Minutes of CALCOM Meeting Friday August 14, 1998 ================================================ Agenda: Patrizia Rossi - LAC gain monitoring system results Cole Smith - Elastic ep scattering analysis Stepan Stepanyan - Tracking efficiencies in ep => eppi+pi- Kim Egiyan - Analysis of empty target data from e1 run at 1.6 GeV ---------------------------- Postponed till next meeting: Volker Burkert - Status of missing mass resolutions in inclusive and exclusive reactions Costy Loukachine - Time-of-flight offsets ------------------------------------------------------------------------------- Comments: I. There seems to be now a reasonable understanding of the tracking efficiencies for events with few particle in the final state. This comes from various sources (1) The extracted inclusive and exclusive elastic ep cross section are very close to the world data and the difference in inclusive vs exclusive indicates a 95% tracking efficiency for elastic protons (Cole Smith). (2) The same conclusions can be drawn from exclusive reactions involving pions and protons covering large kinematical regime (Stepan Stepanyan) (3) Similar conclusions were drawn from the analysis of Joe Manak and Richard Thompson (see Joe's minutes of clas_offline meeting) II. Cole Smith's analysis also indicates that electron acceptances may be parametrized with a simple angle and momentum-dependent phi acceptances that uses reconstructed momentum vectors only. The agreement with the world data appears to be at the level of ~5% now (my conclusion!, V.B.). III. Kim Egiyan presented clear evidence for strongly enhanced event rates in the 1.6 GeV data coming from the downstream target cell window. The data maybe explained by water-ice buildup on the aluminum windows. This finding also explains (qualitatuvely) the increase in empty target rate observed during part of the e1 running. However, there is no reason for being alarmed, the data can still be analysed. Volker Burkert ------------------------------------------------------------------------------- Individual contributions: ========================= Patrizia Rossi: -------------- The results of the gain monitoring system of the Large Angle Electromagnetic Calorimeter (LAC) are presented. The criteria we had in mind when designing the monitoring system were: 1) The system should monitor the PMT gains as a function of time with an accuracy of 2-3%. 2) The light delivered to the PMT by the system should be in the range of that delivered during physics runs. 3) The system should be easy to use and fast. So to control the stability of the 512 PMT of the LAC we choose radiactive light pulsers, YAP:Ce-241 Am of 3.0 mm x 0.15 mm (outer diameter 4.0 mm x 1.0 mm) having a count rate of ~20 c/s. The light pulsers are in contact with the PMT's photo-cathode surface and using a mechanical system we can move them back and forth in order to increase or decrease the amount of light on the photo-cathode. Using the same electronic chain used during physics runs, we measured the gain stability of the 512 PMTs of the LAC after 1 days, 16 days and 27 days. The results obtained are summarized in the table below: (mean value and sigma of the peak variation distribution) SECTOR 1 SECTOR 2 days mean(%) sigma(%) mean(%) sigma(%) ---- ------------------- ------------------- 1 -0.36 0.53 -0.38 0.62 16 -0.39 1.06 -0.60 1.09 27 -1.25 1.33 -1.08 1.32 A gain monitoring system run takes ~ 5 min of time acquisition ( to collect ~ 200,000 events for both sectors) and a ~ 2 hours to analyze the output file. Cole Smith Elastic ep scattering analysis ---------- Inclusive and exclusive elastic cross sections were presented from analysis of 26M triggers from the Feb 98 e1 run. The integrated luminosity was 43 pb -1. Beam energy was 1.645 GeV with 60% torus. A simple fiducial cut by Volker Burkert designed to avoid the irregular edges of the Cerenkov acceptance was used to define the solid angle. Within this cut elastic electrons make up roughly 6% of the total triggers. A target empty run indicated that non-LH2 backgrounds are about 25% of the total triggers and 17% of the exclusive e-p elastic yield. After target empty subtractions, the ep/e elastic ratio is around 95%. Overall agreement of the inclusive angular distribution with the radiated theoretical cross section is excellent. The Q**2 dependence shows a departure from the dipole shape ranging from -10% at Q**2=0.35 to +10% at Q**2=1.0, with a crossover occuring at Q**2=0.77 GeV**2. This is roughly in agreement with previous experiments. Stepan Stepanyan: ----------------- Tracking efficiency from reactions (e,e'), (e,e'p) and (e,e'ppi+pi-). Run #8099 (December, 2.445 GeV, 