| 研究課題/領域番号 |
16K18341
|
|---|
| 研究機関 | 国立研究開発法人量子科学技術研究開発機構 |
|---|
研究代表者 |
Bierwage Andreas 国立研究開発法人量子科学技術研究開発機構, 六ヶ所核融合研究所 核融合炉システム研究開発部, 主幹研究員(定常) (10584691)
|
|---|
| 研究期間 (年度) |
2016-04-01 – 2020-03-31
|
| キーワード | Abrupt large events / Anomalous self-heating / Energetic particles / Hybrid simulation / Gyrokinetic simulation / Tokamak plasma |
|---|
| 研究実績の概要 |
1) A self-consistent simulation of the spontaneous onset of an Abrupt Large relaxation Event (ALE) in a JT-60U tokamak plasma was successfully performed for the first time using the MHD-PIC hybrid code MEGA. The results were reported as an oral presentation at the 26th IAEA Fusion Energy Conference in Kyoto. A paper describing the model and numerical methods used was published. One important implication is that the ALE trigger mechanism can be studied using the standard MHD model for the bulk plasma response. Unlike initially expected, the implementation of kinetic thermal ion effects is not essential, at least qualitatively.
2) Using the linear gyrokinetic code LIGKA, the feasibility of the Alfven acoustic self-heating channel was studied using an ALE-scenario of JT-60U. The simulation results allowed us to make quantitative estimates concerning the efficiency of Alfven acoustic self-heating. Based on these results, it is concluded that the Alfven acoustic self-heating channel in the form we had originally envisioned is probably negligible in comparison with the self-heating caused by direct kinetic damping of Alfven waves, which is estimated to be several orders of magnitude more efficient. The results were reported as an invited talk at a US-Japan Workshop held at General Atomics, San Diego, USA, and a paper was submitted.
3) The fact that we have successfully simulated Alfven acoustic couplings with a standard gyrokinetic model (LIGKA) implies that parallel magnetic fluctuations dB_|| play no essential role (other than cancelling a high-pressure drift correction term).
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
A major part of the original plan for FY2016 was to carry out simulations with an extended hybrid model. This was postponed because several successful achievements required immediate attention. We have successfully achieved three of the goals that were set in the original research plan, although in a different order:
1) Performed a self-consistent simulations of the spontaneous onset of an ALE. This was originally part of the plan for FY2017. We are now ready to proceed with the study of the physical mechanisms that trigger such kinds of abrupt relaxation events and validate the simulation results against experimental measurements, using the JT-60U data base.
2) Performed quantitative estimation of Alfven acoustic self-heating. This was originally part of plan for FY2019. We were able to complete this task early because we used an antenna-driven linear model instead of numerically expensive and physically complicated nonlinear simulations. Nonlinear simulations would be required to obtain the absolute amount of channelled energy. The antenna-driven linear simulation we used only gives the relative amount of energy transferred via two different channels. However, this was sufficient to demonstrate the inefficiency of the Alfven acoustic channel and identify a more important channel.
3) The results in item 2) indirectly show that further investigations of the role of parallel magnetic fluctuations dB_|| are no longer required. Therefore, we may regard the study of dB_|| effects as completed (part of the plans for FY2016 and FY2018).
|
| 今後の研究の推進方策 |
Based on the achievements summarized in 1)-3) above, we revised the research plan as follows.
* The role of dB_|| will not be investigated further.
* Instead, more emphasis will be put on the kinetic damping of Alfven waves and the modification of Alfven wave spectra by fast ions. In FY2017-2018, we will investigate these effects using an extended version of the LIGKA code, which includes finite-orbit-width effects. This is expected to yield important new insights that are relevant for several issues. For instance, we anticipate that these insights will allow us to propose a new explanation for experimentally observed modes known as "energetic-ion-driven geodesic acoustic modes (EGAM)", which are also thought to be important for noncollisional self-heating of fusion plasmas. In addition, the results will allow us to compute the accurate structure of continuous Alfven wave spectra in the presence of fast ions. This information will be needed to interpret results of nonlinear simulations and construct reduced models for so-called energetic particle modes (EPM), which is part of the research plan for FY2019.
* The use of the nonlinear code MEGA extended with kinetic thermal ion (KTI) effects will be postponed again. One reason is that such simulations are very expensive and it will be more efficient to wait for the new QST supercomputer to be installed in FY2018.
* Meanwhile, in FY2017-2018, we will use the limited resources we have on K and ICE-X supercomputers to perform reproducibility checks for ALE simulations. This is necessary for publishing these important results.
|
| 次年度使用額が生じた理由 |
When it became clear that no further study of parallel magnetic fluctuation effects is required (see research achievements item 3), it was also deemed unnecessary to visit UC Irvine, USA in FY2016. Therefore, the budget that was originally allocated for that trip became available for other purposes and future use.
|
| 次年度使用額の使用計画 |
The FY2017 plan includes a trip to Germany with the purpose to collaborate with Dr.Lauber at Max Planck IPP in Garching, who developed the LIGKA code that is used extensively in the present research. The funds transferred from FY2016 (230,210 JPY) will be used to extend the duration of the PI's trip to Garching. This will allow the PI to better familiarize himself with the latest version of LIGKA, contribute to the further development of the code, and thereby raise the possibility to achieve and even exceed the goals of this research project. The trip is expected to cost 650,000 JPY. Visits to QST Naka lab for validation and experimental data analysis will cost 200,000 JPY. The remaining funds (about 170,000 JPY) will be used to replace old hardware (PC, storage) used for this project.
|
