Master by Research (MRes) project of Bogdan Mishchenko :"Simulation, reconstruction, Physics sensitivity studies for the SiD(Sillicon Detector) "

Project Aim

The main aim of the project is the Physics sensitivity studies(Higgs physics, top quark Physics, beyond standart model physics) at future ILC(International Linear Collider). The project focused on simulation, reconstruction, physics analysis of of the results Tracking system from SiD(Sillicon Detector). Below you can find link to the brief report on the progress:

• Brief_report.pdf

SiD detector model

SiD detector mentioned above is the one of two detector concepts proposed for ILC. Below you can see layout of the detector. My thesis is focused mostly on the Tracker subsystem. Please refer to the report that will be attached later on for more details on it.

This is angle coverage for SiD Tracker system. For low angles from beamline(<40^{0}) acceptance will be mostly by endcaps(inner outter vertex disks and tracker disks), while for angles greater then 40^{0} will be mostly Barrel(Vertex and Tracker).

Generation/simulation chain

Prior simulation of detector response of the SiD, we need to generate particles. In the simple case it can be running script that specifies PDG number of particle, 4 vector of particle and other usefull parameters.

How to generate file with blank of particles

Assuming that you have installed lcgeo(linear collider geometry package). If not, follow instruction on https://twiki.ppe.gla.ac.uk/bin/view/LinearCollider/GlaSiDGettingStarted . On the fresh login to terminal execute :

source /cvmfs/sft.cern.ch/lcg/releases/gcc/4.8.4/x86_64-slc6/setup.sh
source /cvmfs/ilc.desy.de/sw/x86_64_gcc48_sl6/v01-17-10/init_ilcsoft.sh
cd lcgeo/
python example/lcio_particle_gun.py (where you specify PDG and incident angle \theta  and number of generated particles and other parameters)
Do the following command to have a look inside the file:
anajob mcparticles.slcio (prints MCParticle collection for each event)
or 
dumpevent mcparticles.slcio (prints details of MCParticle collection for each event)
Run the simulation:bogdanmishchenko
ddsim --compactFile=SiD/compact/SiD_o1_v01/SiD_o1_v03.xml --runType=batch --inputFile mcparticles.slcio -N=10  --outputFile=testSiD_o1_v03.slcio
anajob testSiD_o1_v03.slcio (prints MCParticle collection and hit collections for different subdetector parts (SiTrackerBarrelHits of SimTrackerHit type) and other collections )

Digitization/reconstruction chain

MC TruthTracking

Aim of the MC(Monte Carlo) Truth Tracking is to get make sure that we can get sensible results from analysis of the data from the simulation/reconstruction chain.

Definitions

Description of the reconstruction code, How to run it.

Reconstruction code for SiD was adopted from CLIC (Clicperformance package). It has following modifications: For layer residuals code we are using hard coded SiDecoderString: "subdet:6,side:3,layer:4,module:12,sensor:1" for decoding. Additionally, We do not do refitting and extrapolating of tracks, what have been implemented in the CLIC code.

Assuming that you are on the fresh login in terminal:

source /cvmfs/sft.cern.ch/lcg/releases/gcc/4.8.4/x86_64-slc6/setup.sh
source /cvmfs/ilc.desy.de/sw/x86_64_gcc48_sl6/v01-17-10/init_ilcsoft.sh
 cd lcgeo/
wget https://twiki.ppe.gla.ac.uk/pub/LinearCollider/SiDTrackingEfficiencies/Tracking.tar.gz   (To get the reconstruction code)
tar xvzf Tracking.tar.gz Tracking/ (To unpack archive)
cd Tracking/
find(to list all files in the directory)
./src(directory with Marlin processor source codes)
./src/Global_residuals.cc (processor that output residuals, pulls for hits)
./src/HitResiduals.cc (processor that output layer residuals )
./src/ResidualsCalculator.cc( processor that calculates trackreconstruction efficiencies versus, pt(transverse momentum, theta, eta(pseudorapidity)
./include ( directory that contains header files)
./include/Global_residuals.h 
./include/HitResiduals.h
./include/ResidualsCalculator.h
./include/LinkDef.h
./include/fpcompare.h
./CMakeLists.txt ( CMake configuration file )
./SiDReconstruction_161910_new_bogdan.xml (Marlin steering file)
./treeDraw.cxx  (script to draw residuals per layer with specifying root file)
./make+run.sh~
./treeDraw_x.cxx (script to draw x layer residuals)
./pulls.cxx ( script that draw pull distributions)
./make+run.sh ( bash script to compile code and to run Marlin with steering file)
./histograms500GeV_ttbar_10000.root (root file with tree structure that contains track  reconstruction efficiencies)
./analysis_100_40_10_mu.root (root file with tree structure that contains residuals and pool distribution)
./hitresiduals_layer_40_10_v02_SiD_100.root (root file that have layer residuals )
./SiD_02_v01_10_10_85.slcio (slcio file that can be used as input for running Marlin)
./gear_sid.xml ( gear with detector specification that attached to the Marlin steering file) 

