Test and development of a Cherenkov di usion detector
Contents
1. so e Place the tile to be cut as preferred and switch on the saw Set the saw speed to at least 350 ft min this increases stability and minimizes the risk of breaks Always set speed with moving saw only Otherwise the saw could be damaged 3 2 Baking silica aerogel As emphasized before aerogel produced by Airglass is extremely hydrofillic This evidence is shown in the table where the weight variations of the samples after the baking process are reported In the table m is mass of the aerogel sample before the baking an is the decrease of the mass after the baking time indicated in the table sample m g SA SA FAC m m 8 hrs 16 hrs 24 hrs The baking process is necessary to eliminate soaked water This proce dure improves the optical properties of the aerogel as shown in Fig 4 where are reported the transmittance curve by the aerogel before and after the baking 32 h In our experiment an air circulation based oven was used 7 This oven is programmable by the user and the following cycle was set e 4 hours upward ramp from ambient temperature to 300 Celsius de grees e 32 hours soaking at 300 Celsius degrees e 4 hours downward ramp from 300 Celsius degrees to ambient temper ature 1 Although the mentioned oven was the only one available at the time of our work a 500 Celsius oven is planned to be acquired to be acquired for the bigger aerogel detector 7 smittance an z f2. trigger scintillators position This increased the total diffusing area Indeed the mean number of photoelectrons slightly increased 10 Figure 5 Separate ADC spectra for each PMT calibrated at channel 100 Un baked aerogel was installed inside the prototype Figure 6 Separate ADC spectra for each PMT Run with a 32 hours baking process of the aerogel tiles 11 Figure 8 Sum spectrum for non baked aerogel of the six calibrated ADC This refers to high energy muons trigger 3 included 12 ob TE Mat 1000 25C 1500 Figure 9 Sum spectrum for baked aerogel of the six calibrated ADC This refers to high energy muons trigger 3 included Note the great improvement in collected photoelectrons 6 Effect of magnetic field on photoelectron number During the measurements described above a mu metal shield was installed on each PM to let the photoelectrons numbers to be not changed by the presence of a magnetic field The magnetic field around the detector was measured to be 0 5 Gauss at its maximum With the mu metal shields removed as can be seen by fig 11 12 the single contribution of each phototube and their sum slightly decreases of an average factor of 6 7 Conclusions This study showed that a baking procedure on hydrofillic ac
3. used in the past Nevertheless aerogel as Cherenkov radiator has several drawbacks such as poor light transmission in the UV visible wavelength region a region in which most phototubes are sensitive and extreme fragility In addition acrogel tiles produced by Airglass Inc are strongly hydrofillic water par ticles soaked by acrogel and absorbed in its structure tend to degrade its transparency and its optical properties For this reason it is important to develop 1 a cutting technique which preserves the integrity of cutted pieces of the tiles and their optical properties 2 a baking procedure which elim inates as much water as possible from aerogel and finally 3 a post baking procedure which minimizes the air exposure of the tiles until their installation in the detector 2 The aerogel Cherenkov diffusion box pro totype The aerogel prototype detector 1 2 depicted in Figs 1 and 2 consists of six 14 stage 5 BURLE 8854 Quantacon photomultiplier tubes 3 arranged in two rows of three phototubes opposite to each other collecting the Cherenkov light produced by a 40x40x9 cm acrogel layer connected to a diffusion zone of 45x45x20 cm All inner surfaces of the detector were covered with Millipore paper as a reflector A cross sectional drawing of the detector is shown in Fig 1 and a schematic view of the experimental set up is shown in Fig 2 The present measurements have been taken with acrogel produced by Airglass Inc with ref
4. JLAB TN 00010 April 2000 Test and development of a Cherenkov diffusion detector prototype using Airglass aerogel at AF L Lagamba R Iommi B Wojtsekhowski Test and development of a Cherenkov diffusion detector prototype using Airglass aerogel at TJNAF Luigi Lagamba Dipartimento Interateneo di Fisica and Sez INFN Bari Italy Thomas Jefferson National Accelerator Facility Riccardo Iommi INFN Gruppo Collegato Sanita Roma Italy Thomas Jefferson National Accelerator Facility Bogdan Wojtsekhowski Thomas Jefferson National Accelerator Facility November 18 1999 Abstract In this note some operational procedures applied to aerogel tiles are described These tiles are currently installed in the Cherenkov diffusion box prototype built by INFN groups of ISS Roma Bari and Lecce and used to aid the design of a bigger detector to be installed in Hall A Particular attention is focused on the baking and cutting procedures of the tiles Moreover the analysis of data obtained with baked acrogels shows a significant improvement in the average number of detected photoelectrons when compared with results from unbaked tiles A factor 4 increase in the number of produced photoelectrons is obtained 1 Introduction Since the early 1980 s silica aerogel has been widely used as a radiator for Cherenkov detectors Its low refraction index and density allows particle identification in experiments where gaseous radiators have been
