In a tokamak fusion reactor, confining plasma hotter than 100 million degrees with magnetic fields requires a current of several million amperes flowing through the plasma. The magnetic field created by this current combines with the field from external coils to form the helical magnetic structure that confines the plasma. In conventional tokamaks, however, most of this current has been generated by transformer action in a large electromagnet called the central solenoid. Because this relies on the same principle as a transformer, the available change in magnetic flux is limited, and the discharge must eventually end. Overcoming this constraint has long been a challenge for continuous operation of fusion power plants.
A research team including General Atomics in the United States has, for the first time, captured experimental evidence of "helicon current drive," in which radio-frequency electromagnetic waves generate current inside the plasma without relying on electromagnetic induction. The results are reported in the peer-reviewed journal Physical Review Letters in the paper "First Experimental Evidence of Helicon Current Drive". However, the experiment did not achieve continuous fusion power or net energy generation. The radio-frequency power coupled to the plasma was 370 kW, and the driven current is estimated at 20±10 kA, concentrated near the magnetic axis. Even so, it supports the idea that radio-frequency waves can be sent deep into a high-density plasma to drive current externally, making it an important proof of principle for future fusion reactor designs.
The limits of transformer action and momentum transfer by a "fast wave"
Continuous tokamak operation requires "non-inductive current drive," which sustains plasma current without electromagnetic induction. This calls for technologies that supplement the bootstrap current, which arises naturally from pressure gradients inside the plasma, by injecting particle beams or radio-frequency electromagnetic waves from outside.
Representative methods include current drive using Lower Hybrid Waves and Electron Cyclotron Waves. In the hot, dense plasmas expected in future fusion reactors, however, each faces challenges. Lower hybrid waves have limited propagation to the plasma core as density rises, and electron cyclotron current drive faces issues of equipment cost and power conversion efficiency at high power.
Helicon waves, long studied as an alternative, are a type of Fast Wave in the lower hybrid frequency range. Also called whistler waves, these electromagnetic waves propagate readily to the core even in high-density plasmas and are thought to be able to give electrons both energy and toroidal momentum (momentum along the circumference of the torus). Specifically, electrons moving along magnetic field lines at speeds close to the wave's phase velocity receive energy from the wave through interactions such as Landau damping. By selectively launching waves that propagate in a particular direction, electrons moving in one direction are preferentially accelerated, generating current.
Although the principle itself has been known for some time, experimental demonstration took many years. It was difficult to develop an antenna that could couple radio-frequency power stably and propagate waves in the intended direction even as the boundary conditions of a hot plasma change. The DIII-D team built its experimental setup by combining a comb-line traveling-wave antenna installed on the inner wall of the vacuum vessel with a 476 MHz radio-frequency source. The source uses a klystron rated at 1.2 MW, but the power actually coupled to the plasma in this current drive experiment was 370 kW. The rated output of the equipment and the power delivered to the plasma need to be distinguished.
How heating-induced resistivity changes were separated from current drive
When radio-frequency waves are injected into plasma, it is not easy to tell whether current increased because electrons were accelerated in a particular direction or because the plasma was simply heated. The electrical resistivity of plasma is roughly inversely proportional to the electron temperature to the 1.5 power. Even if the radio-frequency waves caused only heating, a rise in electron temperature would lower resistivity and change the distribution of the existing inductive current. This effect could be mistaken for newly driven current.
To address this, the research team led by Jeff Lestz and colleagues conducted control experiments to separate the effects of heating and current drive. Using stable L-mode discharges that are easy to compare, they compared three types of discharge conditions:
- Reference discharges: Basic discharges sustained by 2.3 MW of neutral beam injection (NBI) and 1.5 MW of electron cyclotron heating (ECH).
- Helicon-applied discharges: Discharges in which 370 kW of 476 MHz helicon waves was injected for one second on top of the reference conditions.
- Electron-heating control discharges: Discharges with no helicon injection, in which additional ECH equivalent to the predicted helicon absorption was applied instead.
When the helicon waves were injected, the electron temperature near the magnetic axis rose by about 1 keV. If the change in current distribution could be explained by the temperature rise alone, the control discharges with a similar amount of ECH heating should have shown the same change. The team examined this difference through the magnetic field structure inside the plasma.
The analysis used magnetic field measurements from the Motional Stark Effect (MSE) and the equilibrium reconstruction code EFIT. The team focused on the distribution of the safety factor ($q$), which describes how twisted the magnetic field lines are. The safety factor indicates how many times a field line travels around the toroidal direction while going around once in the poloidal direction (the direction around the torus cross-section). The analysis showed that in the helicon-injected discharges, the safety factor profile near the magnetic axis flattened clearly faster than in the reference discharges or the ECH heating control discharges.