 Then: 
mkdir build  lib ( make directories)
cd build/
cmake -DCMAKE_CXX_COMPILER=`which g++` -DCMAKE_C_COMPILER=`which gcc` \
  -DILCUTIL_DIR=/cvmfs/ilc.desy.de/sw/x86_64_gcc48_sl6/v01-17-10/ilcutil/v01-03/ -C $ILCSOFT/ILCSoft.cmake ..
make 
make install
After successful compilation:
cd ..
export MARLIN_DLL=$MARLIN_DLL:$PWD/lib/libTrackAna.so  (to link Marlin with compiled object files)
Marlin SiDReconstruction_161910_new_bogdan.xml(with specifying slcio file (SiD_02_v01_10_10_85.slcio))
Marlin -c  SiDReconstruction_161910_new_bogdan.xml(check given  steering file for consistency) 

How to produce Hit residuals

How to produce Hit pull distributions

If you will specify in the in the  pulls.cxx script  root file : (TFile *residuals = new TFile("analysis_100_40_10_mu.root")) and then:
root pulls.cxx

Below you can find Pulls distribution track parameters, which indicate good performance of the MC Truth Tracking(Gaussian distributions with mean close to 0 and sigma close to 1(small box) ):

How to produce Tracking efficiencies

Efficiency of track reconstruction depends on the theta angle(angle between projection of momentum on z axis and momentum) and transverse momentum. You can see below dependence on theta for SiD_o1_v03 version of detector model. As you have noticed efficiency of reconstruction falls significantly after 40 due, to the problems with Forward Tracking Disks (after 120^{0} ) and Endcaps(both Tracker and Vertex) (< 40^{0} ).

This is tracking efficiency for momentum range [1,500]GeV

• Tracking efficiency for SiD_o2_v01:
  

How to produce track parameter resolutions (d_{0}, z_{0}, pt)

D0_resolution can be produced by running following script D0_resolution_angle.cxx:

Z0 resolution obtained using Z0resolution.cxx script

On the figure below you can see Transverse momentum resolution obtained in the new framework and comparison with Detailed Baseline Design report from 2012(following figure) * This is Pt resolution obtained with SiD_o1_v03 model using momentumres_angle.cxx script:

• This is Pt resolution obtained with SiD_o2_v01 module:
  

Pattern recognition

Pattern recognition is a field of applied mathematics, which aim is the classification of objects. For instance, position measurements along particle trajectories can be classified as different physical objects. Please refer to the report that will be attached later on for more details on it.

How to run Conformal tracking code

Tracker optimization studies

The aim of the optimization studies is to optimize particularly, Tracking system for the best price/performance ratio. For example, We can look how each layer of tracker perform or how the tracking system perform with pixels versus performance with strips. For more details please refer to the report that will be attached latter on for more details on it.

Physics sensitivity studies

Physics studies for SiD can be following: precise Higgs physics measurements or searches for new physics like composite Higgs models, top quarks physics.Please refer to the report that will be attached later on for more details on it.

Marlin tips

For more options of running Marlin:

Marlin -? 
Particularly useful:
Marlin -c SiDReconstruction_v03_160711.xml  (check given  steering file for consistency)