Progress on DART code optimization. Page: 2 of 4
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
DART (Dispersion Analysis Research Tool) is the unique mechanistic model for
the assessment of dispersion fuel behavior for oxide, silicide, and other new dispersant
phases (vhd alloys).
However, due to its evolution as an R&D tool, it was not developed as a user-
friendly code. Besides, each simulation done to study the effect of changing a particular
parameter or operating condition, demands the iteration of a process consisting in input
preparation, DART run, extraction of calculated quantities from program output file and
plotting. Each step of the calculation process has an extension of a couple of minutes to
many hours, depending on complexity of the problem. For an analysis covering a
multitude of parameters and/or operating conditions, this is a very long and tedious
The conversion of DART into a parallel architecture version was an ideal
situation to improve calculation power, due to the increasing tendency in IT technology
to appeal to multithreading techniques and parallel processing.
As a part of SISTERLAB agreement, ANL and CNEA have proposed several
topics for mutual collaboration. One of them is related with modeling. It consists of a
full revision of DART models and version codes and the inclusion of new models in the
framework of the development of a unique parallel architecture version for DART code.
Several aims are pursued, namely to enhance DART input/output structure in order
to increase its availability and usefulness in the international community and to afford
the opportunity to develop an interface whereby the user can monitor the evolution of
various calculated quantities "in situ."
In addition, it will provide the possibility for the user to change values of various
parameters and/or operating conditions during the course of a run. The user/code dialog
will become highly optimized and the analysis procedure will be more efficient. It is
intended to parallelize a variety of calculations performed as a function of operating
conditions and fuel morphology, like
* Evolution of the fission-gas bubble size distribution and meat thermal
* Fuel-meat matrix interaction
* Evolution of fuel microstructure
* Stress/strain analysis,
and other issues. These processes, if parallelized, will provide a much more efficient
calculation. It was thought that a project like this also would allow the merging of all
different versions of DART into a single code. The project include the aim of
developing new models such as superplasticity, elastoplastic feedback, improved
models for the calculation of fuel deformation and fuel microstructure evolution. As a
whole this project implies a rigorous inspection and overhaul of DART bringing the
user and developer of the international community a very valuable benchmark and also
it will form the basis of a code for the analysis of dispersion fuel during transient
(and/or accident) conditions.
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Taboada, H.; Rest, J. & Solis, D. Progress on DART code optimization., article, October 2, 2001; Illinois. (digital.library.unt.edu/ark:/67531/metadc724843/m1/2/: accessed December 11, 2018), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT Libraries Government Documents Department.