Speed up simulation configurations by stopping particles being tracked well beyond the detector.

[simonge]

Is your feature request related to a problem? Please describe.
The simulation runs very slowly for simple configurations without calorimeters or complex materials which should.

This came up when tracking single beam electrons from the IP through the vacuum pipe which was running at about 3Hz due to world volume extending beyond the vacuum, magnets and pipe implemented in the geometry. Adding a volume which just absorbs the tracks at the end of the detector geometry sped the simulation up to 1000Hz.

This is unlikely to have a large effect on the full detector simulation of events in the central detector, however if only tracking layers are being simulated the effect could be significant, or for Low-Q2/low Pt events where the electron and proton will be tracked out the end of the beampipe.

Describe the solution you'd like
There are a few solutions to this problem.
Reduce the world volume, and/or create a more complex world volume shape.
Shape the world volume to match each configuration.
Add track killing volumes after the last layer of detectors in any given configuration.