Differences also appeared in the "sawtooth oscillation," a type of magnetohydrodynamic instability that occurs in the plasma core. In general, the safety factor near the magnetic axis reaching 1 is closely tied to the onset of sawtooth crashes. In discharges where helicon waves were injected in the same direction as the existing plasma current (co-current), sawteeth appeared earlier than in the reference and heating control discharges. This result is consistent with the interpretation that additional co-directional non-inductive current near the magnetic axis accelerated the decline of the safety factor.
The team also considered whether the observed current drive came from the helicon wave itself or from mode conversion to a lower hybrid wave. In similar past experiments, the possibility that an unintentionally generated Slow Wave drove the current had been an issue. Under the conditions of this experiment, however, the density profile is such that a 476 MHz slow wave cannot propagate to the plasma core. The changes observed near the magnetic axis therefore cannot be explained by indirect current drive from lower hybrid waves, supporting direct current drive by helicon waves.
| Discharge group | Added conditions / power | Purpose of experiment | Observed behavior | Physical interpretation and caveats |
|---|---|---|---|---|
| Reference discharge | NBI 2.3 MW + ECH 1.5 MW | Establish baseline conditions for comparison | Safety factor changes gradually as inductive current diffuses | Baseline state with no additional current drive from helicon waves |
| Helicon-applied discharge | Reference conditions + 370 kW helicon waves (1 second) | Detect current drive effect of helicon waves | Safety factor profile flattens rapidly; sawteeth appear earlier | Suggests added co-directional current that heating alone cannot explain, but is not a direct measurement of total driven current |
| Electron-heating control discharge | Reference conditions + additional ECH equivalent to predicted helicon absorption | Separate the effect of resistivity changes from the temperature rise | Safety factor change is close to the reference discharge; sawteeth are not similarly advanced | Reduced resistivity from a temperature rise alone cannot explain the changes seen with helicon injection |
データを表で見る
| Power (MW) | |
|---|---|
| Baseline heating NBI | 2.3 |
| Baseline heating ECH | 1.5 |
| RF source rating | 1.2 |
| Helicon power actually coupled | 0.37 |
The 370 kW of helicon power coupled to the plasma in this experiment is about one-tenth of the 3.8 MW baseline heating power from NBI and ECH combined. A key point of this study is that the effect of this relatively small additional power was detected by combining measurements of temperature profiles and magnetic field structure with control experiments. However, because the heating methods and energy transfer processes differ, current drive efficiency cannot be compared on this power ratio alone.
About 20 kA of driven current near the magnetic axis, and its uncertainty
How much current was actually driven by the helicon waves? Understanding this figure requires attention to the measurement method and the limits of the analysis. Because an ammeter cannot be placed directly inside a hot plasma, the team estimated the driven current by combining magnetic field measurements with numerical models.
The analysis used data from 1360 to 1560 milliseconds after the start of the discharge, a stable period before sawteeth appear. The team compared a total of seven shots: five with helicon injection plus the reference and ECH control discharges. From the MSE diagnostic and EFIT equilibrium reconstruction based on magnetic measurements, they obtained the parallel current density profile inside the plasma. They then subtracted the calculated Ohmic inductive current, bootstrap current, neutral beam current, and electron cyclotron current to derive the current component attributable to the helicon waves. The resulting current density was concentrated near the magnetic axis.
The current was concentrated in the region where the normalized radius is about 0.2 or less. Here, is a dimensionless radial coordinate, with the magnetic axis at 0 and the outermost magnetic surface of the plasma at 1. The helicon-driven current integrated over was estimated at 20±10 kA. With an uncertainty of ±50% on the estimate, it cannot be treated as a precise measurement.
The main reason for the large uncertainty is that the current density profile obtained from the magnetic equilibrium reconstruction contains unphysical oscillations. Even if the oscillations in the current density itself are small, their effect becomes non-negligible when integrated over the radius. Especially far from the magnetic axis, the area being integrated grows, so errors increase. For this reason, the team limited the integration to $ ho \approx 0.2$, where the reconstructed current density first changes sign, and reported the result to one significant figure.
Meanwhile, this local current distribution broadly agrees with the predictions of GENRAY, a ray-tracing code that calculates radio-frequency propagation. GENRAY predicted 21 kA within and a total of 26 kA driven during the first pass through the plasma. However, the peak of the current density predicted by GENRAY is sensitive to the ray trajectories set in the simulation and can change with the calculation conditions. Agreement between experimental and calculated values supports the theoretical model, but does not by itself constitute independent additional proof.
