Search

Start Over Domain Natural Sciences

121. Toward fusion plasma scenario planning for NSTX-U using machine-learning-accelerated models

Author(s):
Mark D. Boyer
Abstract:
One of the most promising devices for realizing power production through nuclear fusion is the tokamak. To maximize performance, it is preferable that tokamak reactors achieve advanced operating scenarios characterized by good plasma confinement, improved magnetohydrodynamic (MHD) stability, and a largely non-inductively driven plasma current. Such scenarios could enable steady-state reactor operation with high \emph{fusion gain} --- the ratio of produced fusion power to the external power provided through the plasma boundary. Precise and robust control of the evolution of the plasma boundary shape as well as the spatial distribution of the plasma current, density, temperature, and rotation will be essential to achieving and maintaining such scenarios. The complexity of the evolution of tokamak plasmas, arising due to nonlinearities and coupling between various parameters, motivates the use of model-based control algorithms that can account for the system dynamics. In this work, a learning-based accelerated model trained on data from the National Spherical Torus Experiment Upgrade (NSTX-U) is employed to develop planning and control strategies for regulating the density and temperature profile evolution around desired trajectories. The proposed model combines empirical scaling laws developed across multiple devices with neural networks trained on empirical data from NSTX-U and a database of first-principles-based computationally intensive simulations. The reduced execution time of the accelerated model will enable practical application of optimization algorithms and reinforcement learning approaches for scenario planning and control development. An initial demonstration of applying optimization approaches to the learning-based model is presented, including a strategy for mitigating the effect of leaving the finite validity range of the accelerated model. The approach shows promise for actuator planning between experiments and in real-time.
Type:
Dataset
Issue Date:
May 2020

122. Vertical forces during VDEs in an ITER plasma and the role of halo currents

Author(s):
Clasuer, C; Jardin, S; Ferraro, N
Abstract:
Vertical displacement events (VDEs) can occur in elongated tokamaks causing large currents to flow in the vessel and other adjacent metallic structures. To better understand the potential magnitude of the associated forces and the role of the so called ``halo currents'' on them, we have used the M3D-C1 code to simulate potential VDEs in ITER. We used actual values for the vessel resistivity and pre-quench temperatures and, unlike most of the previous studies, the halo region is naturally formed by triggering the thermal quench with an increase in the plasma thermal conductivity. We used the 2D non-linear version of the code and vary the post-thermal quench thermal conductivity profile as well as the boundary temperature in order to generate a wide range of possible cases that could occur in the experiment. We also show that, for a similar condition, increasing the halo current does not increase the total force on the wall since it is offset by a decrease in the toroidal contribution.
Type:
Dataset
Issue Date:
February 2020

123. Visual Analogy Extrapolation Challenge (VAEC)

Author(s):
Webb, Taylor; Dulberg, Zachary; Frankland, Steven; Petrov, Alexander; O'Reilly, Randall; Cohen, Jonathan
Abstract:
Extrapolation -- the ability to make inferences that go beyond the scope of one's experiences -- is a hallmark of human intelligence. By contrast, the generalization exhibited by contemporary neural network algorithms is largely limited to interpolation between data points in their training corpora. In this paper, we consider the challenge of learning representations that support extrapolation. We introduce a novel visual analogy benchmark that allows the graded evaluation of extrapolation as a function of distance from the convex domain defined by the training data. We also introduce a simple technique, context normalization, that encourages representations that emphasize the relations between objects. We find that this technique enables a significant improvement in the ability to extrapolate, considerably outperforming a number of competitive techniques.
Type:
Dataset and Image
Issue Date:
2020

