10MINOS Preliminary: 3.2
D.3 U and V
xis not a fundamental property of the geometry of the detector: it is naturally oriented in U and V. Again, it is simpler to look at flat edges, so a U cut was applied to the V plots and a V cut applied to the U plots so only the flat edges are selected. The analyzed regions are shown in insets in Figure D.5. This figure also shows that the effect can be isolated by looking only at track end positions along the length of the strips. Note, the plane only measures position based on which strip a hit occurs in and cannot measure position along a strip. This means the effect only occurs in the
Figure D.2: The types of planes in the Near Detector with their component modules.
Figure D.3: These plots show Figure D.1 broken up into full and partial U and V planes. As you can see, the discrepancy appears equally in Full U and Full V planes but not in either Partial U or Partial V planes.
Figure D.4: Here we are looking at V and U planes (from the colored segments of the insets, again looking at the flat edges) broken up into calorimeter and spectrometer. The discrepancy appears in bothz segments, so it is not an effect of the lack of alignment in the spectrometer.
154 Detector Edge Study direction not directly measured by the plane. In the dimension the plane actually measures, there is excellent agreement between data and simulation as alignment has been performed on the data in this direction.
There is a second effect apparent in Figure D.5 that strongly suggests a mechanism that might produce the difference in track end position observed. Not only do data hits extend further out in the direction parallel to the strips, there are also more data tracks ending on those planes. The effect is striking since the data and simulation follow each other while they rise, the Monte Carlo just hits a cut-off earlier. The combination of these effects suggests that the strips in the simulation do not extend as far out relative the perpendicular plane as they do in the data. Figure D.6 shows an illustration of what the effect might look like.
However, there are several underlying effects that could produce this result. For example, the strips could be too short in the Monte Carlo, or the, strips could be offset along their length, or the plane itself could be offset. Additionally, the extra length could be an effect of the reconstruction (reconstructing points outside the physical detector).
Figure D.5: Viewed in U and V directions (as opposed to the linear combination of the two) features of the discrepancy become apparent. When position is measured along a plane’s axis (perpendicular to the direction of the strips) the Data and Monte Carlo agree well on where each strip is located, and thus where the detector edge is. However, when position is measured perpendicular to a plane’s axis (along the length of the strip), the discrepancy from Figure D.1 is clear. Note also that there is more Monte Carlo in the top plots and more Data in the bottom plots. Not only is there more data in the bottom plots, but the data and Monte Carlo curves follow each other and the Monte Carlo just falls off sooner than the data.
Figure D.6: A proposed explanation of the effects seen in Figure D.5. If the Monte Carlo strips did not quite extend to the edge of plane but the real strips did, it would produce the effects seen. First, the data edge would be further out than the Monte Carlo edge. Also, there would be more data hitting perpendicular strips since there is more strip at that edge than in Monte Carlo. Essentially, the Data particle, after passing through a parallel strip would go on to hit one more perpendicular strip while the Monte Carlo would appear to just stop in the parallel strip since there is no perpendicular Monte Carlo strip to hit that far out.
Figure D.7: These are the top and bottom of the detector inywith a selection onxas shown in the inset to get the flat segments. Referring to Figure D.2, U Top is the reflective end of the Ju module, V Top is the readout end of Mv and Nv modules, U bottom is the readout end of the Mu and Nu modules, and V bottom is the reflective end of the Jv module. While there is no apparent discrepancy in the yendpoint, note that there appears to be a small excess of data in the readout ends (V Top and U Bottom) and a small deficit of data in the reflective ends (U Top and V Bottom).
156 Detector Edge Study Coordinate Plane Modules
U Full U Ju
V Full V Jv
U Full V Kv and Lv
V Full U Ku and Lu
Table D.1: The correspondence between plots in Figure D.5 and the modules in Figure D.2.
One way to disambiguate an offset in strip position and an overall change in strip length is by looking at the other end of the strips. Unfortunately there is no data for the far end of the K and L modules, but there is data at the top and bottom of the detector which can tell us about the J modules. The plots can be seen in Figure D.7. There is no clear offset in the top edge of the detector.