VSAT

My VSAT project was to evaluate the relative luminosity for the 1995 scan. This implied the following tasks: recovery of events with mismatched minibunch numbers, energy calibration and measurement of the beam parameter variation.


The events

The Very Small Angle Tagger is an electromagnetic sampling calorimeter consisting of four rectangular modules called F1, F2 in the forward region and B1, B2 in the backward region. The distance of the modules from the DELPHI origin is approximately 7.7 m. The luminometers are placed symmetrically around a short elliptical section of the beam pipe:




The process seen by the detector is Bhabha scattering at small angles (5-7 mrad), i.e. electrons and positrons emitted back-to-back carrying approximately the beam energy. An event is therefore a coincidence of signals between a module in the forward region and a module in the backward region. Two diagonals are defined for the trigger: diagonal 1 (modules F1-B2) and diagonal 2 (modules F2-B1).



Minibunch mismatches

In 1995 period 2 an energy scan of the Z0 peak took place. For this run LEP had modified its previous mode of single bunch operation to a scheme of tightly spaced bunch-trains, i.e. instead of having one bunch crossing every 25 microseconds there were 2 to 4 bunch crossings within a fraction of a microsecond every 25 microseconds. Most of the scan data taking was done with 3 minibunches separated by about 250 ns.

During the data-taking, the VSAT detector assigned wrong minibunch numbers to a number of cases. The following figure shows their relative importance by comparing their populations to these of the so-called good assignements, i.e. events where both diagonal modules gave the same minibunch number. Plot (a) shows the number of events measured by module F1 as a function of the minibunch number in B2 and the minibunch number in F1. It includes both good assignments and all mismatch cases. The only mismatch which is visible in this plot is the case where high energy minibunch 3 data were flagged as minibunch 2 events in B2. The energy calibration for these events is described in this paper . The rest of the mismatches are shown in plot (b). Their treatment is described in this paper . Plots (c) and (d) show the corresponding distributions in diagonal 2. The analysis was performed over six data sets, one for each beam energy and each corresponding to two periods: before and after the radiation accident of September 15, 1995.

Before this analysis, the mismatched events represented 10 percent of the total sample. Their recovery was rather successful, leaving less than 0.1 permille of unretrieved data.









Beam Parameters

The VSAT monitoring of the beam parameters is used as an on-line cross check of similar measurements done by the general monitoring system of LEP at the DELPHI interaction point. These measurements are used in order to optimize the luminosity, which depends strongly on the beam parameters. Among the DELPHI detectors, the VSAT has the highest accepted cross section and can thus collect sufficient statistics in a time interval relevant for on-line control.

The off-line estimation of the variations of the beamspot position as well as those of the directions of the incident particles are used to optimize the alignment of the different DELPHI detectors. The beam parameter monitoring also provides an important cross check when it comes to the VSAT luminosity measurement: the Bhabha cross section in the detector's acceptance depends on the beam parameters, so it has to be corrected for their variations. Since the correction function is determined by the detector simulation (FASTSIM), it is important to cross check the validity of the model used in FASTSIM by comparing the description of the variations of the beam parameters that we obtain from VSAT data with data from LEP or other DELPHI detectors.

Beam parameter plots for 1993 and 1994 VSAT, VD and TPC data (as well as a detailed derivation of formulae) can be found in this report . A comparison of VSAT results with the LEP 1993-1994 acollinearities has been reported in this DELPHI note . The estimation of the beam parameter variations requires a number of steps, which are described in this paper. The following plots show beam parameter VSAT measurements which are relevant to the 1995 luminosity calculation. The x-beamspot estimation and its variation with fill number is shown here:




The x-beamspot measurement has been corrected for variations of the beam width, divergence and acollinearity in the (y,z) plane. It does not include corrections for the acollinearity in the (x,z) plane, however, which affects the normalized difference between the VSAT and the VD measurements (plot (b) below). Nevertheless, the linear relation between the two measurements is evident (plot (a)):




The following four plots give the corresponding distributions for the z-beamspot and its comparison to the TPC measurement.






The variations of the x- and y-acollinearity, the x- and y-tilt and divergence have been reported in this DELPHI note. As 1995 was the first year LEP ran in the minibunch scheme, it was of interest to examine whether the three minibunches exhibited differences in their beam parameter characteristics. Since the VD had not supplied beamspot information per minibunch, the calculation of the acollinearities has not been performed. However, the distributions of the x- and z-beamspot, as well as the mean tilts in the two planes, have been extracted from VSAT data alone for the three minibunches. The results are shown in the following figure, which shows no dependence on minibunch number (solid histogram = minibunch one, dashed histogram = minibunch two, dotted histogram = minibunch three).








Luminosity

The luminosity measurement in DELPHI is performed by two luminometers: the STIC detector, which calculates the absolute luminosity and the VSAT detector, whose task is to obtain a relative measurement with lower statistical error. For the 1995 scan, the main use of the VSAT measurement was to cross check the results of STIC. The agreement between the two luminosity measurements has been confirmed by the normalized differences (pulls) at the three energy points and for the three minibunches separately:




The two measurements have clearly followed each other, as can also be seen in the following correlation plots:




The results of the luminosity analysis have been reported in this DELPHI note and were used by the DELPHI collaboration in the lineshape analysis, which was published in this paper.

My work on the VSAT resulted in a filosofie licentiat, which was presented on October 2nd, 1998.


[My homepage]