Working Group 2 - Heating, Current Formation and Sustainment by Fast Mechanisms

D. B. Batchelor (chair), R. Majeski, T. Bigelow, D. D’Ippolito, R Pinsker, L. Grisham, T. Intrator, T. K. Mau

with contributions from

M. Carter, D. Darrow, P. Efthimion, D. Jassby, J. Menard, P. Moroz, W. M. Nevins, A. Ram

Working group 2 identified 5 basic objectives for our discussions (some of which were accomplished):

1) Identify the possibilities for various heating and current drive approaches

4) Identify hardware, diagnostic, or theory needs to test the various possibilities

5) Determine required plasma conditions for a given technique to function

Four types of technique were considered:

1) RF in the ion cyclotron range of frequencies:

2) Neutral injection

3) RF in the electron cyclotron range of frequencies:

4) Current ramp by bootstrap

Although strictly speaking the last of these does not involve fast particle interactions, the success of the technique will rely on the ability to heat rapidly and to maintain a controlled pressure profile. It therefore fit appropriately within our working group.

RF in the Ion Cyclotron Range of Frequencies

The primary critical issues for use of HHFW in NSTX relate to compatibility of HHFW with neutral injection and antenna issues. Because of the large plasma dielectric constant (i.e. short wave length) fast waves can couple to energetic ions even up to quite high harmonics. For example in 2D full wave calculations including a hot ion component with , and up to 40% of the power was absorbed by the hot ion component. High bulk ion temperatures can also pose a problem from the standpoint of linear cyclotron harmonic absorption and conversion to high harmonic ion Bernstein modes. The degree to which this is a problem depends sensitively on the hot ion concentration, energy spectrum and profile, and requires further study. The concern is both that the electron deposition and current drive will be reduced and that the fast waves will produce trapped energetic ions and cause losses. There are however possibilities that synergies exist between the waves and energetic ions which can be beneficially employed.

Other possibilities for use of ICRF in NSTX: for Alfven waves, RF helicity injection and non-linear mechanisms such as rotating magnetic field current drive.

Research needs include a number of areas of theory and modeling:

Experimental research is also needed, particularly in the areas of antenna and edge plasma data base: CDX-U, DIII-D, PEGASUS (future)

Diagnostic needs include:

We suggest the following general question for both ICRH and NBI: Is there a credible non-inductive ramp-to-full-performance scenario, interfacing with either CHI or ECH startup?

Neutral beams are a proven technique for heating and current drive in tokamaks of conventional aspect ratio. It is planned to install one TFTR neutral beam line on NSTX as early as possible. Not only does the NBI provide heating, current drive and some central fueling, but is also needed for diagnostics such as MSE and CHERS. The beam line will be operated at 80keV for 5 MW of total injected power. Modeling with TSC and TRANSP indicates ~300kA of beam current drive.

As with ICRF, neutral injection on NSTX differs from conventional tokamaks as a result of the low aspect ratio and low magnetic field. These two differences give rise to large trapped particle fractions, as well as large ion gyroradii, , and trapped particle orbits. Hence, there is concern about orbit and ripple loss and the resulting localized heat loads on plasma-facing components. Specific loss mechanisms of concern for NSTX include:

NSTX is also new territory with respect to beam induced instabilities such as fishbones. Of particular concern is susceptibility to TAE instabilities since the Alfven speed for NSTX is so slow, 0.4 < vbeam/vA < 5., and the low aspect ratio widens the Alfven eigenmode gaps. There is also the possibility, as yet unexplored, of velocity space instabilities due to the magnetic geometry.

With a single beam, central fueling is quite modest. It would therefore be desirable to have pellets for central fueling, density control and for shine-through reduction in target plasma. For purposes of optimizing bootstrap currents, density control and profile peaking may be particularly important.

