ATL-TILECAL Jul 1998

Transcription

1 European Organization for Nuclear Research ATL-TILECAL Jul 1998 Bicron bers: choice of the UVA concentration for the Barrel Module 0 M. David, A. Gomes and A. Maio LIP and Univ. Lisbon, Portugal 1

2 1 Introduction Historically, in 1993, the rst bers used in the prototype modules of the Tilecal calorimeter were Bicron single clad (SC) BCF91A, and in 1994 the bers used to equip the prototypes were multi clad (MC) Kuraray Y11. Both of these bers had not UV Absorber (UVA). Tests with the prototypes have shown a large enhancement of the calorimeter response to pions and muons in the crack region between modules. The enhancement was due to the light produced in the bers (direct response to particles hitting the bers), and to the strong non-uniformity of the rst tiles produced for the Tilecal modules. Those tiles had grooves at the edges for the tile-ber coupling, and their response was much higher near the ends than in the center. The ber producers were asked to add a small amount ofuva to the bers in order to suppress part of the ber direct response, and both Bicron and Kuraray produced sets of sample bers with dierent UVA concentrations for optimisation purposes. The UVA used by both producers suppressed partially the direct response of the bers to particles as it was envisaged, but they also cause some degradation in the response to scintillator light. For the Kuraray Y11 bers the degradation of the response to scintillator light is negligible, but the UVA concentrations required to reduce the direct response of the Bicron BCF99-28 bers to the same levels of Kuraray Y11 bers lead to the suppression of a non negligible part of the ber response to scintillator light. The results were published in previous notes [1] [2]. In 1995, Y11(200)MS bers with 600 ppm of UVA were selected as the best compromise between high light output and small direct response. Adding 600 ppm of UVA to the Y11 bers resulted in a substantial decrease of the direct response of the bers, a decrease of about 60%, with negligible loss for the scintillator response. The response of the 1995 prototypes to pions showed a drop of the signal of the order of 5% in the crack region, but the tiles used in 1995 also contributed to this drop. These tiles were the best we have used in terms of attenuation length, and the choice of the UVA concentration did not take into account their improvement of uniformity. The results with the Y11 bers with 600 ppm of UVA were used as a \guide-line" for the choice of the UVA concentration of the Bicron bers to reinstrument the Barrel Module 0 during

3 2 Strategy for the choice of the UVA concentration for the Bicron bers to be used in the re-instrumentation of the Barrel Module 0 in 1998 Bicron MC bers were a candidate to be used in the Barrel Module 0 being reequipped in For all the samples tested until mid 1997, the light output of the Bicron bers without UVA was not optimised: the light output at distances of the order of 1.5 m was about 30% below the light output of Y11(200)MS bers with 600 ppm of UVA produced for the 1995 prototypes. The attenuation length of the Bicron bers was acceptable, i.e., of the order of 2.8 m. From previous results [2] it was known that for SC BCF99-28 bers, 60 ppm of UVA would suppress about 10% of the scintillator light, and no data was available for smaller UVA concentrations. Since the response of the modules in the crack region was below the module average, the direct response could increase a little, but we could not allow further scintillator light loss. In order to study the UVA eect in the low concentration range, bers with smaller amounts of UVA were ordered to Bicron, but some production and labelling accidents made impossible to use the bers produced for the intended study. The same happened to the MC bers ordered for the Barrel Module 0 (the amount of UVA asked was 20 ppm of UVA, a \guess" to get the best light output at expenses of the electron/scintillator response ratio). These last ones, labelled BCF99-28-G MC, showed a small attenuation length (about 2.5 m), resulting in a light output at a distance of 1.7 m about 30% smaller than the one obtained with the Y11 bers of the 1995 prototypes. At this distance, even SC bers of the same type showed a slightly higher light output. These bers could not be used in the instrumentation of the Barrel Module 0 and were rejected. The problem was discussed with Bicron, and about 6 weeks later Bicron sent us new MC BCF99-28 bers with 20 ppm of UVA, and these last ones showed an acceptable attenuation length ( cm) and a good light output. Compared to the Y11(200)MS bers with 600 ppm of UVA produced for the 1995 prototypes, the light output at distances of the order of 1.5 m was only about 5% smaller, and it was about 5% above the light output of 3

4 the Y11(200)MS bers used in the Modules 0. But the new Bicron bers with 20 ppm of UVA showed a direct response to electrons about 3.8 times higher than the Y11 with 600 ppm of UVA. There was no time to produce new samples with dierent UVA concentrations, so it was necessary to decide the amount ofuvabased on the existing data at the time. 3 Procedure to combine all the data The existing data in January 1998 resulted from several sets of measurements of dierent bers. A set of MC Y11 bers with UVA concentrations 0, 30, 100, 300 and 1000 ppm was tested in 1995, at the same time as a set of SC BCF99-28 bers with UVA concentrations 0, 60, 300, 600, 2500, and ppm [2]. The results are shown in tables 2 and 3. In January 1998 the new MC BCF99-28 with 20 ppm of UVA were tested at the same time as the Y11(200)MS with 600 ppm of UVA from the 1997 Extended Barrel Module 0. Since the data is in arbitrary units, a normalisation of the data was required before the analysis could be done. The scintillator is often changed, and the results obtained change from one measurement to the other depending on the quality of the scintillator used, on the scintillator-ber distance and on the damage caused to the scintillator by the source that is permanently in front of it. The analysis was done using only a typical distance of the ber excitation point to the PMT. It was chosen a distance of 140 cm as in the Tilecal note 95 [2]. The ber Y11(200)MS with 600 ppm of UVA was not tested in 1995, but the change of the light output with the UVA concentration is well behaved as can be seen in gures 1 and 2, so the points for the 600 ppm concentration were built by interpolation using the neighbour concentrations of 300 and 1000 ppm. The resulting points are plotted (solid circles) in the same gures. The points built in the plot for the Y11(200)MS with 600 ppm of UVA can be used now to convert all the values of the note to the arbitrary units of January 1998, with the multiplicative factors obtained from table 1. 4