124. A scalable real-time framework for Thomson scattering analysis: Application to NSTX-U

Author(s):
F. M. Laggner, A. Diallo, B. P. LeBlanc, R. Rozenblat, G. Tchilinguirian, E.Kolemen, the NSTX-U team
Abstract:
A detailed description of a prototype setup for real-time (rt) Thomson scattering (TS) analysis is presented and implemented in the multi-point Thomson scattering (MPTS) diagnostic system at the National Spherical Torus Experiment Upgrade(NSTX-U). The data acquisition hardware was upgraded with rt capable electronics (rt-analog digital converters (ADCs) and a rt server) that allow for fast digitization of the laser pulse signal of eight radial MPTS channels. In addition, a new TS spectrum analysis software for a rapid calculation of electron temperature (Te) and electron density (ne) was developed. Testing of the rt hardware and data analysis soft-ware was successfully completed and benchmarked against the standard, post-shot evaluation. Timing tests were performed showing that the end-to-end processing time was reproducibly below 17 ms for the duration of at least 5 s, meeting a 60 Hz deadline by the laser pulse repetition rate over the length of a NSTX-U discharge. The presented rt framework is designed to be scalable in system size, i.e. incorporation of additional radial channels by solely adding additional rt capable hardware. Furthermore, it is scalable in its operation duration and was continuously run for up to 30 min, making it an attractive solution for machines with long discharge duration such as advanced, non-inductive tokamaks or stellarators.
Type:
Dataset
Issue Date:
March 2019

125. Application of transient CHI plasma startup to future ST and AT devices

Author(s):
Hammond, K.C.; Raman, R.; Jardin, S.C.
Abstract:
Employment of non-inductive plasma start-up techniques would considerably simplify the design of a spherical tokamak fusion reactor. Transient coaxial helicity injection (CHI) is a promising method, expected to scale favorably to next-step reactors. However, the implications of reactor-relevant parameters on the initial breakdown phase for CHI have not yet been considered. Here, we evaluate CHI breakdown in reactor-like configurations using an extension of the Townsend avalanche theory. We find that a CHI electrode concept in which the outer vessel wall is biased to achieve breakdown, while previously successful on NSTX and HIT-II, may exhibit a severe weakness when scaled up to a reactor. On the other hand, concepts which employ localized biasing electrodes such as those used in QUEST would avoid this issue. Assuming that breakdown can be successfully attained, we then apply scaling relationships to predict plasma parameters attainable in the transient CHI discharge. Assuming the use of 1 Wb of injector flux, we find that plasma currents of 1 MA should be achievable. Furthermore, these plasmas are expected to Ohmically self-heat with more than 1 MW of power as they decay, facilitating efficient hand-off to steady-state heating sources. These optimistic scalings are supported by TSC simulations.
Type:
Dataset
Issue Date:
February 2019

126. Blob wakes in NSTX

Author(s):
Zweben SJ, Myra JR, Diallo A, Russell DA, Scotti F, Stotler DP
Abstract:
Transient small-scale structures were identified in the wake of blobs movingpoloidally through the SOL of high-powered H-mode plasmas in NSTX, using the gaspuff imaging (GPI) diagnostic. These blob wakes had a poloidal wavelength in therange 3.5 cm, which is significantly smaller than the average blob scale of~12 cm, and the wakes had a poloidal velocity of 1.5 km/sec in theelectron diamagnetic direction, which is opposite to the blob poloidal velocity inthese shots. These wakes were radially localized 0-4 cm outside the separatrix andoccurred within ~50 microsec after the passage of a blob through the GPI field of view.The clearest wakes were seen when the GPI viewing angle was well aligned with thelocal B field line, as expected for such small-scale structures given the diagnosticgeometry. A plausible theoretical interpretation of the wakes is discussed: theobserved wakes share some features of drift waves and/or drift-Alfven waves whichcould be excited
Type:
Dataset
Issue Date:
July 2019

127. CKavity Library: Next-Generation Sequencing

Abstract:
Protein sequence space is vast; nature uses only an infinitesimal fraction of possible sequences to sustain life. Are there solutions to biological problems other than those provided by nature? Can we create artificial proteins that sustain life? To investigate this question, the Hecht lab has created combinatorial collections, or libraries, of novel sequences with no homology to those found in living organisms. These libraries were subjected to screens and selections, leading to the identification of sequences with roles in catalysis, modulating gene regulation, and metal homeostasis. However, the resulting functional proteins formed dynamic rather than well-ordered structures. This impeded structural characterization and made it difficult to ascertain a mechanism of action. To address this, Christina Karas's thesis work focuses on developing a new model of libraries based on the de novo protein S-824, a four-helix bundle with a very stable three-dimensional structure. The first part of this research focused on mutagenesis of S-824 and characterization of the resulting proteins, revealing that this scaffold tolerates amino acid substitutions, including buried polar residues and the removal of hydrophobic side chains to create a putative cavity. Distinct from previous libraries, Karas targeted variability to a specific region of the protein, seeking to create a cavity and potential active site. The second part of this work details the design and creation of a library encoding 1.7 x 10^6 unique proteins, assembled from degenerate oligonucleotides. The third and fourth parts of this work cover the screening effort for a range of activities, both in vitro and in vivo. I found that this collection binds heme readily, leading to abundant peroxidase activity. Hits for lipase and phosphatase activity were also detected. This work details the development of a new strategy for creating de novo sequences geared toward function rather than structure.
Type:
Dataset

