Minutes of the CALCOM analysis meeting August 15, 1997. ======================================================= Agenda: A. Vlassov - New analysis of CLAS Cherenkov detector response to electrons B. Niczyporuk - Comments on reconstruction efficiency in SDA (no report received) V. Burkert - Comments on CLAS commissioning run September/October ================================================================== Comments: V. Burkert Two analyses of the Cherenkov counter response to electrons were presented (see A. Vlassovs' report). The most reliable electron identification is obviously achieved in elastic scattering and only this process should be used for such an analysis at low energies. Using these it was shown that -by and large- the detector response is close to what was expected from the GEANT simulations. However, significant deviations seem to be present at very forward angles. These could be due to the difficulties in simulating all details of the support and shielding structures (which was not even attempted) which may obstruct some of the detector acceptance at small angles. Also, the distribution of photo-electrons as a function of phi within one Cherenkov element showed somne deviations. The results were presented in terms of #photo-electrons. As these numbers depend on an accurate knowledge of the single photo-electron peak a direct check of the results is highly desirable. This could be achieved by the direct measurement of the efficiency = #elast. electrons with matched CC Ph > 0 / #elast. electrons in same bin ====================================================================== Individual reports: =================== A. Vlassov: ----------- The results of Cerenkov efficency analysis was presented. It was based on two possible electron identifications : 1) Electron elastic scattering peak (hit-based tracking in recsis) 2) eid0 electron identifiations (using EC, SC, CC matched hits). In both cases the run 3166 was analysed (1.6 GeV initial energy). The efficiency was estimated by comparison of the obtained mean number of photoelectrons with GEANT estimated one. It was shown, that: * At zero approximation the characteristics of Cerenkov detector are very close to estimated in GEANT simulations (in term of mean photoelectron number) althow it will be useful to increase HV values to see well-separate one_photoelectron peak. * Two different types of electron identification shows mainly the same result. The estimated number of photoelectrons at first segment in case of eid0 identification is larger, than predicted by GEANT simulations. Recent investigations by Lee Cole Smith (see http://apollo.phys.virginia.edu/clas/june1.html ) shows that the source of that is gammas and electrons from electron interaction with the support structure around beam pipe. * At large angles ( large Cerenkov segment numbers ) the event statistics is very small. It looks like the mean number of photoelectrons is diminished here with respect to GEANT estimations. Possible reasons were discussed, but it is evident, that much more statistics needed to prove one of them. It was mentioned, that the direct investigation of the detected fraction of elastically scattered electrons is very important. Transparencies presented are shown at: http://www.cebaf.gov/~vlassov /cc/e_rec.html Home page : http://www.cebaf.gov/~vlassov/ V. Burkert ---------- Thoughts on the CLAS commissioning run September/October: This will be the last commissioning run for CLAS. By the end of this run and the subsequent analysis period we need to understand the most important parameters of CLAS. The boundary conditions during the run will be as favorably as they can be: - Hall B is the only hall receiving beam and we can thus define the operating conditions (within the limits given by the machine operation) without interference from other users - We will have the low current (1 nA) beam monitors installed and hopefully operational - The drift chambers will have undergone major repair work and should be in best possible shape - The cryo target has been surveyed. A new cell design is being worked on and should be in place. This new design is expected to minimize halo-related background produced in the target support structure - Additional shielding is being prepared to reduce the large fraction of false calorimeter triggers. These appear to be due to mostly neutral energy from secondary interactions in magnet support structures -------- I propose that we treat this run as if it were a physics run! -------- This implies that we have all the controls in place to tell us if somethings goes wrong, that all relevant information is recorded (this was e.g. not the case in all previous runs!!) and we are able to extract quantities such as the elastic and inclusive inelastic cross section during the run. As a goal I propose we aim for a 10% accuracy in determining these quantities during the run. This will be the ultimate test of the performance of the entire system, the beamline, the detector, the data acquisition system and the offline reconstruction software. We need to have the calorimeters, TOF counters, and Cherenkov detectors reasonably well calibrated using the June run data and cosmic ray data. There is much beam data that can be used e.g. for min.ionizing calibration for EC and TOF counters (If someone is looking for a project in this area I would be happy to suggest specific runs and projects for the analyses). Also, we need to train the shift leaders and shift workers better so that they know what their responsibilities are, what tools they have available to judge the quality of a data run, etc.. and we need to define procedures that every shift must comply with.