Higgs mass in the MSSM
-
jyotiranjan
- Posts: 24
- Joined: 24. Aug 2016, 11:09
Higgs mass in the MSSM
Dear Florian,
I am trying to generate a spectrum in the MSSM with SPheno-3.3.8.
I observe that SPheno gives tachyonic Higgs masses and I checked
the same in FeynHiggs and I do not see any error. The parameter
point is attached. Could you suggest if I am missing some
important settings in the input file.
Best regards,
Jyotiranjan
I am trying to generate a spectrum in the MSSM with SPheno-3.3.8.
I observe that SPheno gives tachyonic Higgs masses and I checked
the same in FeynHiggs and I do not see any error. The parameter
point is attached. Could you suggest if I am missing some
important settings in the input file.
Best regards,
Jyotiranjan
- Attachments
-
- inpMSSM-SPheno-3.3.8.txt
- Input file for SPheno
- (3.63 KiB) Downloaded 235 times
-
- spc-FeynHiggs-2.13.0.txt
- Output SLHA file from FeynHiggs
- (15.16 KiB) Downloaded 246 times
Re: Higgs mass in the MSSM
Hi,
after taking a look at your point, there are at least three important remarks:
1) You should turn on
which decouple SUSY states at M_SUSY and perform a Higgs mass calculation within an effective SM, see also 1703.03267
This solve the tachyonic problem at tree-level because SUSY states are no longer evolved down to M_Z to get the mass spectrum.
2) At the loop level, there are still problems with tachyons. The difference to FeynHiggs is that M_A is not on-shell renormlised. Thus, to get the same pole masses for H_2, A^0 and H^+ you need to use higher input values for MA2. This gives then a well defined spectrum
3) There are still tachyons in the calculation of the flavour observables. The reason is that those are performed at Q=91 and 160 GeV. These are artefacts because in these calculations the SUSY states are still taken into account as degrees of freedom and not integrated. Unfortunately, there is no way to circumvent this at the moment but turning off the flavor observables.
Best,
Florian
after taking a look at your point, there are at least three important remarks:
1) You should turn on
Code: Select all
66 1 # Two-scale matching
67 1 # EFT Higgs mass
which decouple SUSY states at M_SUSY and perform a Higgs mass calculation within an effective SM, see also 1703.03267
This solve the tachyonic problem at tree-level because SUSY states are no longer evolved down to M_Z to get the mass spectrum.
2) At the loop level, there are still problems with tachyons. The difference to FeynHiggs is that M_A is not on-shell renormlised. Thus, to get the same pole masses for H_2, A^0 and H^+ you need to use higher input values for MA2. This gives then a well defined spectrum
3) There are still tachyons in the calculation of the flavour observables. The reason is that those are performed at Q=91 and 160 GeV. These are artefacts because in these calculations the SUSY states are still taken into account as degrees of freedom and not integrated. Unfortunately, there is no way to circumvent this at the moment but turning off the flavor observables.
Best,
Florian
-
jyotiranjan
- Posts: 24
- Joined: 24. Aug 2016, 11:09
Re: Higgs mass in the MSSM
Thanks, Florian. It solves the issue.
Best regards,
Jyotiranjan
Best regards,
Jyotiranjan
-
jyotiranjan
- Posts: 24
- Joined: 24. Aug 2016, 11:09
Re: Higgs mass in the MSSM
Dear Florian,
I have a few more naive questions. I could generate the spectrum in
SPheno-3.3.8 with around 6TeV of M_A. However, It seems that
I would require even larger M_A if I raise mu and TU(3,3) to a bit
larger value. Is this an acceptable/required feature?
Secondly, the flags you mentioned in your previous mail, i.e.,
does not seems to have any impact in the version I am using
(SPheno-3.3.8 with SARAH generated MSSM model). Do I have to
use SPheno-4 for that?
While building SARAH-4.12.1 generated MSSM in SPheno-4.0.3
using gfortran compiler, I see following error message
Could you shed some light on these issues and their possible
solutions.
Best regards,
Jyotiranjan
I have a few more naive questions. I could generate the spectrum in
SPheno-3.3.8 with around 6TeV of M_A. However, It seems that
I would require even larger M_A if I raise mu and TU(3,3) to a bit
larger value. Is this an acceptable/required feature?
Secondly, the flags you mentioned in your previous mail, i.e.,
Code: Select all
66 1 # Two-scale matching
67 1 # EFT Higgs mass
(SPheno-3.3.8 with SARAH generated MSSM model). Do I have to
use SPheno-4 for that?
