The higher energy and luminosity of future HL-LHC imposed the development and testing of a new type high-rate detector known as GEM (Gas Electron Multiplier). A monitoring system designed to measure the radiation dose and particle fluence absorbed by the GEM detectors has been produced and installed at the CMS detector of LHC. It consists of a readout-control module, to which there can be connected up to 12 radiation monitors (RADMON). There are in each unit two types of sensors: RadFETs, measuring the total radiations dose and p-i-n diodes - for particle fluence. For the first test, a group of 6 GEM chambers was placed at the inner CMS endcap in March this year with one RADMON controlling the dose. After about 4 months of operation, the first results are analyzed. They show that for the integral luminosity 46 fm-1 the dose and the particle fluence are low (about 0,5 Gy and 3.1010 cm-2 1 MeVneq fluence) and only two more sensitive sensors are giving measurements and providing data that can be analyzed. Nevertheless, the experimental results confirm the dose and fluence values simulated by FLUKA. This is an important result for the radiation hardness test plans of the GEM part of the CMS muon detector.
The increase of the energy and luminosity during the coming upgrades of the CERN LHC creates a more hostile environment for the detector used. In the muon system of the CMS the heaviest conditions will be created in the region with pseudorapidity 1,6 < η < 2,2, (fig. 1), where for the high- luminosity phase of the LHC, Monte Carlo simulation gives particle rates of several kHz/cm2. This imposes severe restriction on the technology that can be used.
To solve the problem it was decided [
After numerous tests [
The installed RADMON (fig. 2) is a little different from the described in [
The basic data of the radiation sensors used are shown in Table 1. The relation between the
RadFET’s gate threshold voltage shift Vth and the radiation dose D is strongly dependent on the electric
field in the oxide during irradiation and on the process parameters, especially the thickness of the gate
oxide. It is nonlinear and best approximated by Vth = a×Db, (resp. D = (Vth)1/b/a), where the coefficients
a and b depend on the RadFET type as well as on the measured dose range. For this purpose, the entire
operating dose range is divided into zones for each of which a and b have different values [
The shift of the p-i-n diodes forward voltage VF is proportional to the 1-MeV neutron equivalent particle fluence Φ [cm-2]. The relation is generally linear – Φ = cVF, where c depend of the diode type. Table 1 shows that LBSD Si-1 is about 50 times more sensitive than BPW34S.
The readout system measures the voltages on the sensors using current pulses of different
amplitudes and durations prescribed by the producer [
Here we shall try to analyze the first data, read during the period of 15.05 to 1.11.2017 (the last
data are received after the NEC-2017 symposium, during the preparation of this talk for publication
and
for the operating ranges of the sensor used. This can be seen from the FLUKA simulations of the
dose and fluence distribution in CMS and cavern (fig. 5, [
These results immediately show that there will be no useful data from the sensors REM 130 and BPW34S, whose operating ranges begin from a few Gy and 2.1012 cm-2 respectively. Nevertheless, we found, that the voltage values received from them are not zero, which means that these sensors are operating and the control system read their data normally.
For our investigation, we took the experimental data from 5 different days during the period of LHC operation. In Table 2 for each of these days there are shown: the LHC integral luminosity until this moment; the total radiation dose – simulated by FLUKA and measured by the RadFET REM250; the 1MeVneq-fluence – simulated and measured by the p-i-n diode LBSD Si-1. These results are illustrated and linearly approximated in fig. 6. The errors are relatively high because of the low values of the dose and fluence – the maximal measured voltage shifts are only about several tens of millivolts. REM250 - Absorbed dose Si-1 - 1MeVneq fluence
Due to the relatively low integral luminosity delivered by LHC until now, the measured dose and fluence values are in the initial operating range of the sensors. For this reason, only the more sensitive sensors give useful data until now though with lower accuracy. There is good agreement between the data from FLUKA and the experiment, which is a confirmation of the reliability of the CMS BRILL dose simulation at the position of the GE1/1.
All radiation sensors are active and with the increasing of the LHC integral luminosity the data from all the RADMON sensors can be analyzed and compared with the FLUKA.
The "Radiation Monitoring of the GEM Muon Detectors at CMS" is part of the "CMS MUON ENDCAP GEM UPGRADE" project, which is financed by the Bulgarian Scientific Fund at the Ministry of Education, Youth and Science – grant DCERN 01/2 25.11.2011.