2250A) was analyzed with two versions of tracking code (I named them RECSIS and PROD-1-3). Both versions have pointer bug fixed, in addition PROD-1-3 has also new scheme for LR ambiguity determination. The ratio of (e,e'p)/(e,e') in the elastic peak region (where acceptance for proton is not an issue) showed that both versions have more then 97% efficiency of reconstruction of protons. Although the comparison of two analyses were not done in the correct manner, the ratios of (e') and (e'p) distributions for both cases showed in average the same efficiency for event reconstruction. In case of PROD-1-3 one can see also slightly better resolution. Reaction ep=>eppi+pi- were studied at 4.045 GeV in order to extract efficiency of reconstruction of protons, pi+ and pi-. About 20M triggers were analyzed. Final state was selected using the missing mass and missing momentum cuts. As an efficiency of reconstruction the angular and momentum dependencies of the ratio (eppi+pi-)/(eh1h2) were studied in a narrow (+-10degree) region of phi angle. Here (h1h2) either (ppi+), (ppi-) or (pi+pi-). For protons that have relatively high momentum average efficiency in the "good acceptance" region was above 95%, while for pions, that have mainly low momentum average efficiency was about 90%-95% (that shows complicate dependence of acceptance from momentum and angle). Final conclusion of efficiency studies is that average efficiency for track reconstruction is more then 95% and it stays the same for different reactions. Using the same 4 GeV data at elastic scattering region studies of reconstruction of vertex position were done for protons and electrons. Z-distribution (along the beam) for both cases showed systematic dependence from sector to sector, that could be an effect of beam position. These needs more studies with different beam positions. ============================================================== Kim Egiyan: The First Look of Empty Target Data ----------- The "Empty Target" (ET) data from the Run 8938 ware analyzed. Data from the 8937 "Full Target" (FT) Run were used also. 1. In (e,e') scattering at Q**2 = 0.4-1.0 (GeV/c)**2 range the cross section ratio ET/FT = 20-22% was obtained. 2. In (e,e'p) scattering at the same Q**2 range the cross section ratio ET/FT = 16-18% was obtained. 3. Expected value of ET/FT ratio is close to the 3% (for 2x15 micrM Al wall tickness). 4. Thus, there is large unexpected yield in ET data. 5. To find out origin of these additional data, the (e,e'p) data were analyzed. 5.1 It was obtained that for ET the vertex coordinate along the beam direction-Vz- is very different from expected one. Instead of two small peaks at two target cell-wall positions, the 2 very intensive peaks was obtained in downstream wall position. This means that near this wall there is "something" which is a source of additional events. 5.2 Some physics analyses was used to find out the nature of this source. It is known that for nuclear targets the (e,e'p)-event distribution on the proton angle with virtual photon direction is a broad peak-type with the max at the TETA_pq =0. For the H-target this distribution has a very sharp peak, again at the TETA_pq =0. For the 8938 ET Run superposition of above mentioned two type distribution was obtained which is direct indication that in ET data there are two components - from the H and from the nuclear sources. Detailed analysis of ET and FT data show that H-contribution in ET data is approximately 5% off FT-data (data were normalized to the beam intensity). According the LH target manual, in "Empty Target" regime 2% of gas H (to the LH) should be expected (this value experimentally can be checked). Therefore 3% H-events are coming from the additional source at the downstream cell-wall. The remaining (12-14)% events have a nuclear origin. From this (12-14)% about 3% events are produced in 30 micrM Al walls, therefore, the origin of remaining (9-11)% should be obtained. 5.3 It is obvious, that thies are nuclei in (H_i,X_j)_n molecules. These can be molecules of some plastic (H_i,C_j)_n and/or molecule of water in Ice phase. The ET data can be explained very well in the second (Ice)-version. At this version 90% of nuclear (e,e'p) events are connected to the O_16 nucleus and only 10% on Al_27. The Following Conclusions was made.: 1. In the 1.645 ET Run data shows the nuclear effects, which can be published; 2. These data must be analysed in very detail allso for e2 experiment preparation; 3. It was shown that 90% of ET nuclear events in Run 8938 are from the the O_16 (and/or C_12) nucleus and only 10%- from the Al. 4. Vertex analysis show that there are still alignment problems in CLAS.