The most immediate research needs for NBI include more modeling to choose a beam aiming angle which optimizes CD, beam ion confinement, and shine-through. Shine-through calculations are needed for wall armor design. There are also theory needs for validating guiding center codes against full Lorentz force equation orbit calculations for the low magnetic fields and aspect ratios of NSTX.

Initial estimates of the minimum target plasma density needed for NBI are central (/volume average?) . This will however depend on profiles and wall armor design so more studies are needed.

A number of specific diagnostics will be required to assess fast ion confinement and loss:

The importance of neutron diagnostics to determine the effectiveness of heating and current drive techniques and to verfiy the soundness of plasma modeling was pointed out. Neutron "shortfall" correlates closely with fast-ion loss, presence of beam injected or RF produced tails, and can be used to detect small changes in Ti. It has been proposed that the TFTR fission-chamber neutron detection system be used on NSTX with little modificaiton. It is estimated that rates down to n/s could be measured.

There are a number of basic beam/plasma physics issues which can be studied advantageously on NSTX. These include:

It appears feasible on NSTX to use ECH for pre-ionization, plasma and current buildup and even bulk heating and current drive. At present there are only plans to provide 15-20 kW of power at 18 GHz for 10-20 msec for pre ionization. However NSTX-relevant ECH hardware available from ORNL could provide significantly more power at minimal costs - two 28 GHz gyratrons providing up to 400 kW. These gyratrons could likely also be operated at lower frequency using a different cavity mode, 15.3GHz, although probably at lower power.

ECH pre-ionization is an established technique which will provide reliable initialization of NSTX plasma to control timing and location of the initial current channel. There are no significant issues related to that.

Another exciting possibility is to use Electron Bernstein Wave (EBW) for plasma buildup and current drive or for bulk heating and current drive during high density conditions. Absorptivity of 3 harmonics appears quite high. Three options for wave accessibility show promise but need further investigation:

Related to these techniques is a proposal to use radiated EBW as an electron temperature diagnostic in lieu of ECE. Again because of the large value of the first 5 or so cyclotron harmonics are cut off and cannot be used for ECE. Estimates indicate that the EBW should be optically thick on NSTX. There are 2 approaches for detecting the EBW: 1) employ an electrostatic antenna to receive EBW propagating to the edge, 1) Detect X-mode which has been converted from EBW at the upper hybrid layer. This is essentially the inverse of the heating process proposed above.

A further option is to use fundamental absorption of the 28 GHz ECH at the inside plasma edge to drive an edge current by the standard Fisch-Boozer mechanism. Such an current is needed to in some of the NSTX scenarios to supplement the bootstrap current and provide edge stability. By absorbing the power on the inside edge, trapped particle effects are reduced. This technique would probably require a more elaborate launching system of reflectors to launch the power from the high field side.

The immediate research need is for detailed analysis with NSTX geometry to evaluate each of the above scenarios. In particular it is important to assess the various startup and current ramp techniques as an alternative or supplement to CHI. Further experiments on DIII-D, C-MOD, and CDX-U would also be helpful. It appears that a proof of principle for the EBW temperature diagnostic could be performed on CDX-U. Of course a critical issue is for high power ECH to me made available on NSTX.

 T. Intrator, R. Fonck, B. Lewicki, T. Thorson, K. Tritz, G. Winz

University of Wisconsin - Madison

Jon Menard

High frequency fast waves are to play an important role in heating and driving current in high beta NSTX plasmas. Of major concern is the effect of the extreme field line pitch (up to 60 degrees) of the equilibrium magnetic field on the coupling physics of the fast wave. First, one would like to know how (or if) the angle between the straps and magnetic field affects how much power goes into different wave branches. Second, even if most of the power goes into the fast wave independent of strap angle, there may be a loading resistance variation with both phasing and angle. Third, even if all the power goes into the fast wave and loading variations are moderate as the field-line pitch and/or phasing change, what happens to the heating/current drive deposition profiles as these parameters change?