5. ad several tiles each 3 cm thick but with greater square area So in order to cut tiles with the desired dimensions we used the AB Marvel band saw located in the High Bay Area of the EEL Building Before the cut each aerogel tile was put on a cardboard sheet to avoid con tact with the saw working table The saw is mostly used to cut aluminium or metal foils and on its table surface a certain amount of oil coolant water and metal debris can be present even after an accurate cleaning In addition at the beginning of the operation each worker should wear appropriate pro tective clothing gloves protective glasses and optionally a mask to protect from silica powder even if silica particles are declared as being non toxic and not harmful to lungs 6 To minimize contamination a vacuum cleaner was used while the saw was moving to capture powder thrown in air by the saw itself After the cuts the saw and the environment around it was to be cleaned One might follow the checklist below during the operation e Turn cooling off using the switch located on the front panel e Close the coolant circuit operating the switch located near the band itself This is not to be done while the saw is moving e Switch on the saw and let it run for a few minutes in order for the residual coolant to fall down After if possible switch off and clean the band and the work table with dry towels Do not use water The latter operations requires 1 2 hour or
6. an be seen in the following table Mean of Distribution A A Single Photoelectron Peak Aes it is wise to arrange things in order not to exceed one hour of total work time inside 3Looking forward at the installation of the bigger detector a new and more performant dehumidifier has been recently acquired 1 010 0 990 1 105 1 015 0 985 1 218 1 025 0 976 1 186 1 055 0 948 1 119 where n is the minimum velocity to produce light in the aerogel Omin 0 9941 is the minimum 3 for a muon to traverse the 2 lead filter and f is the correction factor for each 8 5 Results In the following pictures some example spectra according to the analysis described in the previous section are presented In figure 5 6 and 7 the con tribution due to the single six phototubes is shown In figure 8 9 and 10 the sum spectrum belonging to each run each one presenting different conditions For the August 14th run an unbaked aerogel layer was installed while for the August 21st run we used the aerogel baked in the way described before As one can easily see although the absolute result in the photoelectron number is not a smash the improvement is impressive a factor 4 in collected p e s confirming the usefulness of the baking procedure Before the August 24th run additional millipore paper strips were added on the tray which contained the aerogel covering the corners of the layer a not interesting area since it was excluded by the
7. e not baked Figure 4 Transmittance spectra measured every 10 nm between 190 and 900 nm for the Airglass aerogel sample before and after a 32 h baking 3 3 Aerogel installation Since Airglass aerogel tends to absorb a great amount of water in a relatively low time interval it is wise to protect the layer to be installed from air exposure therefore from humidity exposure as much as possible For this purpose as a dry room was not available an ad hoc technique was developed to protect aerogel consisting in the following steps e after the baking process the aerogel tiles were covered with properly bended stainless steel plates eventually sealed with duck tape This was done inside the oven itself since after the baking the temperature inside is well above the ambient temperature and it takes a sufficiently long time to make a thermal equilibrium between the oven and the environment A relatively high temperature helps limiting the amount of water soaked by the tiles e A little dry chamber was built using flexible and transparent plastics to separate the installation area from the external environment The latter showed to be a good procedure However since the air circulation inside the little chamber is somewhat limited and human bodies exploit oxygen in similar conditions e The air inside the chamber was made dry by means of a small automatic dehumidifier The relative humidity during the whole installation was mon
8. itored by an hygrometer It never exceeded 42 while outside it fluctuated around 50 e The installation of each needed acrogel tile was performed using gloves and masks since the human body might emit water particles just for the act of breathing 4 Data analysis In order to obtain the average number of photoelectrons the charge spectrum coming from each phototube was equalized and calibrated cach pedestal was shifted to channel zero and each spectrum was multiplied by a different factor to let the one photoelectron peak to lie on channel 100 From the summed PMT spectra the average number of photoelectrons can be computed as 1 1 where A A y is the ratio of a single photoelectron peak to the mean of a single photoelectron distribution 3 and for our data is equal to 1 12 Of course we could distinguish between two results the one obtained by requiring a trigger between the first two scintillators and the second in which also the third scintillator was allowed as a trigger The same software cuts described in ref 1 were applied in obtaining u From the raw results for high energy events 0 35 photons must be subtracted as a correction this term comes from the Cherenkov light produced on air and millipore paper inside the detector 1 Finally as described in ref 1 the results were extrapolated to a 8 1 particle beam to consider cosmic rays 8 distribution This extrapolation requires a multiplying factor of 1 039 as c
