USBPO Mission Statement: Advance the scientific understanding of burning plasmas and ensure the greatest benefit from a burning plasma experiment by coordinating relevant U.S. fusion research with broad community participation.
Announcements Research Highlight O. Meneghini, S.P. Smith Schedule of Burning Plasma Events Contact and Contribution Information
New ITER job announcements
Several positions have been opened at ITER Headquarters in Saint Paul-lez-Durance, France, including two in Machine Operations that might be of interest to our readers. For more information, including instructions on how to apply and a complete list of current (11 today) openings, please see http://www.iter.org/jobs. Note that new jobs are posted frequently, and the application periods tend to be short, so those of you who are interested in working at ITER should check this website often (or enroll in email updates for job openings that is offered through this website).
Confinement and Transport Topical Group, Leaders: Gary Staebler and Saskia Mordijck
A flexible comprehensive integrated modeling framework capable of supporting both experimental analyses as well as predictive simulations is critical for whole device modeling research activities with present-day tokamak experiments and ITER. Much progress has been made recently in the development of the OMFIT framework and its many applications to support front-end scientific research at many institutions. In this month's Research Highlight by Orso Meneghini and Sterling Smith, an overview of the OMFIT framework is given including key elements of the framework that are critical for its quick adoption by a wide user community and many positive scientific impacts it is making.
A sustainable path towards successful integrated modeling
O. Meneghini1, S.P. Smith1, for the OMFIT development team1
1 General Atomics, PO BOX 85608, San Diego, CA 92186-5608, USA
Figure 1. Illustration of an OMFIT integrated modeling workflow with dynamic pedestal. Coupled modules include core transport (turbulence and neoclassical), current and source evolution, magnetostatic equilibrium, and pedestal stability. The only inputs to the model are the initial plasma equilibrium, the rotation profile, the electron density at the top of the pedestal, the setup of the heating sources, and an initial guess for the normalized plasma beta. Simulation for a DIII-D discharge shows good agreement with experimental measurements across the entire plasma radius.
One Modeling Framework for Integrated Tasks (OMFIT)  is a comprehensive integrated modeling framework which has been developed to enable physics codes to interact in complicated workflows, and support scientists at all stages of the modeling cycle. OMFIT development follows a unique bottom-up approach, where the framework design and capabilities organically evolve to support progressive integration of the components that are required to accomplish physics goals of increasing complexity. To date OMFIT has been used by over two hundred fusion scientists worldwide, in support of their experimental and theoretical research, for tasks that range from the automation of routine analyses to one-of-a-kind frontier research.
Notable usage examples are the workflows for generating full kinetic equilibrium reconstructions that in addition to magnetic and motional Stark effect measurement constraints utilize kinetic profile information including the fast-ion pressure modeled by a transport code. OMFIT is also used to streamline edge-stability analyses and evaluate the effect of resonant magnetic perturbations on the pedestal stability. The development of a set of neural-network-based models to efficiently reproduce the turbulent transport flux and pedestal structure calculations of computationally expensive first-principles codes was also supported by the framework. Concerning predictive transport simulations, the framework made possible the design and automation of a workflow capable of calculating the steady-state self-consistent solution to coupling the core transport, pedestal structure, current profile, and plasma equilibrium (see in Fig. *). The workflow leverages state-of-the-art components for collisional and turbulent core transport, equilibrium and pedestal stability. Testing against a DIII-D discharge shows that this workflow is capable of robustly predicting the kinetic profiles (electron and ion temperature and electron density) from the axis to the separatrix in good agreement with experiment . Finally, OMFIT provides an interactive front-end to a wide range of physics codes, spanning across all areas of magnetic fusion research. Such a rich set of results  provides a tangible evidence of how OMFIT has enabled a fast track to integrated modeling solutions that are functional, cost-effective, and in sync with the research efforts of the community.
Figure 1. Global statistics for the number of OMFIT sessions, users, and commits over the course of the past year. The usage of the framework is accelerating. Peaks in the number of users occurred when hands-on workshops were held. The project is vibrantly active, with an average of over 20 code commits per week. Dips in the metrics occur during holidays. To learn more about the OMFIT project visit http://gafusion.github.io/OMFIT-source.
From a technical standpoint OMFIT is a lightweight framework, meaning that it does not ship with the executable of the physics codes. Instead it can be thought of as "the glue" that holds codes, scripts, and data together. The idea at the heart of the framework is to treat files, modeling data, experimental databases, and scripts as a uniform collection of objects organized into a tree structure (the so called "OMFIT tree"). The framework then provides a consistent way to access and manipulate such a collection of heterogeneous objects, independent of their origin. Having a uniform data structure allows for a single top-level Graphical User Interface (GUI) to be used to interact with the tree data structure, carry out simulations, and analyze data interactively. Within OMFIT, modeling tasks are then organized into modules, which can be easily combined to create arbitrarily large multi-physics simulations.
The OMFIT approach does not specify, a priori, which components are to be coupled and how. Importantly, this allows it to leverage existing tools that have been developed over the years within the magnetic fusion community. This is absolutely critical for saving time and effort while being at the forefront of scientific research. The development model of the framework is such that the capabilities of the framework and of its physics components organically evolve to ensure that mission-critical capabilities are the first ones to be completed and refined. Progressively, the physics components that are required to solve problems of increasing complexity are integrated into OMFIT.