To begin to address some of these concerns regarding high frequency fast waves on NSTX, a 2 strap antenna which can be manually rotated between shots was installed in the CDX-U tokamak. Thus far, most of the RF experiments performed on CDX-U have concentrated on the loading issues mentioned above. In particular, it has been observed that the plasma loading resistance is nearly independent of strap angle with respect to the equilibrium magnetic field for 0-pi phasing. This is in agreement with a coupling code used for modeling fast wave loading on CDX-U (ANT1D). This code does, however, predict a significant increase in FW loading for 0-pi/2 phasing (i.e. current drive phasing) as the straps become increasingly parallel to B. This loading dependence on phasing and angle has not yet been tested on CDX-U, but will be in the near future. If there is a significant loading dependence on phasing and strap angle, using fast waves for both startup and current sustainment in NSTX may be difficult due to different matching requirements during different parts of the discharge.

Thus, while this may already be obvious to everyone else, these considerations only highlight the obvious need for controlled FW experiments on NSTX which test the phasing and angle dependence of the coupling physics. The phasing variation is trivial, while the angle dependence is more difficult, since one would like to vary the current to change the poloidal field magnitude while keeping the temperature and density profiles fixed at the edge. Thus, B-dot probes might be very useful for measuring wave polarization, loading can be determined from the usual power balance and strap currents, and edge profile diagnostics are very important here as well if any modeling is to be pursued.

 D. A. D'Ippolito and J. R. Myra

In conclusion, a number of techniques exist for controlling RF sheaths and experiences on other experiments indicate that some effort should be expended in this area in order to have optimal FW heating and current drive.

7. J.A. Rice, C.S. Hazelton, D. A. D’Ippolito, and J. R. Myra, DOE SBIR Phase I Final Report (1997).

M.D. Carter, E.F. Jaeger, and D.B. Batchelor

Oak Ridge National Laboratory

An appropriately designed HHFW system can allow experiments to be conducted with significantly different current profiles for turbulence and transport studies. Substantially different current profiles are possible for some plasma regimes by alternatively driving the HHFW system in the direction of, or opposite to the direction of the bootstrap current. Toroidal phase control for the antenna system can also provide current profile control in plasma regimes that produce moderately strong first pass absorption.

The results to date can be summarized as follows: (1) The theoretical current drive efficiency for HHFW can be very high, and this study provides a good target for the actual antenna design. Non-zero poloidal mode excitation provides the best efficiency because of improved penetration to the high field region where trapped particle effects are reduced. However, limitations on antenna placement and phase control can substantially reduce the efficiency. (2) During plasma initiation, the primary issue for HHFW is loading since the CD capability is good. More profile information during start-up is needed to complete this study. (3) Standard launch from the equatorial plane leads to marginal CD performance at high density and full current, but advanced antenna designs exploiting the theoretical results show promise. For example, poloidal phasing with equatorial launch can add roughly 30% to the CD efficiency. Placement of the antenna away from the equatorial plane near the top or bottom of the tokamak can improve efficiency by about 100%. Other issues that have been addressed include the variation in strap loading during the current ramp and the strap-to-strap loading required for stable phase control of the antenna.

Perhaps the most important issue to still address is the edge phenomena that may limit the power handling capability of the antenna. HHFW produces, and in fact requires substantial electric field components parallel to B. These parallel fields may require careful consideration when designing the lateral protection for the antenna to avoid sheath rectification and heating. Large parallel fields may also lead to substantial ponderomotive effects because of the relatively low confining magnetic field.

[1] E.F. Jaeger, D.B. Batchelor, D.C. Stallings, Nuclear Fusion 33 (1993) 179.

[2] M.D. Carter et.al., Nucl. Fusion 36 (1996) 209.

 Improved Modelling of HHFW Current Drive Scenarios

 T.K. Mau*

University of California, San Diego

--------------------------------

*In collaboration with S.C. Chiu, General Atomics.

[1] M. Ono, Report PPPL-3091 (1995).

[2] T.K. Mau, et al., EPS Top. Conf. on RF Heating and Current Drive of Fusion

Devices, Brussels (1992) 181.

