The model was used to examine possible bounce trajectories. Traces were run for energies ranging from near 0 up to the beam energy on paths that make it through the exit slit. These traces are shown in blue. The incoming beam is shown in green. Possible single bounce paths exist where the green paths and the blue paths intersect at a conductor. With the analyzer set near the beam energy the only single bounce paths are at the input collimator.
On the low side (analyzer selects a lower energy than the source), the main beam hits the bottom plate past the exit slit. There are some single bounce trajectories that could originate at the input collimator. Electrons from the main beam would have to bounce backward to reach the exit slit and would have to bounce more than once to arrive at an accepted angle. Despite this, the change in slope seen at about .45 may be evidence that scattering off the bottom plate does contribute. At ratios of less than .45 the beam does not hit the bottom plate at all. It goes through a screen into the trap area.
On the high side, the main beam gets forced into the bottom plate. There are single bounce paths that can originate where the main beam hits. At ratios of 2 or less the overlap area is fairly wide. There are no single bounce paths from the collimator.
At ratios above 2.5 the area where the main beam hits no longer overlaps with single bounce paths accepted by the slit. In fact there are no single bounce paths for ratios above 2.5. Unfortunately the data doesn't support this model. There is no sharp decrease seen at ratios above 2.5. Double bounce paths can reach the slit for ratios up to 4.5