Loading...
Please wait, while we are loading the content...
Similar Documents
Ga-a 24513 High Performance Stationary Discharges in the Diii-d Tokamak
| Content Provider | Semantic Scholar |
|---|---|
| Author | Wade, Michael R. Ferron, John R. Politzer, P. I. A. Hyatt, A. W. Sips, A. C. C. Murakami, Masami |
| Copyright Year | 2006 |
| Abstract | Recent experiments in the DIII-D tokamak [J.L. Luxon, Nucl. Fusion 42,614 (2002)l have demonstrated high p with good confinement quality under stationary conditions. Two classes of stationary discharges are observed low 495 discharges with sawteeth and higher 495 without sawteeth. The discharges are deemed stationary when the plasma conditions are maintained for times greater than the current profile relaxation time. In both cases the normalized fusion performance ( & H 8 9 p / 4 6 5 ) reaches or exceeds the value of this parameter projected for Qfus = 10 in the International Thermonuclear Experimental Reactor (ITER) design [R. Aymar, et al., Plasma Phys. Control. Fusion 44, 519 (2002)l. The presence of sawteeth reduces the maximum achievable normalized p, while confinement quality (confinement time relative to scalings) is largely independent of 495. Even with the reduced p limit, the normalized fusion performance maximizes at the lowest 495. Projections to burning plasma conditions are discussed, including the methodology of the projection and the key physics issues which still require investigation. ... GENERAL ATOMICS REPORT GA-A245 13 111 HIGH PERFORMANCE STATIONARY DISCHARGES IN THE D1II-D TOKAMAK I . INTRODUCTION T. C. Luce, et al. Discharges with normalized fusion performance well in excess of the International Thermonuclear Experimental Reactor (ITER) [ 11 and Fusion Ignition Research Experiment (FIRE) [2] baseline designs have been obtained under stationary conditions in the DIII-D tokamak [DIII-D]. The criteria by which this assessment is made are discussed later in this section. Several physics issues have motivated these experiments. Design of proposed future experiments is typically based on analysis of multi-machine databases. It is important to validate the design choices made on this basis on present-day machines. Further, it is important to validate the performance under stationary conditions with respect to relaxation time scales of the plasma. For example, recent experiments in the Tore Supra tokamak [3] show spontaneous oscillations in the plasma temperature in discharges operated under feedback control [4]. It is important to examine whether this is a generic feature of the underlying physics of the tokamak system. Finally, it is an interesting scientific question to determine what is the maximum possible fusion performance of a tokamak with the minimal required feedback control, e.g., without any active stabilization of plasma instabilities. The experiments reported here do not represent a definitive answer to these questions but in many respects determine the state of the art of such investigations. One additional practical motivation for these experiments is the desire in the ITER program to carry out limited nuclear testing of some components. The objective would be to maximize the neutron fluence. Obviously, the maximum fluence would be obtained if ITER could be operated continuously, even at significantly reduced fusion power and gain. Practical issues in the design and operation, such as the electrical and fuel costs of continuous operation and the high capital cost of a heat sink capable of continuous low gain operation, have led to the concept of “hybrid” operation. This pulsed operational scheme would maximize the fluence per pulse and would be limited by the installed heat sink and the available inductive flux. Discharges discussed here project to fluence per pulse approaching The performance claims for these experiments are based, in part, on dimensionless figures of merit for fusion performance and duration. With respect to duration, a necessary condition for steady-state operation is that no inductive flux is required to sustain the plasma current. The key figure of merit to assess this is the fraction of noninductive current = ZNI/Zwhere Z is the total plasma current andZNI is current sustained noninductively. The discharges discussed here typically have 0.5 and therefore are not consistent with steady state operation. However, they become stationary in the sense that the plasma parameters such as density, pressure, and current density are MW . a/m2, assuming heat sink limits of 3000 s. GENERAL ATOMICS REPORT GA-A245 13 1 T.C. Luce, et al. HIGH PERFORMANCE STATIONARY DISCHARGES IN THE D1II-D TOKAMAK not changing in time. The longest time scale in these plasmas is set by the equilibration of the current density profile. The figure of merit for reaching stationary conditions is taken to be the diffusive time for the lowest radial moment of the current profile at constant current zR = po (0) A/7ckfl, where (0) is the average conductivity over the cross-section of area A and kl l is the first zero of the Bessel function J1 [5]. For the discharges discussed here, zR 1-2 s. The validity of this figure of merit has been verified by monitoring the current profile evolution directly using motional