Calibrating the Monte Carlo model

 The probability per straw to observe a high-threshold hit on the reconstructed track as a function of the threshold in keV is shown in fig. 5.16 for pions and electrons in a 1.56 T magnetic field. The testbeam measurements are represented as horizontal bands with a width corresponding to the measurement errors. The discriminator threshold was at the time the data was taken calibrated to a 5% accuracy using data taken with a 5.9 keV ⁵⁵Fe X-ray source.

  

Figure 5.16: Probability to observe a high-threshold hit in a straw crossed by a pion or electron track as a function of the value chosen for the high threshold. The results are shown as horizontal bands for the data, which were taken at a fixed threshold of approximately 6 keV, and as curves for the Monte Carlo simulations. For the simulation of electrons, the results are shown for relative radiator performances of 80% and 90% (see text).

  

Figure 5.17: Pion efficiency at a fixed electron efficiency of 90% as a function of the high threshold. Results from measurements in a 1.56 T magnetic field are compared with Monte Carlo predictions for relative radiator performances of 80% and 90%.

For pions, the high-threshold hit probability is measured to be 4.93 $\pm$ 0.07% , which corresponds to a value of 6.00 $\pm$ 0.06 keV for the high threshold, in excellent agreement with the ⁵⁵Fe measurements. Once this scale is fixed, the simulation can be compared to the data from electrons which produce transition radiation. The Monte Carlo model used for the production of transition radiation is valid for an ideal radiator with perfectly regular foil spacings and foil thicknesses. In reality, the prototype radiator geometry is not perfectly regular, and a tunable overall reduction factor was applied to the amount of transition radiation produced. This factor reflects the relative radiator performance with respect to an ideal radiator, and was found to be 88.4 $\pm$ 2.2 $\pm$ 2.0% for a 6 keV threshold, as shown in fig. 5.16. The first quoted error is statistical and the second one is an estimate of the systematic error, based on the analysis of data taken at different periods during the testbeam run in 1996. This measured relative radiator performance was found to be stable as a function of the choice of GEANT cutoffs used in the Monte Carlo model.

The measured and predicted pion efficiencies at a fixed electron efficiency of 90% are shown in fig. 5.17 as a function of the high threshold. Data taken in a 1.56 T magnetic field are compared to simulations with relative radiator performances of 80% and 90%. This figure shows that the relative radiator performance varies from $\sim$ 80% for a threshold around 4.5 keV to $\sim$ 90% for a threshold above 6 keV. The shape of the transition radiation spectrum is therefore not reproduced accurately enough by the simulation. More detailed comparisons between measured and predicted transition radiation spectra and spectra can be found in [57].