128. Climate Impacts from Large Volcanic Eruptions in a High-resolution Climate Model: the Importance of Forcing Structure

Author(s):
Yang, Wenchang; Vecchi, Gabriel; Fueglistaler, Stephan; Horowitz, Larry; Luet, David; Muñoz, Ángel; Paynter, David; Underwood, Seth
Abstract:
Explosive volcanic eruptions have large climate impacts, and can serve as observable tests of the climatic response to radiative forcing. Using a high resolution climate model, we contrast the climate responses to Pinatubo, with symmetric forcing, and those to Santa Maria and Agung, which had meridionally asymmetric forcing. Although Pinatubo had larger global-mean forcing, asymmetric forcing strongly shifts the latitude of tropical rainfall features, leading to larger local precipitation/TC changes. For example, North Atlantic TC activity over is enhanced/reduced by SH-forcing (Agung)/NH-forcing (Santa Maria), but changes little in response to the Pinatubo forcing. Moreover, the transient climate sensitivity estimated from the response to Santa Maria is 20% larger than that from Pinatubo or Agung. This spread in climatic impacts of volcanoes needs to be considered when evaluating the role of volcanoes in global and regional climate, and serves to contextualize the well-observed response to Pinatubo.
Type:
Dataset
Issue Date:
2019

129. Comment on ‘Numerical modeling of tokamak breakdown phase driven by pure Ohmic heating under ideal conditions’

Author(s):
Yoo, Min-Gu; Na, Yong-Su
Abstract:
In this comment, we point out possible critical numerical flaws of recent particle simulation studies (Jiang et al 2016 Nucl. Fusion 56 126017, Peng et al 2018 Nucl. Fusion 58 026007) on the electrical gas breakdown in a simple one-dimensional periodic slab geometry. We show that their observations on the effects of the ambipolar electric fields during the breakdown, such as the sudden reversal of the ion flow direction, could not be real physical phenomena but resulting from numerical artifacts violating the momentum conservation law. We show that an incomplete implementation of the direct-implicit scheme can cause the artificial electric fields and plasma transports resulting in fallacies in simulation results. We also discuss that their simple plasma model without considering poloidal magnetic fields seriously mislead the physical mechanism of the electrical gas breakdown because it cannot reflect important dominant plasma dynamics in the poloidal plane (Yoo et al 2018 Nat. Commun. 9 3523).
Type:
Dataset
Issue Date:
June 2019

130. Competitive learning modulates memory consolidation during sleep

Author(s):
Antony, James W.; Cheng, Larry Y.; Brooks, Paula P.; Paller, Ken A.; Norman, Kenneth A.
Abstract:
Competition between memories can cause weakening of those memories. Here we investigated memory competition during sleep in human participants by presenting auditory cues that had been linked to two distinct picture-location pairs during wake. We manipulated competition during learning by requiring participants to rehearse picture-location pairs associated with the same sound either competitively (choosing to rehearse one over the other, leading to greater competition) or separately; we hypothesized that greater competition during learning would lead to greater competition when memories were cued during sleep. With separate-pair learning, we found that cueing benefited spatial retention. With competitive-pair learning, no benefit of cueing was observed on retention, but cueing impaired retention of well-learned pairs (where we expected strong competition). During sleep, post-cue beta power (16–30 Hz) indexed competition and predicted forgetting, whereas sigma power (11–16 Hz) predicted subsequent retention. Taken together, these findings show that competition between memories during learning can modulate how they are consolidated during sleep.
Type:
Dataset
Issue Date:
November 2018