DIII-D RESEARCH OPPORTUNITIES FORUM FOR THE 2008 EXPERIMENTAL CAMPAIGN
Login | Review | Submit | Logout | Help
Questions about this website? Contact Andrew LeBlanc Questions about ROF? Contact Chuck Greenfield
| Title | 200: Edge Current Control of ELMs | ||
| Name: | James D. Callen ( |
Affiliation: | University of Wisconsin |
| Research Area: | ELM Control & Pedestal Physics | Presentation time: | Requested | Co-Author(s): | R.J. Groebner, who will present at ROF |
| Description: | OBJECTIVE: Control ELMs via reduction of the edge current density at rho ~ 0.85 using ECCD. | ||
| Experimental Approach/Plan: | PROPOSED EXPERIMENT: The basic idea would be to use (counter) ECCD just inside the top of the pedestal to arrest the recovery of the local current density there (from about 20 A/cm^2 to 40 A/cm^2) after an L-H transition and/or an ELM crash. That is, the objective would be to prevent the current density just inside of the pedestal top from becoming large enough and broad enough to add sufficient peeling drive to destabilize the peeling-ballooning mode and precipitate a Type I ELM. It would be desirable for the ECCD to be injected at about rho of 0.85 and be radially localized within about 10% of the plasma radius (FWHM ~ 7 cm?) about this point. The ECCD would hence be applied to a cross-sectional area of about 2500 cm^2. Thus, to achieve a desired local current density change of about 20 A/cm^2 at rho ~ 0.85 would apparently require about 50 kA of ECCD. But less than this might be sufficient to determine if Type I ELMs at low collisionality could be affected and/or controlled by ECCD at about rho of 0.85. The ECCD might also be able to arrest the slow radially inward propagation of the top of the density pedestal. | ||
| Background: | Recent analysis of DIII-D shots 118897 (Groebner) and 98889 (Osborne) have shown that after an L-H transition or an ELM, while the edge pedestal electron density and temperature profiles come into a "transport quasi-equilibrium" in less than about 10 ms, the pedestal density continues to evolve slowly -- the top of the density pedestal moves radially inward and increases in magnitude slowly (on 30 to 100 ms or longer time scales). The edge current density also continues to evolve slowly as the edge inductive electric field induced by the pedestal bootstrap current decays resistively on a commensurate local magnetic skin diffusion time scale. This slow current profile evolution may be the underlying cause of the continuing slow density pedestal evolution. A typical current profile evolution (from EFIT and ONETWO modeling) was presented in viewgraph 16 of Mickey Wade's 1999 Year End Review talk (http://fusion.gat.com/diii-d/ExpPlans00 ). Then, Type I ELMs apparently occur due to peeling-ballooning modes when the edge pressure gradient and current density are large enough AND broad enough (i.e., reach far enough into the plasma -- to rho < 0.85?). The experimental proposal would be to inject counter ECCD to arrest the increase of the current density at rho ~ 0.85 and thereby arrest the inward propagation of the density pedestal, prevent the increase of the current density just inside the edge pedestal, and thus prevent the plasma from evolving toward a peeling-ballooning instability that would precipitate an ELM. | ||
| Resource Requirements: | Shots similar to 118897 which had a long ELM-free period after an L-H transition, and/or 98889 which had a long period (~ 36 ms) between roughly equally spaced Type I ELMs. | ||
| Diagnostic Requirements: | Thomson, CER, CO2 Interferometer, edge MSE? | ||
| Analysis Requirements: | For developing a detailed scenario, revisit ONETWO analysis that was presented in viewgraph 16 of Mickey Wade's 1999 Year End Review talk including ECCD at rho ~ 0.85, and explore its possible effects on stability of peeling-ballooning modes. And after the experiments are performed redo such analysis for the particular shots with the specific ECCD that is feasible. | ||
| Other Requirements: | DIAGNOSTICS: Direct measurement of the temporal development around rho ~ 0.85 of the current density, q or poloidal magnetic field (via edge MSE?) would be highly desirable to see if the ECCD does produce the desired reduction in the local current density there.
ANALYSIS: The space-time development of such ECCD and edge pedestal bootstrap current effects which are radially localized to such a small region of the plasma (less than 15% of the plasma radius) would require either a large number of radial points in ONETWO (>~ 200?), or packing of the spatial grid near the edge. This would likely require some careful work and testing of the ONETWO code to efficiently handle such complications. | ||
Print this page