Research Needs for Magnetic Fusion Energy Sciences - US Burning ...

to carry out the validation program, we will enlist a hierarchy of new and existing experiments,
including a spectrum of laboratory-scale devices along with the major confinement experiments
with their complement of sophisticated diagnostics. logically, the comparisons should proceed
in phases, proceeding from the simplest systems to the most complex (see Figure 2). each level
represents a different degree of physics coupling and geometric complexity. The most basic tests
attempt to isolate particular physical phenomena in the simplest geometry (a “Unit Problem”).
dedicated experiments, with specialized diagnostics, will be required at this stage and some comparisons
with analytic solutions may be possible. as experience is gained, researchers can be more
confident when modeling more “realistic” cases (“complete systems”). note that as the hierarchy
is traversed, the number of code runs and experiments tends to decrease. The quality and quantity
(especially spatial coverage) tend to decrease as well, while experimental errors and uncertainties
increase. information on boundary and initial conditions decreases as well. These trends
suggest that at least as much attention should be paid to the lower levels of the hierarchy as to the
top. These arguments may lead to requirements for new small-scale devices and/or substantially
improved diagnostics on existing ones. an effort to identify opportunities at this scale should
be undertaken to identify focused physics issues that could be addressed, stressing the “universality”
of the issues and their relevance in fusion plasmas. a further challenge will be to provide
codes that have appropriate geometry and work in appropriate regimes for each class of problems,
a considerable task. on the larger experiments, which probe regimes of direct relevance to fusion
energy, adequate run time and support must be made available.
The actual process of validation will be a collaboration among theorists, computationalists, and
experimentalists led by dedicated analysts who are not tied to particular code development or experimental
teams. These bridge the gap between specialized groups and are well placed to provide
unbiased, dispassionate assessments. This role is well recognized in fields where codes of high
consequence are employed. The analysts would have the primary role in defining validation tests
and diagnostic needs, in coordination with modelers and the experimental teams, and would carry
out much of the analysis and documentation of those tests. They would help marshal the substantial
computer time and experimental run time required by the validation program. a particularly
important activity would be the development of visualization tools, post processors and
synthetic diagnostics to make the comparison between codes and experiments more direct and
quantitative. The results of the validation experiments would feed back into the model development
activities, guiding efforts to explain and resolve the discrepancies that are uncovered.
Scale of Effort and Readiness for Thrust
significant resources would be required to pursue this Thrust with real vigor. each element described
would benefit from increased attention, better coordination and more funding. Theory is
at the foundation, providing the conceptual models and mathematical formulations, interpreting
the results of computations by identifying important physical processes. There will be a need for
both small and large code teams, each working on problems appropriate for their level of organization.
small teams are essential for exploratory work where flexibility and agility are critical
for success and where parallel efforts must be maintained. access to specialists in applied math
is beneficial to groups of all sizes, providing support for numerical methods and algorithms that
can qualitatively improve computational speed. as models mature and prove their utility, greater
effort is justified and a greater range of skills is required to make codes more robust and suitable
282