8.4 Physics Results and Model Validation
Historically, modeling and experiment with a poloidal divertor have gone in parallel
all the time, with modeling closely following the experiment and helping to interpret
the experimental data. A comparison of the modeling results with experimental data
serves as code validation and different groups pursue this activity permanently,
following the improvements in the experimental diagnostics and development of
the model. However, given the obvious lack of comprehensive physical description
of the edge plasma, see Chaps. 6 and 7, this comparison can be rather tricky. In this
chapter, we give examples of the application of modeling tools to several problems
of the plasma edge physics, aiming at the qualitative understanding of relative
importance of different processes involved and at confronting the experiments to
validate the models. This is not a comprehensive review and we apologize for having
not touched upon many other applications – in particular, the kinetics or turbulence
codes.
8.4.1 2D Transport Modeling
Looking broadly, one can see a good qualitative agreement between the general
trends in the edge plasma in experiment and modeling, even without fine-tuning of
the models. Such features are, for example, evolution of the power loading and
particle flux onto the divertor target along with an increase of the edge density
(fueling) or impurity level (seeding) [65], appearance of plasma detachment at
sufficient density or radiation level [66–68], or reduction of the upstream plasma
density at the separatrix by impurity seeding [19, 69]. In particular, the effect of the
E
! Â B
!
and ∇B drifts on the detachment asymmetry in H-mode was confirmed by
comparing the UEDGE code results with DIII-D data from Thomson scattering
diagnostics by different directions of the toroidal magnetic field [62]. However, a
successful quantitative comparison of a modeling run with a single experimental
shot, which is usually termed “model validation”, is not easy. As was noted in [70],
the codes can reproduce the experimental data satisfactorily for Ohmic and L-mode
discharges with no significant plasma detachment from the targets. Since then, a
considerable effort has been made to model experiments with H-mode plasma and
detachment (see e.g. [19, 21, 62, 71–73]).
The most severe challenges in modeling the experiment are the in-out asymmetry
of detachment and the appearance of the high-density region at the top of the inner
divertor target in the far SOL. An example of modeling-to-experiment comparison is
shown in Fig. 8.4, where the density and temperature profiles in two DIII-D H-mode
shots with different orientation of the toroidal magnetic field, obtained experimentally and modeled with UEDGE, are shown. All four cases feature the same electron
density at the separatrix in the outer mid-plane. For the experimental profiles, the
data from the Thomson scattering diagnostics covering well the divertor region were
8.4 Physics Results and Model Validation
215
Historically, modeling and experiment with a poloidal divertor have gone in parallel
all the time, with modeling closely following the experiment and helping to interpret
the experimental data. A comparison of the modeling results with experimental data
serves as code validation and different groups pursue this activity permanently,
following the improvements in the experimental diagnostics and development of
the model. However, given the obvious lack of comprehensive physical description
of the edge plasma, see Chaps. 6 and 7, this comparison can be rather tricky. In this
chapter, we give examples of the application of modeling tools to several problems
of the plasma edge physics, aiming at the qualitative understanding of relative
importance of different processes involved and at confronting the experiments to
validate the models. This is not a comprehensive review and we apologize for having
not touched upon many other applications – in particular, the kinetics or turbulence
codes.
8.4.1 2D Transport Modeling
Looking broadly, one can see a good qualitative agreement between the general
trends in the edge plasma in experiment and modeling, even without fine-tuning of
the models. Such features are, for example, evolution of the power loading and
particle flux onto the divertor target along with an increase of the edge density
(fueling) or impurity level (seeding) [65], appearance of plasma detachment at
sufficient density or radiation level [66–68], or reduction of the upstream plasma
density at the separatrix by impurity seeding [19, 69]. In particular, the effect of the
E
! Â B
!
and ∇B drifts on the detachment asymmetry in H-mode was confirmed by
comparing the UEDGE code results with DIII-D data from Thomson scattering
diagnostics by different directions of the toroidal magnetic field [62]. However, a
successful quantitative comparison of a modeling run with a single experimental
shot, which is usually termed “model validation”, is not easy. As was noted in [70],
the codes can reproduce the experimental data satisfactorily for Ohmic and L-mode
discharges with no significant plasma detachment from the targets. Since then, a
considerable effort has been made to model experiments with H-mode plasma and
detachment (see e.g. [19, 21, 62, 71–73]).
The most severe challenges in modeling the experiment are the in-out asymmetry
of detachment and the appearance of the high-density region at the top of the inner
divertor target in the far SOL. An example of modeling-to-experiment comparison is
shown in Fig. 8.4, where the density and temperature profiles in two DIII-D H-mode
shots with different orientation of the toroidal magnetic field, obtained experimentally and modeled with UEDGE, are shown. All four cases feature the same electron
density at the separatrix in the outer mid-plane. For the experimental profiles, the
data from the Thomson scattering diagnostics covering well the divertor region were
8.4 Physics Results and Model Validation
215