Because the EFIT reconstruction shows unphysical oscillations away from the magnetic axis, the total driven current across the entire plasma could not be determined with sufficient accuracy. The 20±10 kA obtained here is only the current estimated in a limited region near the magnetic axis and does not represent the driven current of the whole plasma. This distinction is also important when applying the experimental results to future fusion reactors.
What this experiment has not demonstrated
This result supports the principle of helicon current drive in an actual tokamak plasma, but many challenges remain before it can be applied to a steady-state fusion reactor. In particular, three points need further verification: the direction of current drive, the absorption of injected power, and operation in high-performance plasmas.
First, there are constraints from the antenna's operating conditions. In this experiment, only one side of the traveling-wave antenna was working properly, so it was not possible to directly compare "co-directional injection," which drives current in the same direction as the existing plasma current, with "counter-directional injection," which drives it in the opposite direction. If it could be confirmed that counter-directional injection reverses the change in current distribution, the conclusion that helicon waves drive current could be tested more directly.
Second, the amount of radio-frequency power absorbed and the current drive efficiency have not been sufficiently quantified. GENRAY calculations predict that, in this current drive experiment, about 47% of the injected power is absorbed by the helicon waves on the first pass. The remaining power may be absorbed during multiple passes through the plasma after reflection from the wall, but the team has not been able to accurately pin down that process.
The study estimated the radio-frequency power absorption profile from the electron temperature response and also confirmed consistency with the theoretical model. However, because changes in the temperature profile from heat transport and losses from multiple passes could not be sufficiently separated, a precise evaluation of absorption was not achieved. For this reason, it also cannot be distinguished whether the observed driven current is smaller than the theoretical expectation because the power was not sufficiently absorbed or because current drive efficiency for the absorbed power was low. The team lists this as an issue for future verification.
Third, the current drive experiment was limited to L-mode discharges. In future fusion reactors, H-mode, with improved confinement, and advanced tokamak operating regimes with a higher fraction of bootstrap current will be important. DIII-D has also reported helicon wave injection and core heating in H-mode plasmas, but the quantitative demonstration of current drive shown in this paper was carried out in L-mode. Whether radio-frequency waves can be absorbed at the intended location and drive sufficient current in plasmas with higher pressure or different current profiles remains to be verified.
Of course, this experiment also did not demonstrate net energy generation from fusion reactions or long-duration steady-state operation. The experiment was carried out in discharges lasting several seconds, and helicon waves were injected continuously for one second. There is still a large gap between confirming the physical mechanism of current drive and sustaining the current needed in a power reactor over a long time.
What must be verified for steady-state operation of power reactors
Even with these limitations, the result is significant. It provides evidence that helicon current drive actually works as a means of reducing dependence on the central solenoid, which has hindered continuous operation of tokamak fusion reactors. Combined with existing bootstrap current, neutral beam, and radio-frequency current drive, it widens the options for maintaining the current profile needed in a future reactor core.
Helicon waves are predicted to be potentially more efficient than existing current drive methods such as electron cyclotron waves in the hot, dense plasmas of the future. In particular, if current can be supplied to a targeted location deep in the plasma, it could be useful for maintaining magnetic stability while tuning the current profile. However, this experiment alone cannot conclude that helicon waves will be more efficient than existing technologies in future fusion reactors.
The following verifications will be important for practical use:
- Comparison of bidirectional injection: Inject radio-frequency waves from both directions of the antenna and directly measure how the current profile changes between co-directional and counter-directional injection.
- Quantifying the absorption process: Obtain an absorption profile corrected for heat transport effects, and distinguish power absorbed on the first pass from power lost through multiple passes or at the wall.
- Evaluating current drive efficiency: Quantify the current drive efficiency with its uncertainty, and advance comparisons with electron cyclotron current drive and neutral beam current drive.
- Demonstration in high-performance regimes: Confirm in H-mode and advanced tokamak regimes with a high bootstrap current fraction that helicon waves propagate to the intended location and stably drive the required current.
The estimated 20±10 kA is small compared with the multi-million-ampere currents needed in a fusion power reactor. But its significance lies not in the absolute value of the current, but in the fact that control experiments and magnetic field measurements support the ability of helicon waves to generate non-inductive current deep in the plasma. If quantification of absorption efficiency and driven current advances, and similar effects can be confirmed in high-performance plasmas, helicon waves could become an important current drive tool for tokamaks aiming at steady-state operation.