ARIES Project Meeting Minutes
8-10 January 2003
University of California, San Diego
Documented by L. Waganer
Welcome - Farrokh Najmabadi welcomed the team to UCSD and informed them of the University facilities. He explained that the meeting was being transmitted over the web and also through a phone hookup with DOE and PPPL members participating remotely.
Status of ARIES Program - Farrokh Najmabadi told the group it is difficult to have the group working on two different concepts, each with differing groundrules and objectives. Almost all of the IFE papers documenting the past two years’ efforts have been received. These papers will be published in a special issue of Fusion Science and Technology.
Farrokh noted that the US/Japan Workshop would be held in Tokyo, March 24-26. He would like to see suggested topics for presentation. There will be a limited number of people to attend from ARIES, so the topics should be examined to present the most valuable data for the workshop.
Farrokh mentioned the Compact Stellarator (CS) research plans would commence in FY03 and run for three years. The first year will improve the modeling capability and investigate the configuration options. The second year will refine the configuration and parameter space to select the most appropriate baseline approach. The final year will focus on a more detailed and optimized design assessment for the commercial power plant.
Next Meeting/Conference Call - The next project meeting was selected to be held in early May in conjunction with an ARIES Town Meeting (TM), focused on the analyses and results of IFE liquid wall protection approaches. Attendance in this Town Meeting would encompass ARIES and IFE communities to discuss and exchange information. This co-joined meeting would best be held on the West Coast - Wayne Meier offered to host it off-site at LLNL and make the TM arrangements/invitations. Tentatively, the Town Meeting would be around 1-½ days, the IFE around a day and the CS around a day. Les Waganer will send out an e-mail firming the dates for the week of May 5-8.
To establish a longer-range calendar of events, the following meeting was planned for September 3-4-5 in Atlanta at the Georgia Tech site, or off-site, at a new conference center.
The next project conference calls were selected to be Tuesday, February 11 (IFE team only), Tuesday, February 25 (CS E-meeting conference call), and Thursday, March 20 (IFE only). Separation of the IFE and CS team members on the call will help keep the calls to a manageable length and retain focus. The CS team also needs more detailed results and the E-meeting should help. Please send electronic presentations to Farrokh ahead of the call date of 2/25 so he can install the files on the ARIES server. Les Waganer will notify the team of the conference call numbers.
General MFE Discussions
Summary of Lehman Committee Findings on ITER Cost Assessment - Les Waganer reported on the findings of the Lehman Committee that was commissioned by Dr. Ray Orbach, Director of the Office of Science. The objective of the committee was to assess the validity and reasonableness of the ITER cost estimate for construction of ITER. Les summarized the Snowmass Summer Study that, in part, also addressed the construction cost of ITER (and FIRE). The Snowmass Cost Assessment results were that the ITER cost estimate was detailed and reasonable as a basis to negotiate the allocation of resources and the participation levels of the Parties.
The Committee met in late November 2002 with Dr. Aymar and the ITER team to understand the procedure for the ITER cost estimate and the resultant cost basis. Again the result was that the ITER estimate was very detailed and reasonable as a basis for negotiating the work packages. If the US decides to rejoin ITER, the committee recommended that for those hardware elements the US might potentially supply, cost estimates should be reestimated by vendors to include sufficient contingency, escalation, and management support. Conversion of the total ITER estimate “values” into a more-traditional US estimate containing appropriate contingencies and escalation factors may be inappropriate as the Parties will finance elements of ITER in their own currency. A total cost of ITER would never be quantified or reported in any single currency.
ARIES-IFE Study Presentations
Systems Analysis and Integration
Status of Heavy Ion Driver Point Design - Wayne Meier reminded the group that the HI driver was paired with the thick liquid wall chamber approach and the indirect drive target. The driver arrangement consists of 60 beams on each side of the chamber arranged with a maximum ½ angle of 24°. The main beams are arranged on the outside of the array configuration, the prepulse beams are closer to the center and the center is vacant to allow for target injection or beam-formation lasers. Individual beams are assembled to form the proper time- and energy-phasing for target heating and compression. The driver ion has been raised from Xenon to Bismuth to make the beams stiffer and easier to focus. The driver efficiency is now around 38%. With driver energy of 7 GJ, the target gain is 57 and the yield is 400 MJ.
