"A novel method of automatic particles identification for large CsI(Tl) detection systems"
Joanna Borgensztajn, Institute of Physics, University of Zielona Góra
(id #83)
Seminar: YesPoster: No
Invited talk: No
CsI(Tl) scintillators are often used in nuclear and high energy physics as charged particles detectors. Optical properties of CsI(Tl) allow simultaneous measurement of different types of particles and quite good discrimination between them. The last possibility is related to the fact that duration of the light pulse emitted by some materials is determined by two time constants. The first of them corresponds to so-called fast component of emitted light and varies very strongly with particle mass, charge and energy. The second corresponds to so-called slow component and is regarded as independent on particle type.
By recording the pulse shape and applying Pulse Shape Discrimination Method [1] it is possible to find for each event two digital variables proportional respectively to the fast and slow component. Data analysis is usually done by displaying these variables on two-dimensional fast-slow plot (see examples in [1-3]) and by putting graphical cuts or masks separating isotopic branches.
The idea of this method is very simple but it should be noted that for two or more identicallyfabricated detectors we never obtain identical fast-slow plots. Moreover plots obtained for one detector vary with temperature. Ageing effects of scintillators are also present so the procedure of putting masks should be repeated many times, according to the current conditions. Especially for devices consisting of several hundred detectors as CHIMERA [1] or INDRA [3] the method is very time consuming process.
To overcome this difficulty a modification of the commonly used method is proposed. The modification is based on the concept that all fast-slow plots obtained for one or more detectors could be standardized to one pattern. The method seems to be very promising in automatic particles identification, crucial for nuclear experiments on multifragmentation in which huge number of fragments have to be detected and identified. In present work the basis of the method and some important results are described and discussed.
References:
[1] M. Alderighi et al., NIM A489, 257-265 (2002).
[2] S. Aiello et al., NIM A369, 50-54 (1996).
[3] J. Pouthas et al., NIM A357, 418-442 (1995).
By recording the pulse shape and applying Pulse Shape Discrimination Method [1] it is possible to find for each event two digital variables proportional respectively to the fast and slow component. Data analysis is usually done by displaying these variables on two-dimensional fast-slow plot (see examples in [1-3]) and by putting graphical cuts or masks separating isotopic branches.
The idea of this method is very simple but it should be noted that for two or more identicallyfabricated detectors we never obtain identical fast-slow plots. Moreover plots obtained for one detector vary with temperature. Ageing effects of scintillators are also present so the procedure of putting masks should be repeated many times, according to the current conditions. Especially for devices consisting of several hundred detectors as CHIMERA [1] or INDRA [3] the method is very time consuming process.
To overcome this difficulty a modification of the commonly used method is proposed. The modification is based on the concept that all fast-slow plots obtained for one or more detectors could be standardized to one pattern. The method seems to be very promising in automatic particles identification, crucial for nuclear experiments on multifragmentation in which huge number of fragments have to be detected and identified. In present work the basis of the method and some important results are described and discussed.
References:
[1] M. Alderighi et al., NIM A489, 257-265 (2002).
[2] S. Aiello et al., NIM A369, 50-54 (1996).
[3] J. Pouthas et al., NIM A357, 418-442 (1995).
