4. HADRON CALORIMETERY The hadron calorimeter (HCAL) surrounds the electromagnetic calorimeter and acts in conjunction with it to measure the energies and directions of particle jets, and to provide hermetic coverage for measuring missing transverse energy. The CMS calorimeters must satisfy the following requirements: maximum counting rate comparable with 25 ns time interval between bunch crossings at high luminosity $1034cm-2s-1$ non-magnetic absorber must be used not to distort the uniformity of the magnetic field; due to magnetic field in the place where the calorimeters are situated, the photodetectors must be put outside of the magnetic field or be insensitive to it; minimum dead space to have high hermeticy to measure missing transverse energy. Effects due to clustering algorithms, magnetic field, pileup and noncompensating nature of the calorimeters (due to different structure of electromagnetic calorimeters) degrade two jet resolution, therefore, the HCAL energy resolution is not a crucial issue. The radiation dose in the hadronic calorimeter according to calculations is not large at a rapidity of Thus the option based on plastic scintillator is reading out with WLS fiber embedded in the scintillator plate proposed in [4.1] meets all the requirements. The design allows to make projective structure with minimum cracks, easily divide the calorimeter in longitudinal direction to measure the shower development. 4.1 Conceptual design The main responsibility of RDMS is the endcap HCAL (HF). The structure of the endcap HCAL is shown in Fig.4.1. Endcap hadron calorimeter must cover a rapidity region between 1.5 and 3.0 with good hermeticy, good transverse granularity, moderate energy resolution and a sufficient depth. A lateral granulirity chosen is shown in Fig.4.2. The basic structure of the endcap calorimeter is the same as for the barrel calorimeter. Electromagnetic calorimeter is placed in front of HCAL for The hadron calorimeter consists of 9 copper plates 5 cm thick then 11 plates 10 cm thick and the last plate 8 cm thick. The total thickness of the calorimeter ( without electromagnetic calorimeter) is 10.8 absorption lengths (20 active layers). The expected energy resolution for this structure is Each hadron endcap is made up of 18 wedges (14 tones per wedge) that cover and close one end of the barrel calorimeter. The wedges are constructed of copper plates, separated by staggered spacers, that are perpendicular to the beam axis. The first and the last plates are made up of stainless-steel for fixation of electromagnetic calorimeter and both of them on the endcap steel support. Scintillating plates are mounted on plastic plates forming the "pizza pan" structures as shown im Fig.4.2 and are installed in gaps of copper absorber. The construction allows easy access to the pizza pans and provides a rigid structure. The special concern for the endcap calorimeter is the radiation environment. The dose rate presented in the Fig.4.3 imposes severe requirements on radiation hardness of the calorimeter, in particular in the range of pseudorapidities 2.6 - 3. It leads to necessity to use radiation hard scintillators or possibility to replace easily the damaged tiles (the design must take into account this possibility). Additional R&D work is necessary to develop rad hard scintillator and WLS fibers. 4.2. Active elements It is assumed that the integrated luminosity over the first ten years of LHC operation will be 5 x $105pb-1$. The ten-year integrated dose is estimated to be 1 kGy (0.1 MRad) at the front of the HB. The maximum dose expected at a rapidity of 2.5 is estimated to be about 20 kGy (2 Mrad). It is known that up to 30% peak damage in the HCAL will not induce a constant term in the energy resolution which is unacceptable. In common with most commercial polystyrene based scintillators, SCSN81 together with a K27 doped WLS fiber such as Kuraray Y11, suffers a light yield reduction of about 30% at 10 kGy (1 Mrad) and an unacceptable 70% at 50 kGy (5 Mrad). The baseline HCAL design uses this combination in the barrel region. In Table 4.1 the performance of this combination is compared with that of several other options. In the endcap region up to either SCSN81 or Kharkov scintillator with Y11 doped fiber is used. Between a green scintillator (3HF) with an orange WLS fiber (O2 Kuraray) could be used. However the low number of photons, combined with the longer decay time constant and the instability of 3HF under illumination, make this a difficult choice. The new Bicron blue scintillator BC499S2 together with doubleclad Kuraray Y11 WLS fiber appears to be a promising combination for the region between . The scintillator produces 70% more photons than SCSN81 and shows a light yield loss of 35% at a dose of 50 kGy (5 Mrad). The light yield in the endcap elements may turn out to be smaller than in the barrel. However, since most of the physics resides in transverse energy, the stochastic term in the resolution, which scales as improves with decreasing angle. This effect will more than compensate for the loss in photoelectron statistics. For the region in the front-endcap ring, yearly re-masking and/or periodic replacement of the scintillator has to be envisaged. An important consideration is the decay time of the scintillator/fiber combination. Typical decay times for blue scintillators with K27 doped fibers are about 10 ns. This means that after 25 ns about 30% of the photons are still to come. There exist newer fluors such as Bicron's G2 which when