[3] S.C. Chiu, et al., Sherwood Theory Meeting (1996).

[4] M. Brambilla, Report IPP 5/66 (1996).

 by Paul Moroz, Univ. Wisconsin-Madison

Spherical tokamaks (ST) normally feature relatively high density, low magnetic field, and low temperature regimes of operation. Moreover, the trapped particle effects are very significant. As a result, many standard RFCD methods and neutral beam injection become not very efficient. CD methods via the fast waves at high harmonic (HHCD) frequencies, presently proposed for use in NSTX, although possible are still difficult and tricky especially in providing an efficient current profile control.

This proposal is focused on theoretical analysis and full-wave numerical simulation of a few non-standard CD techniques that can be applied to NSTX and will supplement the so far leading methods of HHCD and NBI. The idea is to search and investigate the CD methods that fit specifically to the unique ST geometry and parameters. In particular we are planning to focus on RF methods that are due to wave interaction with the bulk plasma and not only with the narrow region (in velocity space) of the resonant particle population. Some of these methods have been discussed in literature although presently none of them is developed enough to be directly used for NSTX. Among them are RF helicity current drive, rotating magnetic field current drive, current drive due to ponderomotive effects, or magnetic field generation due to nonlinear effects. Some other novel CD techniques might be found during the course of research. Such methods might be less affected by the trapping effect and might actually work in an ST better than the standard CD methods.

Our proposal includes theoretical and numerical parts. Also, active collaboration is proposed with NSTX engineers and scientists for development of CD techniques and appropriate antenna systems. For numerical calculations, we are planning to extend the ALFA code for correct treatment of the novel CD methods under investigation. The ALFA code [1-11] has been developed by the author during a number of years and has been used mostly for the analysis of FWCD or Alfven wave CD. We believe that the ALFA code is advantageous for use in calculations for ST configurations. ALFA features fully toroidal geometry, arbitrary shaped plasma, and arbitrary strong poloidal magnetic field effects. This is the hot-plasma code with the full dielectric tensor which includes the Bessel functions. This feature is important when the finite Larmor radius effects are substantial, which is relevant to the NSTX parameters. The ALFA code computes a 3D picture of RF wave fields and absorbed power. Transient processes of current diffusion are included in the model as well [7-8]. The code has an extended diagnostics with the possible output of up to 300 various plots during a single run of the code. The code has been used for RF heating and current drive analysis in TPX [1-4], Phaedrus-T [5-8], IGNITOR [9], JET [10-11], and ITER [6-7]. Our past theoretical studies of RF helicity CD [12-15] add to a solid starting base for the proposed research.

[1] P.E. Moroz, 15th Int. Conf. on Num. Simul. of Plasmas, Valley Forge, PA, 3B2 (1994).

[2] P.E. Moroz, D.B. Batchelor, R.W. Harvey, M. Porkolab, Sher. Conf., Dallas, 1D20 (1994).

[3] P.E. Moroz, D.B. Batchelor, E.F. Jaeger, T.K. Mau, D.R. Mikkelsen, M. Porkolab, in Radio-Frequency Power in Plasmas, AIP Conf. Proc., 289, p. 218 (1994).

[4] P.E. Moroz, 7th Boulder Int. Workshop on RFCD, Boulder, Colorado (1994). [5] P.E. Moroz, 36th APS Meeting, Bull. Am. Phys. Soc. 39, 1630 (1994).

[6] N. Hershkowitz, P.E. Moroz, et al., Proc. 15th IAEA Conf. PPCFR, Madrid (1994).

[7] P.E. Moroz, Proc. 11th Top. Conf. on RF Power in Plasmas, Palm Springs (1995).

[8] P.E. Moroz, 37th APS Meeting, Louisville, Bull. Am. Phys. Soc. 40, 1710 (1995).

