Update the way charge is spread on sTGC strips
While fixing the issue described in ATLASSIM-5456, another bug was found: charge assigned to a strip may become negative for large incident angle. A workaround was added to drop the strips with negative charge. The proper fix was discussed in a Muon Software meeting back in July [1]. Since the changes violate the frozen tier0 policy, it was decided to wait for data, which was becoming available, in order to tune the resolution parameters. However, due to the lack of proper calibration in data, the spatial resolutions are not accurate. Instead, the \sigma_0 and \sigma_{\theta} parameters are taken from analyses of cosmic data.
Best fit of the spatial resolution from cosmic data:
The plot below compares the spatial resolution from simulation including this fix (shown in red) to the best-fitted curve from cosmic data (blue curve):
A good agreement is expected, since the same parameters are used.
Edited: change the target to master (from 22.0)