WHIZARD fails to find phase space channels (!!!) SEVERE

[Juergen Reuter]

For the process e1, E1 => A, A, A, A (four photons at LEP) WHIZARD fails to find valid phase space channels even when the number of off-shell lines is set to an arbitrarily high number! HIGHEST PRIORITY, PLZ!!!

[Juergen Reuter]

The corresponding qq -> 4g phase space is generated. So it looks like a problem with multi-peripheral t-channel diagrams that causes the failure.

[sschmidt]

1.) Might be connected to #164.
2.) I checked some similar processes and these didn't work either:

process eeA4 = e1, E1 => A, A, A, A             ! doesn't work
process eeA5 = e1, E1 => A, A, A, A, A          ! doesn't work
process eeA3Z = e1, E1 => A, A, A, Z            ! doesn't work
process eeA4Z = e1, E1 => A, A, A, A, Z         ! doesn't work
process eeA2Z2 = e1, E1 => A, A, Z, Z           ! doesn't work
process eeA3Z2 = e1, E1 => A, A, A, Z, Z        ! doesn't work

see the attached file for an example sindarin file and for a list of processes that work. It looks as processes don't work if there are at least four final state particles of which at least two are photons.

[sschmidt]

test sin file

[sschmidt]

I did some tests with older revisions of W2 and it turns out that the error appears when switching from r1619 to r1620.

[sschmidt]

As JR just explained to me this is not the solution, it's just that from r1620 processes of the type e+ e- -> Higgs(es) -> photons are excluded. The single t-channel contributions are still missing (and probably were always missing.)

[Juergen Reuter]

In the meanwhile I traced the problem down to the fact that for ee -> nA (n>4) the matching condition cascade1%depth + cascade2%depth + cascade3%depth = cascade4%depth in cascade_match_triplet is never fulfilled for processes with only multiperipheral t-channel topologies. Hence, cascade_combine_keystone is never called, and cascade_set%final_k is hence never associated, which triggers failure in cascade_is_valid.

[kilian]

Sorry guys, but this is another lack-of-documentation issue. There are two phase-space parameters: phs_off_shell and phs_t_channel. Here you have to set manually

phs_t_channel = 3

and it works:

| Integrating process 'a4'
| Phase space file 'a4.phs' not found.
| Generating phase space configuration ...
| ... found 48 phase space channels, collected in 1 groves.
| Phase space: found 1152 equivalences between channels.
| Wrote phase-space configuration file 'a4.phs'.
| Process 'a4': phase space setup complete.

[Juergen Reuter]

Milestone v2.0.0final deleted