[9] P.E. Moroz, 22nd EPS Conf. Contr. Fus. & Pl. Phys., Bournemouth, 19C, III-333 (1995).

[10] P.E. Moroz, J.R. Myra et al., 37th APS Meeting, Bull. Am. Phys. Soc. 40, 1767 (1995).

[11] P.E. Moroz, J.R. Myra, D.A. D'Ippolito, 23rd IEEE Conf. Pl. Science, Boston, 191 (1996).

[12] P. E. Moroz, Proc. 20th EPS Conf. on CFPP Lisboa, Portugal, Vol. 17C, Part III, 849 (1993).

[13] P. E. Moroz, ICRF Modeling and Theory Workshop, PPPL, Princeton, NJ (1992).

[14] P. E. Moroz, University of Wisconsin-Madison, Report PLR-92-17 (1992).

[15] P. E. Moroz, Plasma Phys. and Contr. Fusion 35, 1145 (1993).

 Doug Darrow

Theoretical and experimental studies of fast ion confinement are of importance to all magnetic fusion plasmas, and these subjects have a direct bearing on the efficiency of auxiliary heating and current drive in STs. The subject has been extensively studied in conventional aspect ratio tokamaks, but a number of unique physics issues are raised by the geometry of ST plasmas. These include: a large gyroradius in comparison to the plasma size, large trapped particle fractions, the potential for strong drive of Alfven eigenmodes due to the low toroidal field, the prospect of velocity space instabilities transporting fast ions into loss cones, and new loss cones developing at high beta due to asymmetric trapping of ions.

Measurements needed to address these issues include: IR imaging of the walls, fast ion loss detectors, Mirnov coils, rf probes, spectroscopic or pellet-based measurements of the fast ion distribution, neutron fluctuations, and internal fluctuation diagnostics.

Fast ion measurements will be applicable to discharges with neutral beams and possibly also to rf heated discharges. High beta discharges are likely to be especially interesting.

 P. C. Efthimion

Simple accessibility arguments indentify two potential means of accessing EBW from the electron cyclotron layer. One scenario is for EBW to propagate to the plasma edge and to employ an electrostatic antenna to couple the radiation into a conventional microwave receiver. The waves may not be accessible from both sides of the plasma because of the upper hybrid resonance and the X-mode cutoff layers. EBW should propagate to the edge of the inner half of the plasma. An electrostatic antenna can be placed oblique to the horizontal plane of the machine either above or below the midplane. Another scenario is for the EBW to mode convert into an electromagnetic wave at the upper hybrid layer, to tunnel through the X-mode cutoff, and to propagate out the low field side [2].

[1] J. Hosea, V. Arunasalam, and R. Cano, PPPL Report, PPPL-1362, July 1977.

[2] K. C. Wu, A. K. Ram, and A. Bers, Proc. Am. Phys. Soc. 41, 1425 (1996).

 T.S. Bigelow, D.B. Batchelor, M.D. Carter

ORNL

A number of possible Electron Cyclotron current drive (ECCD) mechanisms are available. Ad hoc scaling of previous ECH pressure-driven current drive results indicate that a 400 kW ECH system may be capable of driving as high as 60 kA with this mechanism. There are two direct ECCD mechanisms that may be applicable to NSTX- the Doppler shifted Fisch-Boozer technique and mode conversion to electron Bernstein waves from X-mode launch [3,4]. NSTX offers an excellent opportunity to further test and optimize these ECCD modes provided an ECH system with the required flexibility to be optimized for a particular process. Additionally, the synergistic effects between ECH and other current drive systems such as helicity injection, RF and inductive transformer action should be fully investigated.

The availability of existing ECH hardware at ORNL and power supply hardware at PPPL makes this research effort particularly attractive due to the significant cost and schedule reductions possible. Sufficient hardware exists to install 400 kW of ECH power at 28 GHz (two 200 kW, 100 ms pulse gyrotrons). A preliminary design for the ECH hardware and several launcher options has been completed. These gyrotrons are also expected to be operable at ~15.3 GHz in the TE01 mode which adds additional experimental flexibility for both startup and heating functions. An 18 GHz, 15-30 kW ECH pre-ionization system is also under consideration for startup assistance in the initial operating phase. The requirements and capabilities of this system have been addressed in a preliminary design.

