Final-state b-quark in the NF=5 at NLO

Hi,

Asking for
> p p > z b
is ill defined at NLO accuracy.
And can not be consistently defined at NLO accuracy (but if you have massive PDF)
Exactly for the reason that you explained.

> But we do not need to showering in our calculation.

Then you should run on fix-order NLO.

> Another possible solution is to mimic MCFM.

I do not know which kind of precision and to which extend what they do is valid.
You should ask them I guess.

Cheers,

Olivier

> On Dec 29, 2016, at 22:33, Keping Xie <email address hidden> wrote:
>
> New question #423181 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/423181
>
> Dear Authors,
>
> I am Keping Xie. Recently, we want to calculate W/Z boson associated with Heavy flavor production at NLO. For example, if we want to calculate "p p > z b" at NLO, the leading channel should be "g b > z b", where b-quark PDF requires mb=0, i.e. NF=5. When we go to NLO, the singularity from "g > b b~" will appear, which is regulated by b-quark mass (mb!=0) in the default of MG.
>
> So, conflict appears! How can we solved it in MG?
>
> One possible solution is to generate "p p > z j" at NLO, and select the "p p > z b". But, we MG remind us the events.lhe.gz is not physical, which can not be used for analysis. Meanwhile, we switch to pythia to do the showering and get the physical events. But we do not need to showering in our calculation.
>
> Another possible solution is to mimic MCFM. In the source code loop_diagram_generation.py, we can set user_filter. How can we invoke this to select the diagram we want and avoid the crashes due to the un-cancelled poles?
>
> Is there any suggestions for me? In the best case, someone can provide me an example of dealing with FS b-quark in NF=5 at NLO.
>
> Best,
> Keping
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

Thank you for you quick answer.

"p p > z b" is ill defined at NLO, but "p p > z j" is well defined at NLO.
I wonder whether and how I can use "p p > z j " at NLO and select out "p p > z b".
The Matrix Element and pole cancellation should be all include in "p p > z j".

MCFM do the calculation (with mQ=0 through out all the calculations)
gQ->ZQ (Q is heavy quark) at 1-loop
qQ->ZQq
gQ->ZQg
gg->ZQQbar
 qqbar->ZQQbar
And combine all the channels to get the NLO results.

I want to compare these two calculations.

Best,
Keping

Hi Olivier,

Thank you for you quick answer.

"p p > z b" is ill defined at NLO, but "p p > z j" is well defined at NLO.
I wonder whether and how I can use "p p > z j " at NLO and select out "p p > z b" at NLO.
The Matrix Element and pole cancellation should be all include in "p p > z j".

MCFM do the calculation (with mQ=0 through out all the calculations)
gQ->ZQ (Q is heavy quark) at 1-loop
qQ->ZQq
gQ->ZQg
gg->ZQQbar
 qqbar->ZQQbar
And combine all the channels to get the NLO results.

I want to compare these two calculations.

Best,
Keping

Dear Keping,

Due to a completely different structure of the code, it's impossible to do exactly the same as in MCFM. In fact, at fixed order, there is currently no solution to generate 'p p > z b' at NLO in the 5 flavour scheme. The reason is that in the 5 flavour scheme, the b quark is formally indistinguishable from any of the other quarks (all are massless), and therefore in MG5_aMC we treat all these quarks identically.

When matching to the shower, the shower will give the b quark a different mass than all the other quarks, and therefore it will be possible to distinguish them from the other quarks. But that's only after you have showered the events.

Best,
Rikkert