 |
|
|
| |||
Calorimeter draft answeresFrom: CommentsDear LHCb colleagues, Here are the outline of answers presented to the LHCC
referees during the meetings on Tuesday next week. Please e-mail any comments to
Jacques Lefrancois ---------------- OUTLINE OF THE ORAL ANSWERS TO THE CALORIMETER-REFEREE QUESTIONS Question: The main point concerns the overall design of the calorimeter system. It seems to me that some attention may still be paid to its optimization by stressing that the three calorimeters (preshower, ECAL and HCAL) must act as parts of a whole system. This could bring to better performance without (or with limited) additional investments. Answer:After having arrived at a plausible set up for the Preshower(PS) ,the Ecal,the Hcal, for the technical proposal we will now concentrate before the TDR on calorimeter optimisation taking into account the necessary system integration. Question:As a practical example, to trigger the discussion on this subject, one could envisage a design like the following one: 1) reduce the instrumented depth of HCAL (to 3-4 lambda?). Keep the same iron thickness for muon filtering. Answer:We have started work on this:Hadrons have been generated at different energies (as function of angles) for a fixed Pt of 2.5 Gev/c.The trigger efficiency was studied for a Pt threshold of 2.2Gev/c The Calorimeter lenght is varied from 3 to 8.5 lambda.The calorimeter response was recalibrated so that the most probable energy response is defined as 2.5 Gev/c. We plan to produce before the JULY LHCC a more complete study by comparing (in a full geant geisha simulation as above)the trigger efficiency of B->pi,pi and B->DsK events vs minimum bias event rejection for two set up :the TDR one with 7.3 lambda and a new one with 4.5 lambda. Question: 2) Increase the longitudinal readout segmentation for ECAL (e.g. by suitable fibre grouping). One could foresee readout at 1 X0, at shower maximum (where the highest dose is released) and at shower tail. 3) The above design does not foresee a separated preshower, whose task would be accomplished by the first longitudinal readout sector of ECAL (at 1 X0). However, the type of photodetector could be different for different sectors. Answer: The above two questions adress the problem of the Preshower-Ecal optimisation.(Since a segmentation in depth in the later part of Ecal while increasing rapidly (doubling?) the PM and electronics cost would bring only a marginal improvement in e/pi trigger rejection compared to the present design) Therefore the main question is to find the most cost effective PS solution. We are pursuing this, however no early(July) result of these studies are foreseen. Question: 4) Pointing or quasi-pointing geometry for ECAL and (possibly) HCAL can be envisaged. The gain in the performance must then be evaluated costwise. Answer: Some experience exist (Delphi) on construction cost increase in Shaslik calorimeter from non pointing geometry to a perfectly pointing geometry (it can be as large as a factor of 2) In view of the vast increase in complication and the modest,to our understanding,gain in physics capacity,we do not foresee to pursue this avenue. However it is clear that the pad chamber ,preshower,Ecal,Hcal cells need to be associated in the trigger.therefore a scaling and a correspondance of cell size has to be implemented.We intend to take this into account in the next mechanical designs.We foresee no difficulty in doing this. Question:i) HCAL is not compensating. This implies e>>p and a (trigger-) bias could appear in the pt measurement, given the energy dependence of the e.m. content of hadronic showers. Answer: The hadronic shower simulation take into account the pi0/pi+- fluctuation and therefore the tail produced by the e/pi effect; We therfore think our simulation takes the e/pi effect into account and there should be no extra bias. ( by the way dear Cal colleagues I thought after the meeting that I did not check if this is taken into account correctly in the Ecal Hcal sum: in priciple an electron of 2. gevs will be seen as a (e/pi)*2 Gev hadron .For typical e/pi values (1.2?) a 2gev electron will pass our 2.4 gev hadron threshold this is contradictory with a 2.34 GEV/c electron threshold which implies that lower energy electrons are rejected. Of course this can be cured by asking that hadrons leave always some energy in Hcal... please Tatsuya can you delete the 8 line parenthesis before forwarding the message to the colaboration.) Question: ii) Pizero reconstruction capability is an important issue for a dedicated experiment like LHCb. How is the calorimeter design optimized for the pizero detection ? (e.g. module x-section, geometry, longitudinal segmentation,..) Answer:The pizero optimisation is one of criteria used for the Ecal
optimisation. For example studies have starded comparing in the ro+pi- and the
ro0pi0 channels efficiency and background rejection (i.e.pizero mass resolution
and Bmass resolution) for two cell size choices :the present one with inner
=4*4cm middle=8*8 cm and outer=16*16cm ;and a new one with inner=6*6cm and
middle and outer= 12*12cm resulting in 25% less channels. It is not foreseen
that a detailed answer can be available for July. Question: iii) How good and
how critical is the calorimeter(s) hermeticity, given the constraints of a
realistic engineering design ? Answer: The crack size among modules is from a
fraction of a mm to 1 mm leading to very small effects (even more so because of
non pointing geometry!) A plot of a scan of a HeraB Ecal module crack can be
produced for July.The only foreseen delicate point is the vertical middle
separation .Studies are continuing on this to make sure the cracks is <
| ||||||||||