Complex Mass Scheme in WHIZARD

[kilian]

This is forked off from #420.

1. Whizard model files should support a syntax that distinguishes between different schemes for a single model. This is already implemented, partly. I'll have to check for any missing pieces.
2. The import_from_whizard routine has to carry an (optional?) argument for the scheme, so the parameters.XXX.f90 code can make use of this. This is not yet implemented.
3. The O'Mega targets.ml source code has to know this; this is a trivial change.
4. O'Mega has to support new propagators for the fermion and vector cases where also the numerator / the gauge terms take the imaginary part (the width) into account; this remains the subject of #420.

[kilian]

r7505 contains the necessary ingredients.

1. The syntax in model files is borrowed from Fortran: select scheme - scheme "foo" - scheme other - end select. (We still can modify this.) It can be used only within the parameter/derived section of the model file. There is a new parameter category 'unused' which may lock variable names that exist in one scheme but not in another one, to avoid confusion. The available schemes have to be listed at the beginning of the model file. See QCD.mdl for an example, where the default scheme works unchanged, while the new 'massive' scheme includes masses for u and d quark (default value zero).
2. import_from_whizard carries a second mandatory argument scheme. This is an integer. If the model does not support schemes, the value will be zero and can be ignored. Otherwise, the value will be the number of the chosen scheme, in the order as they were introduced in the model file. A simple use case can be found in parameters.QCD.f90. Note that the default scheme for a model which does support schemes, is scheme number one.
3. targets.ml has been modified accordingly.

For implementing the actual complex mass scheme, the propagator as generated by O'Mega must be changed, cf. #420. Once this is possible, model file support can be added to the Whizard model library, either in the default SM or as a new model.

Note that the new facility allows for implementing an alternative electroweak (tree-level) scheme without touching O'Mega. This has not (yet?) been implemented in a model.

[kilian]

I should add that the change of the interface of import_from_whizard might affect the FeynRules?/SARAH interface(s). If this is the case, we have to take care with the next release.

[Juergen Reuter]

JRR has notified the authors of SARAH and FeynRules? concerning the changes in the Interface. Florian Staub already promised that this should not be a big deal to have a change depending on the version (<=2.2.8 old interface, >= 2.3.0 new interface) in SARAH.

[Juergen Reuter]

This should be one of our next big goals.

[Juergen Reuter]

As of r7671 the Complex Mass Scheme for the model SM is there and working. Pending are the validation, a possible option to set the electroweak renormalization scheme (definition of alpha as alpha0, alpha_MZ or from GF, MW, MZ), and the extension to other models. Shall we close this ticket or wait for the validation?

[Juergen Reuter]

The implementation has been finished vor v2.3.1. Full validation is still pending, but this will be a new task, and if problems there will be showing up, we should create a new ticket.