The fast ATLAS track simulation (FATRAS)

Various systematic physics and detector performance studies with the ATLAS detector require very large simulated event samples. Since the full detector simulation is a highly CPU time consuming operation, fast simulation techniques are widely used in such applications. Furthermore, the simulation of background events does, in general, not require the very detailed detector simulation and fast simulation techniques satisfy the needed accuracy. In ATLAS, the fast simulation program ATLFAST has been extensively used for such purposes. It is, however, based on the smearing of the initial particle properties and is not capable of producing hits along the track. Tracking relevant studies that include both hit information and pattern recognition effects can not be performed when using ATLFAST. An alternative simulation program, the new Fast ATLAS Track Simulation (FATRAS) has been recently deployed, capable of producing full track information, including hits on track. Initially developed as a validation tool for the ATLAS offline track reconstruction, it has become a powerful engine for various use cases. In general, the CPU time determining factor of the full simulation is the tracking of the particle through the very complex detector geometry, while the event reconstruction including pattern recognition and track fitting is relatively fast. In FATRAS, the simplified reconstruction geometry is used as a simulation geometry model, which leads to a significant speed up of the simulation process. FATRAS uses furthermore mainly common offline track reconstruction code and the reconstruction event data model. It is fully embedded in the ATLAS C++ based software framework ATHENA.