We have found that about 12% of the users become active developers, each contributing to the OMFIT physics modules that are most relevant for their research and expertise. As a result, OMFIT is fostering a growing community of enthusiastic researchers, who are genuinely invested in the project for the good of their own research. These user contributions allow the project to scale at all levels: from harnessing the collective intelligence of a talented group of individuals, to distributing the workload of development and user support, and supporting both code and scientific review. The statistics displayed in Fig.* well capture the positive trajectory of the project.
Establishing OMFIT as a practical framework for integrated modeling has certainly been a learning experience. Unarguably there are many aspects in our methodology that could be further improved, and we are always looking for ways to do so. Nonetheless, in view of our overall positive experience developing OMFIT, we think that the lessons that we have learned could benefit other magnetic fusion projects. Of these, let us underscore once more the importance of valuing user adoption and scientific impact as metrics for success, the adoption of a bottom-up development strategy that encourages community participation, and the crucial role of responsive support in retaining users.
This work was supported by the Office of Science of the U.S. Department of Energy under Contract Nos. DE-FG02-95ER54309 (GA theory), DE-FC02-04ER54698 (DIII-D), and DE-SC0012656 (GA AToM SciDAC).
 O. Meneghini et al., Nucl. Fusion 55 (2015) 083008  O. Meneghini et al,, Phys. of Plasmas 23 (2016) 4
More information concerning the ITPA may be found at the Official ITPA Website.
Integrated Operation Scenarios Topical Group
The 16th IOS-TG meeting (April 26-29) was hosted by IPP, Garching and jointly organized with EUROFusion.
In additional to regular reports on joint activities, presentations at this meeting did focus on (a) use of IMAS for experimental and modeling data sharing, (b) integrated control strategies, (c) experiments in Helium plasma in support of the ITER non-active phase, (d) experiments for ITER baseline demonstration discharges and advanced scenarios (e) experiments and modeling of actuators and (f) a new activity on termination strategies in support of ITER modeling.
(a) At the 2015 fall meeting the IOS-TG started a discussion on the use of IMAS for sharing experimental and simulation data. An action has been put forward to provide an example of the IMAS application for both experiments and modeling before the October meeting. Following the installation of TRANSP into IMAS and the wide use of the code for experimental interpretation, PPPL will provide the first translator from NETCDF TRANSP output and the ITER Data Structure (IDS).
(b) Combined control of current profile and βN control was demonstrated in the DIII-D high qmin steady-state scenarios, for a range of values of the safety factor. AUG showed results and future planning for actuator sharing allocation for combined control of NTM with ECCD, beta and divertor detachment.
(c) There has been an increasing emphasis on experiments in helium in support of the ITER non-active phase, aiming at comparing ELM regimes, pedestal structure, L-H transition power threshold compared to deuterium plasmas. At the fall 2015 meeting C. Kessel presented experiments from C-Mod. At the latest spring meeting experiments from AUG and from DIII-D were presented. Although the DIII-D experiments were very recent and not completely analyzed yet, similarities have been highlighted between AUG and DIII-D. In both cases it was stressed there is difficulty in controlling the density. AUG found that the ELM frequency increased by a factor ten in helium plasmas and the regime was of small ELMs at high divertor neutral densities, although the heating scheme was different (no ECH in helium, ECH in deuterium). On DIII-D type-I ELMs were found with NBI or ECH and power scans were carried out to find the LH threshold and how far above it in power did one need to be to obtain type-1 ELMs.
(d) Common conclusion among metal wall machines is that a way to operate at low collisionality has not been found yet. AUG finds a reduction in H98 by about 20% with tungsten compared to their best discharges with Carbon wall. Notably, AUG runs their baseline scenarios at a slightly higher q95=3.6, since they find difficult operating at q95=3.0. Experiments on DIII-D and EAST indicate that higher pedestal and wide ITBs are mutually exclusive. Also, experiments on ELM stabilization with RMP indicate that in the steady-state hybrid odd parity works better and is insensitive to the value of q95 in the range of 5.9-7.0. Focus on KSTAR was on obtaining stationary operation at high βN: values up to βN=2.5 have been obtained at 2T and 600kA, with 20s sustained H-mode, and fully non-inductive operation at lower current with βN=2, using NBI and on-axis ECH.
(e) C-Mod is operating until the end of September, focusing on expanding the I-mode regime and the EDA H-mode regime up to full field operation. Experiments with RF heating will use ICRF and LH for study of high-Z impurity transport and divertor physics. Recent LH modeling indicated that including a more consistent SOL in ray-tracing calculations would give better agreement with HXR measurement. Experiments with ICRF on AUG indicated reduction in the sputtering level with the 2 and 3 strap antenna and that the 3-strap antenna performs better in helium plasmas if it is tungsten coated. Using top injection valves - as planned on ITER - AUG also did not see clear benefit on ICRF coupling when the antenna is aligned with the valve region along magnetic field lines.
(f) A database of termination scenarios has been put together to identify whether assumptions made in the simulations for normalized quantities are consistent with experimental data. The database includes both large and low aspect ratio tokamaks and also data of automated ramp-down procedures from the recent campaign on NSTX-U.
2016 — 10th Anniversary of USBPO Formation —