combined with a blue scintillator lead to a decay time of about 4 to 5 ns so that a much larger fraction (85%) of the light will be in the 25 ns interval. This fluor, in the liquid state, did not show any radiation damage at 1 MGy (100 Mrad). However the damage was worse, amounting to 40% at 20 kGy (2 Mrad), when used with SCSN81. Radiation damage at low dose rates can be substantially worse than at high dose rates. A major effort is being made at the University of Michigan where some samples have already been irradiated for over one year. This will provide an indication of the damage that is expected over a ten-year period. Similar work has been started at the Institute of Nuclear Physics, Tashkent. IHEP (Protvino), NSCKIPT (Kharkov) and others together with industry are conducting the R&D program to develop radiationhard blue scintillator and fast, radiation-hard WLS fiber by exploring different polymerizations of the plastic. 4.3 Photodetectors. The HCAL photodetectors which convert the optical signals from the 10 mm² fiber bundles corresponding to a tower are required to have a linear dynamic range of $105$ and operate in a uniform 4 T magnetic field. For calibration purposes, the detectors must have the capability of measuring the signal generated by a radioactive source as a DC current to a precision of 3%. In addition, the photodetectors are located inside the detector, adjacent to the HCAL itself, where service access is infrequent thus placing an additional requirement on mean-time-to-failure. The useful lifetime of the photodetector must correspond to ten years of operation at a luminosity of $1034cm-2s-1$. A final requirement on the ratio of the signal to noise follows from the need to measure the signal from a minimum ionizing particle. Progress is being made on the development of two types of photodetectors that can operate in magnetic fields and still provide gain. These are the proximity focused hybrid photodiode (PFHPD) and the semiconductor avalanche photodiode (APD). Proximity Focused Hybrid Photodiode The proximity focused hybrid photodiode is an image intensifier operated in the electron bombardment mode. Photoelectrons emitted from the photocathode are accelerated by an electric field and stopped in a silicon diode target where electron-hole pairs are produced generating the signal. In the device under consideration, the 10 kV electric potential is uniform and the acceleration gap is only 1.5 mm to minimize magnetic field effects. Commercial devices are presently available in standard 18 mm or 25 mm diameter single channel versions. Prototypes have been made in which the diode is subdivided into pixels to make a cost effective multichannel device suitable for reading out fiber bundles corresponding to a number of calorimeter towers. High gain prototype pixel devices using avalanche photodiode targets are also under evaluation. PFHPDs exhibit a gain that is linear with applied voltage, 2000 at 10 kV. In beam tests the gain has been measured to decrease by only 2% in an axial field of 3 T of the RD5 magnet [4.2]. The devices are linear to 2% over the required $105$ dynamic range and exhibit a fast response that is determined by the diode source capacitance. The outstanding questions for these devices are use of fiber optic windows, development of non-magnetic packaging, and reduction of the dark current to levels suitable for measurement of the DC current signal from the calibration sources and of the signal from a minimum ionizing particle. Semiconductor Avalanche Photodiode Large area APDs are under intensive study for readout of the ECAL crystals. For HCAL, these photodetectors could meet the requirements of large dynamic range, operation in a 4 T magnetic field, and precision DC current capability. The device gain is lower than the hybrid photodiode option being only 50 -100, but this is compensated in part by the higher quantum efficiency. Investigations are underway to determine whether the signal to noise and DC current measurement specifications can be met with these devices. Several tens of these APDs have been acquired for beam tests. The issues of gain stability and sensitivity to neutron irradiation for these devices will be addressed in collaboration with the ECAL photodetector development group. APDs with associated preamplifiers are also under development in Institute for Nuclear Research, Moscow and Minsk institutions. The characteristics of these photodetectors will be compared to the presently available commercial devices. 4.4 Front-End Electronics and Services The electronic readout system of the HCAL is based on the CERN FERMI system. The dynamic range requirement (20 MeV to 2 TeV) is similar to that of the ECAL. The HCAL group intends to profit from the adaptation of the FERMI system for the ECAL. Electronics boxes containing the decoder/mixer boxes, the photodetectors and associated HV supplies, as well as their preamplifiers and their low voltage distribution, will be distributed around the outer radius of the transition region from barrel to endcap, close to the HCAL detector itself. The FERMI system will also reside in this region. They will be attached to either the barrel or the endcap and will be able to move along with their own sub-detector. Source driver boxes for both the endcap and the barrel also sit close to the coil in the region. The barrel and endcap are serviced via the 100 mm gap between the two sub-detectors in the region. This region also contains cables from the electromagnetic calorimeter and the tracking detectors. Hadron calorimeter-related services include optical cables from the barrel and endcap megatiles, source tubes servicing each of the megatiles, and possibly quartz fiber bundles transmitting laser signals to each of the individual tiles of a megatile. The electronic boxes and source drivers are connected to the outside world via a cable path that snakes around the barrel to reach the outside center of the detector. 