While building SARAH-4.12.1 generated MSSM in SPheno-4.0.3
using gfortran compiler, I see following error message
Code: Select all
LoopDecays/LoopDecayGlu_MSSM.f90:86.49:
Call hardradiationFFS(Mex1,Mex2,Mex3,MLambda,em,<>StringRiffle[{0._dp, 0._dp,
1
Error: Syntax error in argument list at (1)
LoopDecays/LoopDecayGlu_MSSM.f90:659.49:
Call hardradiationFFS(Mex1,Mex2,Mex3,MLambda,em,<>StringRiffle[{0._dp, 0._dp,
1
Error: Syntax error in argument list at (1)
solutions.
Best regards,
Jyotiranjan
Re: Higgs mass in the MSSM
Hi,
These Flags are available since SARAH 4.11 (independent of the SPheno version). The Riffle problem should only show up with older Mathematica versions. Which one do you use?
Cheers
Florian
These Flags are available since SARAH 4.11 (independent of the SPheno version). The Riffle problem should only show up with older Mathematica versions. Which one do you use?
Cheers
Florian
-
jyotiranjan
- Posts: 24
- Joined: 24. Aug 2016, 11:09
Re: Higgs mass in the MSSM
Thanks a lot, Florian.
I used Mathematica-10.0 and the resulting SPheno files for the MSSM
in SARAH-4.12.1 had problem. However, Mathematica-11.0 seems to
work perfectly.
Yes, the flags (Two Scale matching and EFT Higgs mass) are working
with SPheno routines generated via latest SARAH versions. I am able
to generate spectrum with slightly larger values of DR-bar input for
MA2 ( 3.0E+06) which gives pole mass around 1 TeV that I used
for FeynHiggs input file also.
However, I observe that the resulting Higgs mass is a bit smaller
(115 GeV) compared to FeynHiggs-2.13 reported masses (124 GeV).
I also notice that FeynHiggs gives somewhat larger yukawa couplings
especially bottom yukawa at the scale I am using for spectrum
calculation.
Best regards,
Jyotiranjan
I used Mathematica-10.0 and the resulting SPheno files for the MSSM
in SARAH-4.12.1 had problem. However, Mathematica-11.0 seems to
work perfectly.
Yes, the flags (Two Scale matching and EFT Higgs mass) are working
with SPheno routines generated via latest SARAH versions. I am able
to generate spectrum with slightly larger values of DR-bar input for
MA2 ( 3.0E+06) which gives pole mass around 1 TeV that I used
for FeynHiggs input file also.
However, I observe that the resulting Higgs mass is a bit smaller
(115 GeV) compared to FeynHiggs-2.13 reported masses (124 GeV).
I also notice that FeynHiggs gives somewhat larger yukawa couplings
especially bottom yukawa at the scale I am using for spectrum
calculation.
Best regards,
Jyotiranjan
Re: Higgs mass in the MSSM
Hi,
9 GeV difference are too much. There should be agreement within 3GeV. Did you maybe run MakeSPheno with ReadLists->True? I recognized yesterday that in this case the two loop corrections are missing. Otherwise, you can send me your input again and I'll check next week.
Cheers
Florian
9 GeV difference are too much. There should be agreement within 3GeV. Did you maybe run MakeSPheno with ReadLists->True? I recognized yesterday that in this case the two loop corrections are missing. Otherwise, you can send me your input again and I'll check next week.
Cheers
Florian
-
jyotiranjan
- Posts: 24
- Joined: 24. Aug 2016, 11:09
Re: Higgs mass in the MSSM
I am attaching herewith the input files for SPheno and FeynHiggs.
- Attachments
-
- inpMSSM-FH-2.13.txt
- (575 Bytes) Downloaded 224 times
-
- inpMSSM-Spheno.txt
- (3.7 KiB) Downloaded 296 times
-
jyotiranjan
- Posts: 24
- Joined: 24. Aug 2016, 11:09
Re: Higgs mass in the MSSM
I just used
Start["MSSM"]
MakeSPheno[]
Do I have to use "ReadLists->True" somwhere?
Best regards,
Jyotiranjan
Start["MSSM"]
MakeSPheno[]
Do I have to use "ReadLists->True" somwhere?
Best regards,
Jyotiranjan
Re: Higgs mass in the MSSM
No, no. There is no need for that. It was just a guess what might cause the big difference. Another guess is that it comes from the value of A_top. You use for Tu(3,3) and Atop the same values, but T corresponds to Y times A, i.e. Tu~ 0.9 x 3800 for a running too Yukawa of 0.9.
Cheers
Florian
Cheers
Florian