Stark effect (MSE) [6] spectroscopy [7] . For evaluating fusion performance, the dimensionless parameter G = PN Hggp/q& is a convenient quantity. This quantity is related to fusion gain as follows. For deuterium plasmas, the fusion gain Qfus = pfUs/Pj, is proportional to nTz [8]. In a burning plasma, the situation is more complicated due to the self-heating by the a particles: Qfus = p f u s / ( f j o s s -Pa) where fj,,, is the power transported out of the plasma. Rewriting pfus = 5 Pa, the formula for Qfus can be rearranged to give Pa/f joss = Qfus/(Qfus +5) which is also proportional to nTz. Increases in the quantity nTz are limited by the pressure limit and by the confinement quality. The pressure limit is described by the quantity PN = P/(Z/aB) = nT/(ZB/a). The confinement quality can be described by various scaling laws [9,10,11]. All of the scalings are nearly linear in Z , so z = H Z ( z, , / I ) . Combining these, nTz = (&) ( H Z ) f (P ,B ,R ,u ,...) . (1) The limits on B are technological, but limits on Z for a given B are described by the safety factor 495 = BIZ. So finally a formula related to Qfus is obtained: The power scaling of the confinement is not explicitly factored out in this formula. The loss power is implicit in setting a target value for G based on a specific design, and, like B , it has technological limits based on wall loading rather than a physics constraint. For comparison purposes, the ITER-89P scaling will be used here. Two reasons motivate this choice. It is a global energy Confinement scaling and can be evaluated directly from the stored energy without correction for fast particle content (which is ~ 2 0 % in all cases discussed here). It also has a more realistic P scaling [23,24] than the IPB98y2 scaling [lo]. As will be evident, the high performance discharges discussed here push toward the P limits. Using a scaling which has pessimistic P scaling would inflate the confinement multipliers obtained in the evaluation. The remainder of the paper is organized along the following lines. A description of the various stationary high performance discharges encountered on DIII-D along with the key operational details are presented in Section 11. Section I11 discusses the impact of the 2 GENERAL ATOMICS REPORT GA-A245 13 HIGH PERFORMANCE STATIONARY DISCHARGES IN THE D1II-D TOKAMAK T.C. Luce, et al. magnetohydrodynamic (MHD) instabilities on limiting the current profile peaking, the stability limits and the confinement quality. Projections of specific discharges to ITER and FIRE design parameters are presented in Section IV, including the methodology of the projection. Conclusions based on the data presented are given in Section V. GENERAL ATOMICS REPORT GA-A245 13 3 HIGH PERFORMANCE STATIONARY DISCHARGES IN THE D1II-D TOKAMAK II. TYPES OF STATIONARY DISCHARGES T.C. Luce, et al. The peaking of the current density profile in tokamak discharges is limited, in most cases, by some MHD instability before resistive equilibrium is reached. Given the strong electron temperature dependence of the plasma conductivity = T-"" , the resultant current profile for a constant potential across the entire plasma would have q(0) << 1 except for cases with very high 495 or very large off-axis noninductive currents. The type of MHD instability which limits the current peaking provides a convenient way to classify the various stationary discharges obtained in DIII-D. The most common limiting case is that with a repetitive sawtooth instability [14,15] in the plasma center. Other instabilities which are observed to limit the peaking of the current profile are fishbones [16,17] and tearing modes [18,19]. In this section, high performance discharges from the DIII-D tokamak limited by each of these instabilities will be presented, followed by discussion of the operational recipe to establish these stationary discharges. Comparison with ongoing research in other tokamaks will also be mentioned. \ A. Stationary discharges limited by tearing modes Discharges with stationary pressure profiles for durations up to 3 zR in the DIII-D tokamak have been reported previously [7,20,21]. The current profiles reach stationary conditions with 495 > 4 and qmin 2 1 without sawteeth. The key element appears to be the presence of an m = 3/12 = 2 tearing mode located near the half minor radius. Recently, techniques have been developed to raise p in the discharges. As shown in Fig. 1, a discharge with 495 = 4.4 can be operated for -1 TR with PN = 3.2, which coincides with 4 . t i , which is a simple estimate of the no-wall n = 1 p limit [22]. Attempts to raise p further eventually lead to an m = 2 / n = 1 tearing mode which severely degrades confinement and which can lock, usually leading to disruption. This p limit will be discussed more extensively in Sec. 111. In the example shown in Fig. 1, the presence of an n = 3 tearing mode (likely m = 4 based on the rotation frequency) is sufficient to avoid sawteeth at lower p. As p is raised starting at 2400 ms, the m = 3/n = 2 mode appears. There are magnetic fluc |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://digital.library.unt.edu/ark:/67531/metadc781049/m2/1/high_res_d/828532.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
