Studies for Online Selection of Beam-Gas Events with the LHCb Vertex Locator

The start of the Large Hadron Collider (LHC) is scheduled for the Summer 2008. The accelerator is going to provide unprecedented amount of proton-proton colli- sions with a record center-of-mass energy. The total number of collisions produced in an interaction point is directly connected to a collider characteristic called `absolute luminosity'. The luminosity depends on a number of quantities like the number of particles in a bunch, the bunch size and the number of bunches in a beam. For precise measurements of Standard Model parameters and for the search of New Physics the LHC experiments count on precise knowledge on its luminosity. The absolute luminosity of LHC is going to be measured using various meth- ods, including the recently proposed beam-gas luminosity method. This method counts on the reconstruction of beam-gas vertices for measuring the beam shapes and overlap integral. The beam-gas luminosity method is going to be first tried in the LHCb experiment, making use of its excellent vertex resolution and accep- tance for beam-gas events from both beams. At LHC start-up, residual gas in the beam-pipe of LHCb will produce beam-gas collisions. These will be used for initial studies of this novel method and for monitoring the position and the angles of the beams constantly during the run. In order to collect sufficient beam-gas events, a method must be developed to accept in the High Level Trigger (HLT) events that contai n a beam-gas collision. This must be achieved without notably affecting the core LHCb physics data acquisition. Therefore, a beam-gas trig-gering algorithm needs to be developed which e±ciently selects beam-gas events while rejecting at least 99.9999% of events not containing the desired beam-gas interaction. This trigger algorithm will have to consume a negligible amount of CPU power in the HLT filter farm. The purpose of the studies presented in this thesis is to investigate the beam-gas events and the possibilities of their triggering prior to the decision of injecting gas into the LHC and asking for a dedicated run. Our first algorithm used directly the measurements (clusters) of the vertex de- tector. Its results were not so promising but not pushed very far either. Next, a study with 3D tracks showed that it is possible to achieve reasonable efficiency with 1 ppm retention, but expected excessive CPU power consumption. Therefore we investigated an algorithm based on 2D (RZ) tracks which gave promising results. The actual C++ implementation needs to be done, for timing studies. May have to combine the 2D tracks algorithm with the cluster-based approach(mixed solution).