The repetition rate and driver energy are at, or near, the optimal set of parameters. A dipole magnet is used to prevent the streaming of debris and neutralizing plasma into the final focus magnet region. To help lower the production cost, all magnets are identical in construction. In some magnet locations, the magnets may have an excess capacity. There will be a combination of NbTi and Nb3Sn superconducting materials. The beams must be neutralized just before entering the chamber to focus on the target at the required spot size.
Wayne explained the thick wall approach of protecting the outer chamber structure and the beamlines. The molten metal salt vortex downstream of the final magnet was shown. There was a question as to how the liquid is extracted at the ends and how that structure would be protected. Overall, the shielding results are quite promising with lifetimes exceeding the anticipated plant lifetime.
Chamber and Final Optics Nuclear Analysis
Shielding of the Final Focus System in the HI Point Design - Wayne Meier presented Jeff Latkowski’s work on shielding of the final focus magnet system. His work presently is on the 30°, ½ angle system, and the results are generally applicable. He found the magnets would not quench on a shot by shot basis. The shielding is adequate to keep cooling acceptable and have a good lifetime. Achieving a Class C disposal rating for the magnet was not possible. To solve the waste problem, Laila El-Guebaly suggested averaging the waste disposal rating over the entire assembly, not just the magnet, as disposing the entire assembly as a single unit seems likely. The array configuration (CAD and neutronics models) was shown. Some additional improvements were noted to help the design configuration and reduce the waste requirements.
Target Fabrication, Injection, and Tracking
Indirect Drive Target Materials Selection and Costing Studies - Ron Petzoldt reported that GA has conducted an in-depth cost assessment of the indirect drive target, similar to that previously done for the direct drive target. With the parameters assumed, the production costs for the targets are around 1 cent/DT capsule and 40 cents for the indirect-drive hohlraum assembly. Dealing with recycled, radioactive hohlraum materials could increase the target cost significantly (~ factor of 10). The group questioned the seemingly high number of personnel for the plant. A preliminary layout for the target facility and candidate production processes was presented.
He explained the two fabrication processes for layering the hohlraum were considered: warm assembly and cryogenic assembly. The warm assembly requires a very large tritium inventory and difficult, in-hohlraum layering. Thus, it is recommended the cryogenic assembly process be adopted. The process is described as just-in-time production. A critical step is target assembly in the last few seconds before injection. If there is a production stoppage at that step, the reactor must be shut down within seconds. This should be examined in more detail for a mitigation plan.
Ron described one process that can produce very low density, high-Z materials. The new process is laser-assisted, chemical vapor deposition to in-place create or fabricate the low-density foam within the hohlraum. This process is very capital intensive, but it is an enabling process to rapidly fabricate the low-density foam in-place.
It was suggested that the costs could be further reduced if several fusion power production facilities would share a single target fabrication plant for the non-cryogenic components. Some people felt the target production facility should be self-contained and dedicated to a single plant to offer the advantage of self-sufficient energy production.
Chamber/Beam Physics, Beam Transport, and Chamber Clearing
Integrated TSUNAMI Simulations for the Heavy-Ion Point Design - Simon Yu presented Christophe Debonnel’s charts on the requirements for stringent chamber density control. Neutralized ballistic transport establishes strict limits for the chamber background gas density. Equally important is the gas and particle density deposition in the beam tubes as arcing may occur between the wall and the beam, upstream of the neutralization region. An efficient target chamber should vent debris toward the condensing droplets to keep beam ports as free from debris as possible. Gas dynamics modeling is also important for predicting the liquid pocket response and disruption. The new geometry and target selected for the Heavy-Ion Robust Point Design of 2002 motivated a new set of TSUNAMI simulations.
The TSUNAMI code estimates the gas dynamics behavior during the initial venting process, solving Euler equations for compressible flows using a real gas equation. These “integrated” simulations include the liquid pocket, the venting areas toward the ports and the condensing droplets, and a fictitious centerline tube. (Previously target chamber and beam tube simulations were conducted separately, hence the “integrated.”) Density contour plots and impulse loadings were shown for the cavity and liquid structures.
