Conclusion

Thesis	The work presented in this thesis has two main parts: the development of the LHCb vertex locator (VeLo) and the study of rare semileptonic B decays.
L1 Electronics	The IPHE Lausanne is involved in the development of the VeLo Off-Detector readout Electronics (ODE). The most crucial part is the interface to the second level (L1) trigger farm. The L1 Pre-Processor Interface (L1PPI) has the task of providing zero-suppressed hit positions to the trigger in a reasonable time. It is of particular importance for the trigger performance to achieve almost perfect hit finding while keeping the overall data rate low, i.e. the purity high. This is a difficult task in an environment where correlated noise may become a major source of nuisance. Moreover the very small bunch spacing time (25 ns) may cause overspill between consecutive events. We have developed a linear common mode algorithm (LCMS) which suppresses correlated noise over 32 input channels. We have simulated the effect of the algorithm on the trigger efficiency using simulated hits and raw testbeam data as noise. We show that the trigger efficiency is stable for S/N>10 and for pulse overspill fractions below 30%. These limits are safe compared to the values measured on prototype detectors and readout chips. The LCMS algorithm has recently been adopted by the LHCb Inner Tracker group for their prototype tests. All these results have been obtained using the old TP VeLo geometry and the TP L1 trigger algorithm. The new design of the VeLo is not expected to cause dramatic changes in these results but the definitive trigger algorithm may be less sensitive to noise. Consequently the simulation should be updated for the upcoming trigger TDR.
Testbeam Simulation	At the 1999 VeLo testbeam, the LHCb VeLo group have tested prototype detectors in a 120 GeV muon beam from the SPS accelerator. They achieved the measurement of slope-dependent resolutions at the micron level. It appeared that no simulation was able to reproduce the results. We have developed a detailed simulation of particle interactions in silicon sensors. It involves the description of delta-rays, lateral diffusion and charge coupling between readout strips. This simulation allows the description of the observed resolutions with an accuracy below one micron. We present its predictions for the various detector geometries proposed in the VeLo Technical Design Report.
Rare b → lls,d Decays	We finally assess the physics performance of the LHCb experiment in the field of rare semileptonic decays b → lls,d. These flavour changing neutral current decays occur via loop or penguin diagrams, which makes them very sensitive to physics beyond the Standard Model. Moreover Delta F = 1 transitions are nowadays very weakly constrained experimentally and may thus reveal the limits of the CKM paradigm. Basing on a pre-existing simulation written by CDF for B → μμK and B → μμK*, we have developed an event generator package describing the B → μμXs decay inclusively. The hadronic Xs mass spectrum is described as a sum of strange hadronic resonances and fitted to theoretical predictions. This package is available as a QQ user decay model in the LHCb simulation framework. We have estimated the annual event yields using the GEANT-based full detector simulation. In spite of too low statistics for some background event samples, we have assessed the expected event yields and S/B ratios at a sufficient level of precision to estimate the physics yields. Assuming BR(B → μμXs)=5.7x10-6, we expect 24000 B → μμXs events per year at about S/B = 8. The B → μμXd channel yields 550 events per year with about S/B = 1 assuming a CKM matrix elements ratio of Vtd/Vts=1/30. Several sources of background have been studied. These results use the full potential of 2 inverse fb per year one can achieve using also multiple pp events. We have developed a multi-primary vertex reconstruction algorithm which allows the use of double and triple events. These yields allow the measurement of exclusive and inclusive branching ratios with a relative (statistical) precision of 1-2% for B → μμXs decays and 8-22% for B → μμXd. This is better than the theoretical errors on the Standard Model estimates and thus sufficient to test the SM versus Susy models predicting significant enhancements of these branching ratios. The comparison of the B → μμXs and B → μμXd decays allows to extract the Vtd/Vts CKM matrix elements ratio with a negligible theoretical error. This sets a constraint on Vtd and thus on the unitarity triangle by defining the length of the Rt side opposite to gamma. After one year of data taking we expect a statistical error of 11.5 (+2.8,-3.2) for (Vtd/Vts)2=1/30. Over the whole allowed range of this ratio, the limit of 5% statistical error can be reached in less than 10 years of data taking. This error of 5% is the present theoretical limit on the precision of this ratio using the ratio of Delta ms and Delta md. At the present theoretical knowledge, the method presented here provides thus the most precise extraction of Vtd at LHC. The precisions on CP asymmetries are far above the Standard Model expectations, but of the same order as predictions from supersymmetric theories. A non-zero CP asymmetry measurement would be a clear signal of physics beyond the Standard Model.

More details about the cited facilities and experiments can be found on the links page.