4.5 Calibration and Monitoring. The uniformity of response must, to first order, be assured by the construction and quality control. Experience of CDF and SDC shows that the uniformity of the tile/fiber assembly can be maintained at the 10% level for a large scale production. The assembly can be monitored by radioactive source and by injecting light from UV lamps. Absolute calibration and linearity of the calorimeter will be established by exposing several modules to a hadron test beam. The calibration can be transported to the CMS detector using radioactive sources. Both the quality control function, and the transfer of test beam calibrations to other similar towers, require the incorporation of source tubes crossing every scintillating tile, as in the SDC design. It is assumed that maintenance access to the source tubes in most layers will be available only when the endcaps are withdrawn. We do not intend to calibrate each wedge in a test beam. Instead, we intend to transfer the absolute calibration from a number of wedges exposed to a particle beam to all wedges using the radioactive source calibration. This scheme is currently under test. Radioactive Sources All layers of the hadron calorimeter will be equipped with thin stainless tubes that will route radioactive sources throughout the system. This is a system similar to the one used by CDF and proposed by SDC. A wire with a pointlike Cs source will be pushed through these tubes by a remotely controlled system of drivers. The DC current induced by the source traversing one tile of a tower will provide an accurate measurement of the response of the entire measuring chain. The experience of CDF shows that this measurement can be maintained at the level of 1%. Change of response due to photodetector or electronics will show up as a change of the response of all tiles of a given tower and can be compensated by an adjustment of the overall calibration factor. Change of response due to radiation damage will lead to a change of the measured current that is dependent on the depth of the layer. In order for the moving radioactive source calibrations to be done periodically during collision runs, it will be important that the front-end electronics allow simultaneous digitization of the DC current from the photodetector and of fast pulses, both beam-related and from laser light injection. Photodetectors with high leakage currents would compromise the ability to do the source calibrations. The most convenient location of the source tube drivers is in the transition region between the barrel and endcap. However, this region is in the 4 T magnetic field and conventional driver motors may not work. Piezoelectric wave stepping motors may do the job, and even aircore conventional motors might work if the axle is aligned with the external field. Laser Light Calibration System The main objectives of the HCAL laser calibration system are: to control optical system (fibers, splices, optical connectors etc.), to control stability and linearity of response of each channel (photodetectors, preamplifiers, ADC, HV supply), to control timing of each channel. A relative calibration of the calorimeter can be realized with a nitrogen laser. Fig.4.4 shows the layout of such system. A laser is triggered during the low intensity beam crossing. Part of the light is sent to a scintillator where it is shifted into visible light and detected by a stable photodetector such as PIN photodiode to control the laser output and the timing. The rest of the light goes to a quartz neutral filters to vary the light intensity (about $104$ orders of magnitude) in order to measure the linearity of the channels. To realize the scheme the following requirements on the laser are imposed: the trigger timing is about 10 ns; the light output (for 1 % precision calibration during 1 pulse) is 10000 photons X 10100 (number of towers number of longitudinal segments), i.e. $108$ photons. Taking into account the quantum efficiency of photodetectors and the losses in the distribution system the final number is about $1010$ photons which corresponds to 5 x $10-9$ J. The commercially available lasers satisfy these requirements. We are going to use a quartz fiber 0.3 mm diameter for control system with the next prototype. A study is under way to measure the radiation hardness of the fiber up to 3 Mrad with 0.5 % precision. Calibration Using Data Suitably chosen calibration triggers can be used to monitor the overall stability and/or absolute energy scale of the hadron calorimeter. For example, minimum bias events can be utilized to maintain uniformity of response and to monitor its time stability. Photon/Zjet triggers can be used to provide calibration and the absolute energy scale. 