5 date scintillator response (a.u.) electron response (a.u.) note 95, Jan Table 1: Light output for Y11(200)MS 600 ppm bers, in Tilecal note 95 (interpolation) and measured in January The results of the Tilecal note 95 are converted to January 98 arbitrary units multiplying them by the factors f e =0.727 and f s =0.138 for electrons and scintillator, respectively. electron response (a.u.) scintillator response (a.u.) UVA note 95 Jan note 95 Jan Table 2: Light output for Y11(200)MS bers, at a distance to the PMT of 140 cm, in Tilecal note 95 and converted to January 98 arbitrary units. Since the BCF99-28 bers tested in January 98 are MC and the other BCF99-28 are SC, a correction factor has to be applied to the SC bers to increase their light output to the level of the MC bers. In order to estimate the correction factor, it was assumed that the BCF99-28 MC bers with 20 ppm of UVA had no loss for scintillator light when compared with similar bers without UVA. With this assumption it is possible to set the scale MC/SC (at a distance of 140 cm) for the Bicron bers. The value shown in the note for bers without UVA (2.86) multiplied by f s gives 0.395, which should correspond to 0.50 obtained for the MC ber, so the MC/SC factor f mc=sc is 0.50/0.395=1.266, i.e., the MC BCF99-28 gives about 27% more light at a distance of 140 cm than the respective SC ber. In what concerns the direct response of the bers, it is assumed that the scale factor MC/SC is the same for scintillator and direct response to 5

6 electrons, based on previous results obtained with the bers of Bicron tested in September The results obtained, in the arbitrary units of January 98 for both types of bers, are shown in tables 2 and 3 and in gures 1, 2, 3, and 4. All the results in the gures are for MC bers. electron response (a.u.) scintillator response (a.u.) UVA note 95 Jan note 95 Jan MC Table 3: Light output for BCF99-28 bers, at a distance to the PMT of 140 cm, in note 95 and converted to January 1998 units. A factor f mc=sc was multiplied by the note SC bers to give all results for MC bers. 4 Ratio electron/scintillator and conclusions There are 2 factors to be taken into account in the decision on the amount of UVA to be used in the BCF99-28 bers to equip the modules: the ratio of the electron direct response to the scintillator response the response to the scintillator light, which cannot be degraded by more than a few percent (5%). The ratios were calculated both for Kuraray Y11 and Bicron BCF99-28 bers, and are shown in gures 5 and 6. The electron to scintillator ratio for the Y11(200)MS 600 ppm UVA is about 1.1. A higher ratio can be used, and the test beam results will help 6

7 in the optimisation of this parameter. So an increase to in the electron to scintillator ratio will be tested, being in the middle between the rst bers used by Tilecal (BCF91A without UVA) and the present situation (Y11(200)MS 600 ppm UVA). Such ratio is obtained in the range of 60 ppm of UVA, but a drop of scintillator light of the order of 10% is expected to occur. That means that we expect almost the same light as we have with the Y11 bers of the Modules 0. 5 Final remarks The Barrel Module 0 in 1998 was instrumented with bers from Bicron and bers from Pol.Hi.Tech. The half-module instrumented with Bicron bers is equipped with MC BCF99-28 bers with 60 ppm of UVA in the cells of sampling A, and with MC bers BCF91A in the samplings BC and D. This distribution results from a problem that Bicron had during the production of the BCF99-28 bers with UVA. After the successful production of the bers of the rst preforms with the lengths required for sampling A, they were not able to produce more good bers from the preforms they had prepared with UVA. The most similar bers that Bicron was able to produce in time for the instrumentation of the module were the MC BCF91A bers. References [1] M. David et al., \Comparative Measurements of WLS bers", ATLAS Internal Note, TILECAL-NO-034, 1994 [2] M. David et al., \Systematics of WLS bers with UV absorber", ATLAS Internal Note, TILECAL-NO-095,

8 Figure 1: Y11 direct response to electrons as a function of the UVA concentration. The point at 600 ppm was obtained by interpolation. The point at 1 ppm corresponds to bers without UVA. Distance to the PMT is 140 cm. 8

9 Figure 2: Y11 response to scintillator as a function of the UVA concentration. The point at 600 ppm was obtained by interpolation. The point at 1 ppm corresponds to bers without UVA. Distance to the PMT is 140 cm. The bers without UVA were expected to give more light, and it was not found the reason to give about 5% less than expected. 9

10 Figure 3: BCF99-28 direct response to electrons for MC bers as a function of the UVA concentration. The results were extrapolated from SC bers, except for the 20 ppm concentration (ber tested was MC). The point at 1 ppm corresponds to bers without UVA. Distance to the PMT is 140 cm. 10

11 Figure 4: BCF99-28 response to scintillator for MC bers as a function of the UVA concentration. The results were extrapolated from SC bers, except for the 20 ppm concentration (ber tested was MC). The point at 1 ppm corresponds to bers without UVA. Distance to the PMT is 140 cm. 11

12 Figure 5: Ratio of the ber direct response to electrons by the ber response to scintillator light for the Y11(200)MS bers, as a function of the UVA concentration. Distance to the PMT is 140 cm. The point at 1 ppm corresponds to bers without UVA. 12

13 Figure 6: Ratio of the ber direct response to electrons by the ber response to scintillator light for the BCF99-28 bers, as a function of the UVA concentration. Distance to the PMT is 140 cm. The point at 1 ppm corresponds to bers without UVA. 13