• ooo

Performance simulation studies of a realistic model of theCBM Silicon Tracking System

Silicon Tracking for CBM

Reconstr

ucted U

RQMD event:

centra

l Au+Au co

ll. @

25AGeV

Challenge: • Au+Au collisions , 25 GeV/u • up to 1000 charged particles/event• interaction rates up to 10 MHz

STS requirements:radiation hard and low-masshigh position resolutiontrack reconstruction for all charged particles momentum resolution of the order of 1.5%

at 1GeV/cprimary vertex resolution of about 50 µm

STS Layout

Simulation and signal digitization

8 tracking stations: 30, 40, 50, 60, 70, 80, 90, 100 cm downstream of the target, in a 1 T dipole magnet;

4 m2 active area;The number, positions and

segmentation of the layers are optimized for efficient track reconstruction and high momentum resolution.

The STS stations have a modular structure and are constructed from 4 types of 300 µm thick double-sided silicon micro-strip sensors, arranged on 8 types of carbon-fiber supported ladder structures.

Number of integration

componentsLadders 106

Sectors 1040

Sensors 1292

R/O chips 16640

Channels 2133k

Anna KotyniaGoethe University, Frankfurt, Germany

for the CBM Collaboration http://www-cbm.gsi.de

Reconstruction

The simulations comprise the complete chain of physical processes for charged particles traversing the detector, from charge creation in silicon to the digital output signals [1].

During the digitization process realistic detector response functions are applied, such as:• signal sharing between strips;• charge collection inefficiency;• Lorentz shift due to the presence of the magnetic field;• channel dead time;• random noise added to the signal;

Due to charge sharing between strips, the majority of ctracks is expected to spread signal across several strips. Therefore, a cluster algorithm has been applied to refine the reconstruction of the particles’ positions .

The cluster center is determined using the Center of Gravity (COG) equation.

n

ii

n

iii

ADC

ChADCCOG

1

1

The reconstructed total cluster charge matches the expectation: i.e. 23 ke- in a 300 µm thick silicon sensors.

Illustration of the cluster finding method.The bottom part of the figure shows the charge division sampled at the horizontal line in the middle of the picture.

Occupancies and reconstruction results

The application of the realistic detector responce functions during digitization, of the COG algorithm for space point recontruction and the Cellular Automaton [2] for track finding and Kalman Filter for track fitting, results in:

Track category Efficiency, %Reference set (>1 GeV/c) 95.2All set (≥4 hits,>100 MeV/c) 89.8Ghost (made from fake hits) 6.6

The occupancy has been evaluated in the high luminosity scenario for minimum bias collisions, assuming an interaction rate of 10 MHz.

The chip occupancy is relevant to the data acquisition, while the sensor occupancy influences the tracking performance. Hot regions are observed in a small fraction of the detector, with more than 32 MHz data rate and more than 20 hits/cm2/event.

Hit Finding efficiency: Primary : 98.97 % Secondary : 91.69 %

Point resolution: Vertical : 60 µmHorizontal: 8 µm

References:[1] A.Kotynia, J.M.Heuser „ Performance simulations of the CBM Silicon Tracking

System”, CBM Progress Report 2011 [2] I.Kisel, I.Kulakov, I.Vassiliev, „A standalone package for on-line event selection in the

CBM experiment”, CBM Progress Report 2008 79Simulated track of a minimum ionizing particle in the silicon detector

STS in dipole magnet

Realistic STS system in CbmRoot

Distribution of the number of strips per cluster.

Sensor occupancy Chip occupancy

Average hit density per sensor for the STS station closest to the target.

Distribution of data rates produced by the readout chip of all eight STS stations.