4.6 R&D plans for endcap hadron calorimeter The endcap HCAL engineering plan includes a set of tasks necessary to refine and complete the baseline design. Each of these tasks will require decisions based on analysis of performance and cost. Performance issues are being formulated by particle beam tests and computer simulation. Manufacturing costs are derived from building prototype components in the industry. Full Scale Sector Prototype of HF Instrumented 1/18 full scale prototype detector will be built and tested in 1995-1997 to verify the mechanical design, to gain experience in assembly and to study performance. The basic method for producing the hadron absorber structure outlined in this report is to stack and weld machined copper plates by electron beam welding. Options still exist and are being pursued in parallel with detailing the design of the baseline proposal. Additional studies are being planned in order to understand stress and deflections, mold design, fabrication and support connections with endcap walls. 'Hanging File' Prototypes for Test-Beam Studies Two reconfigurable copper/scintillator calorimeter modules are available for further evaluation and optimization of the HCAL design and both will be exposed to test beams at CERN during 1995 - 1996. Both allow the response of every scintillator layer to be separately recorded, so that the effects of varying the sampling fraction and longitudinal segmentation can be studied. Material representing the coil can be inserted to help determine the optimum design and the effective benefit of the tail catcher. Effects at the wedge boundaries and optimization of the step can be investigated by inserting spacers between the copper plates. One prototype designed and constructed in US makes it possible to study lateral development of hadron showers. The other prototype designed and constructed by RDMS fits inside the RD5 3 Tesla superconducting magnet and will allow the study of the response in fields parallel to the beam up to 3 Tesla. This prototype combined with PWO crystal ECAL and full scale prototype of the forward muon atation MF1 will be used in the integrated test. The RDMS "reconfigurable-stack calorimeter" which contains brass plates as absorber has been built using as the active medium scintillating tile and optical fiber readout. The calorimeter consists of 3 light tight boxes. Two boxes contain stainless steel support frame on which absorber plates and scintillator plates are hanging. The dimensions of the plates (66 X 66 cm²) are determined by the size of the hole in the magnet RD5. The active element consists of 3 x 3 tiles corresponding to 3 x 3 towers. The tile is 4 mm thick scintillator with dimensions 22 cm X 22 cm. 1 mm diameter wavelength-shifting fiber doped with K27 is routed through a key-hole shaped groove milled in the surface of the tile. One end of the fiber is machined by flying diamond cutter and aluminized. The reflectivity of the fiber mirroring is about 80 %. The tile with the fiber is wrapped with aluminized mylar. To the WLS fiber a 4 m long clear fiber is connected to guide the light to a photodetector. The efficiency of the light collection was measured with and photomultiplier FEU-85 with q.e. about 8%. The number of photoelectrons (p.e.) was equal to $N$p.e.=1 (confirmed by measurements with muons) and uniformity with r.m.s. about 5 % was achieved. The fibers from each tower are bundled together and put in a tube. For the measurements in magnetic field proximity focused Hybrid Photo Diodes with 25 mm diameter were used. The HPD outputs were directly connected to preamplifiers and than by 60 m cables to the control room where LeCroy amplifiers were used to equalize the HPD gain. The HPD HV tension was set to 8.08 kV, the preamplifier gain was 0.125 V/$106$ e. The gate length was 150 ns corresponding to the pulse length from HPD. Active Elements The effect of magnetic fields on the response of various scintillators will be measured. Since we expect to take data for ten years or more, we are concerned about the long term stability of any materials used in the construction. We will therefore set up a study of long term stability of new materials. This study will include accelerated aging tests at elevated temperatures, effects of humidity and room lighting, and effects of various handling methods. Photodetectors Tests of only a few samples of the types of photodetectors described above have so far been carried out. 36 APDs were tested with lead tungstate crystals in 1994. Nine PFHPDs were used for the test of the prototype in the 3 Tesla RD5 magnet in 1994. After the beam tests these devices were extensively bench tested (e.g. linearity, dynamic range, etc.). The effect of radiation damage on their performance and more detailed studies in magnetic fields will be carried out. At present these devices have a high cost per channel. The volume production costs still have to be established. A multipixel capability has been developed by DEP to reduce the channel cost of PFHPDs. We intend to pursue this R&D vigorously. 4.7 Cost estimate of Endcap HCAL and RDMS industry involvement The cost estimate of the Endcap HCAL is 10,1 kCHF according to CMS Technical Proposal. For the Endcap HCAL Uzbekistan will provide 500 tons of raw copper. Manufacturing of the plates and assembling of the calorimeter absorber will produce by "Mikoyan" and "Fr.Vyborzhez" plans in Moscow and St.Petersbourg. Ukrainian NPO "Single Crystal" and Kharkov Electro-Mechanical factory are interested in the production and machining of the plastic scintillator tiles with the grooves for WLS fibers. The collaboration has the close contact with these and other organizations. REFERENCES [4.1] V.Kryshkin and A.Ronzhin, Nucl. Instrum. Methods, 247(1986)583. [4.2] G.Bayatian et al., "Study of magnetic field influence on hadron calorimeter response", CMS TN-94-222