9. port R I would like to express his gratitude to Evaristo Cisbani and Mauro Iodice INFN Sanita Roma for their constant and patient help George Lolos and Rob Van der Meer University of Regina for the useful discussions about the cutting technique References 1 J McCann L Taub and B Wojtsekhowski Aerogel Cherenkov Counter Calibration Using Cosmic Rays 2 R Iommi Test of a Diffusion Aerogel Detector at TJNAF 3 M Shepherd and A Pope Investigation of BURLE 8854 Photomulti 8 plier Tube Thomas Jefferson National Accelerator Facility JLAB TN 97 028 1997 4 C Caso et al The European Physical Journal C3 1998 5 Particle Data Group Particle Physics Booklet 1998 6 B Kross Private Comunication 7 Despatch Air Circulation Oven User s Manual 16
10. ractive index n 1 025 The parameters of the detector have been studied using cosmic rays in order to produce Cherenkov light in aerogel With a rather large aerogel square area the behavior of the parameters of the detector smoothly depends on the position of incoming rays so there was no need to study details with a pencil beam Even if the cosmic ray rate is not comparable to that of a particle beam the large square area allowed the collection of good statistics in a relatively low amount of time 24 hours As shown in Fig 2 an upcoming cosmic ray to be detected had to trans TOP VIEW PMT 1 Space for PMT 2 Aerogel EEE PMT 3 Figure 1 Top view of the aerogel detector prototype The central chamber con tains a 40x40x9 cm Airglass aerogel layer with refraction index n 1 025 Cosmic ray muon SIDE VIEW Trigger 1 C E pzz Space for Aerogel A yyy PMTs 1 3 ae PMTs 4 6 CA Trigger 2 Space for y Millipore i A 2 Pb filter 7 C T Trigger 3 Figure 2 Cross sectional view of the detector with the used cosmic ray setup The third trigger scintillator fired only for very energetic gt 810 Mev muons Minumum p vs Lead Thickness Muon Momentum MeV c e Ce LE ami Ps ea CC 100 200 300 400 500 Lead Thickness cm ofe e teed Figure 3 Muon momentum as a f
11. rogel tiles remove a 3 of the weight and improves the light output by a factor 4 In addition we have seen that technique must be developed since baking to 300 Celsius degrees removes a water fraction equal to 3 of the tile weight and after baking the tiles must be opportunely preserved because they acquire weight 13 ral fad n Figure 10 Sum spectrum for baked aerogel with additional millipore paper The increase in diffusing surface carries out a relative improvement in the number of p e s Figure 11 Individual contribution from each phototube having removed the mu metal shield from the detector 14 Figure 12 Sum spectrum for baked aerogel additional millipore paper and re moved mu metal shield The effect of magnetic field results in a decrease of pho toelectron number if exposed to the environment However more study of baking effects such as improvement in light transmission need to be done together with the use of a dedicated saw for the cutting and a professional dry room for the detector assembly 15 8 Acknowledgements The authors wish to thank Franco Garibaldi INFN Sanita Roma who initiated the development of the PID detector for Hall A s kaon experiment L L and R I would like to thank Kees De Jager for his hospitality at TJNAF Bert Manzlak and Brian Kross TJNAF for their advice and sup
12. unction of lead thickness A choice of 60 cm corresponds to 800 Mev c muons verse the first trigger scintillator the detector itself in which it originated the production of Cherenkov light the second trigger scintillator 2 61 cm thick lead filter and finally the third trigger scintillator The electronic chain used in this setup is described in detail in ref 1 Basically only the coincidence signal of the first two trigger phototubes properly timed pro vided the gate signal The third trigger PMT was used to select high energy events during data analysis i e to select those cosmic rays which passed through the lead filter 1 Cosmic rays at sea level are composed mostly of muons which have a mean energy at the ground of 4 GeV 4 In order to find the minimum momentum of muons traversing the lead filter one can refer to the muon energy loss rate for ionization only in lead which depends rather strongly for very low momenta to the incoming particle momentum 5 Then the maximum lead thickness corresponding to a given muon energy or momentum can be easily calculated carrying to the results of Fig 3 A muon momentum of 800 MeV c energy 810 MeV corresponds to 2 lead thickness 3 Operational procedures 3 1 Cutting silica aerogel Silica aerogel is extremely fragile therefore particular care must be used in handling and obviously cutting it For our purpose we needed a 40x 409 cm acrogel tile and we h
Download Pdf Manuals
Related Search