[1] NSTX PAC Review report, 1996

[2] C.B. Forest, et al., Investigation of the formation of a fully pressure-driven tokamak, Phys. Plasmas 1 (5) p1568-75 May 1994

[3] Preinhaelter & Kopecky, J. Plas. Phys, 10 p1 1973

[4] K.C. Wu, A.K Ram, A.Bers., Bull. APS 41 1425 (1996)

 A.K. Ram, A. Bers, K.C. Wu, and S.D. Schultz

Plasma Science and Fusion Center, M.I.T.

ABSTRACT:

 Heating and current drive by electron cyclotron waves in high magnetic field tokamaks is commonly done by exciting the O-mode from the low-field side. However, for NSTX, where for most of the plasma the electron plasma frequency is greater than the electron cyclotron frequency, the O-mode is cutoff near the edge of the plasma and does not propagate into the core. For NSTX-type plasmas we find that, if power is coupled to the X-mode at the edge of the plasma, there can be a subsequent mode conversion to the electron Bernstein wave (EBW) which then propagates towards the plasma core. Since the upper-hybrid layer is followed by a cutoff of the slow X-mode, a triplet mode-conversion scenario (cutoff-resonance-cutoff) exists which can, ideally, lead to 100% mode conversion [1] to EBWs. The EBW could damp on the electrons by Landau damping due to the upshift in the magnitude of the parallel (to the magnetic field) wavenumber and/or at the Doppler-shifted electron cyclotron (or its harmonic) resonance. In this talk we introduce the mode conversion scenario, suggest possible test scenarios on present tokamaks, and outline our planned/ongoing studies on the excitation of EBWs in NSTX.

[1] A.K. Ram, A. Bers, S.D. Schultz, and V. Fuchs,

Phy. Plasmas, vol. 3, pg. 1976 (1996).

Dan Jassby

Scientific issues being addressed

A key test of the ability of plasma simulation codes such as TRANSP to describe plasma performance depends on their prediction of fusion neutron rate, as the latter is sensitive to all plasma parameters including fast ion loss, density and temperature profiles, RF-induced tails, beam deposition profiles, and fast ion loss. Confidence in a simulation code is tentative unless neutron rates can be predicted to within 10-20%.

Measurement of neutron rate is also important for detecting small changes in ion temperature or production of ion tails when RF is applied, where other diagnostics are insensitive to changes (eg. recent IBW work on TFTR).

--------------------------------

Relevance to ST fusion and plasma science

The very low aspect ratio of ST results in a large fraction of trapped particles, including beam-injected ions, RF-induced tails, and the tail of the thermal-ion distribution. Many of these ions will be on loss orbits and thus result in reduced heating, current sustainment, and fusion rates. Neutron measurements will reflect these phenomena. To get agreement with the neutron rates requires accurate calculation of beam-ion deposition, tail formation, slowing-down distribution, and loss processes including ripple loss.q

--------------------------------

Measurements needed

The time dependence of the neutron rate must be monitored. This task can be accomplished with the existing TFTR fission-chamber system (called NE for "Neutron Epithermal"). These detectors can be installed outside NSTX with no impact on NSTX construction or operation. The detectors must be calibrated absolutely by moving a Cf-252 source within the NSTX vacuum vessel.

--------------------------------

Plasma parameters and operating conditions required

The NE system on TFTR can measure neutron rates down to 1xE11 n/s. When installed at NSTX, the detectors will be closer to the plasma and the neutron flux per source neutron much larger, so that rates down to 3xE10 n/s or even lower can be measured accurately. The detection system will be effective when any RF or neutral beam at all is applied, and during high-density ohmic discharges as well.