dottorato
• Tracker b/tau home page
• Computers, Graphics, & Learning
• Nota cms software 1 e 2
• Cms thesis with tracker object model, track parameters, ...
• Just an example of Uml design of a Qt progran
• Edward Tufte

• Java program that draws any of the 41 cylinders/disks. Requires the following data file. After compiling run with the command java Mostratr npart where npart is a number 1-41.
• Java program used with this data file to produce this image