Simulation of the CERN-SPS crystal extraction experiment

Simulation of the CERN crystal-extraction experiments at the SPS, including the real crystal geometry and SPS parameters, as well as multiple turns and passes, has been performed. Two crystals with different geometry have been tried. We argue that in both cases the extraction experiment suffered from the crystal edge imperfection. For the experimental conditions reproduced, the efficiency of the first (twisted) crystal was found to be 12-18% (peak) in the angular range of 140-260 microrad (FWHM), depending on the vertical beam position at the crystal location. For the second (U-shaped) crystal the peak efficiency was much the same, ~19%, while the angular scan FWHM reduced to 70 microrad. These results, as well as the beam profiles, are in agreement with measurements. Finally we suggest some ways to advance this experiment.