Within the tube, liquid vortices offer condensing surfaces for target chamber debris, and magnetic shutters located just before the last final focus magnet divert and trap the remaining energetic ionized particles that have not condensed onto the vortices. The dipole plays the dual role of preventing debris ingression and neutralizing electron streaming into the final focus magnet region. The vortex and the magnetic shutters were analyzed to estimate their effectiveness.
Status of Operational Windows for HIF Chamber Transport Modes - Dale Welch presented David Rose’s presentation. Dale reminded the group the objective of this effort was to develop ion beam transport operational parametric windows to guide chamber design, e.g., radius, gas pressure, shielding and structural materials, and protection approaches. Constraints and design requirements arise from driver beams, targets, chamber, and final focus systems. Early beam transport theory dealt with beam stability. Stability constraints are not being examined by analytical modeling and detailed numerical simulations with aforementioned constraints. The end product maps acceptable chamber initial conditions.
Neutral beam transport places more severe chamber requirements as opposed to channel transport approaches as shown in the accompanying figure.
For the neutral ballistic transport (NBT), the electrostatic stability is being assesses with a detailed comparison between analytic models and numerical simulations. The electromagnetic stability will be assessed in a similar manner, but with large-scale 3-D simulations. The figure to the left is an assessment of neutral beam transport as constrained by first wall radius. This calculation used the radial space charge spreading current and beam stripping as a function of the distance and gas pressure. The neutralization fraction was assumed constant and stability is always stable. The assumptions are being verified as well as further window analyses.
The pinch mode design windows will also be mapped. The Assisted Pinch Transport will be explored with numerical simulations over a broad operational parameter space, per D. Rose’s previous talks. The Self Pinch Transport will likely have a broad pressure window, but significant work remains to assess transport efficiency and stability.
Beamline Design Issues - Dale Welch summarized the beam line design issues below:
Dale stated high perveance HIF beams could be neutralized with localized plasma in the beamline after final focus and entry into the power core chamber. The problem is emittance growth arising from plasma plug electrons being attracted by the beam space charge. The simulation emphasis was upstream of the plasma plug. The simulations indicate the plasma electrons are pulled into unneutralized beam as beam approaches plasma plug and stream back non-uniformly and oscillate. The upstream motion of electrons may be suppressed with solenoidal or dipole magnetic fields. A 2-kG solenoidal field removes most of the emittance growth, but the field needs to be 50-100 cm long. A dipole field of >2 kG and 5-cm long suppresses backstreaming in 3-D Lsp simulations. It is felt a dipole magnetic field just ahead of the plasma plug can suppress the electron motions and reduce beam emittance growth.
Review of the Mechanical Design of the Final Focusing Region of The HIF Point Design - Tom Brown reviewed the design requirements and main magnet parameters from the 2002 Robust Point Design PPPL has been converting into a final focus mechanical design, as shown below.
At the lower edge of the picture, Tom Brown can be seen admiring four magnet assemblies when disassembled. This picture illustrates the new modular design approach that handles all magnets within a section as a single assembly (four assemblies). The prior design had all four magnets for each beamline as a single assembly, each of which was loaded into a large structure that held and aligned the beamlines. The key to the present design is to maintain relative beamline alignment during the cooldown process. Maintainability issues need to be more fully understood from the point of view of component activation and personnel access.
Comparison of Final Focus Magnetic Systems for the Assisted Pinched Transport and the Robust Point Design 2002 - Simon Yu reviewed the principles of the Assisted Pinch Transport and stated the HI beam and hybrid target parameters for the Robust Point Design 2002. He compared the ATP with the RPD and the RPD had many favorable design features, such as:
Simon showed how a racetrack coil design approach can produce the desired field configuration and can be easier to fabricate. Manufactured with Nb3Sn superconductors, the coil can produce the proper field strengths.
Chamber Engineering, Thick Liquid Wall Approach
Summary of Assessment of the IFE Chamber Structural Materials - René Raffray presented the Mike Billone summary of the IFE materials assessment. The reference IFE reactor for the assessment was the HYLIFE concept. The complete report can be found at the ARIES Web Site. It was found that the structural material, 304SS, has major swelling, activation, and He embrittlement issues. As such it is not a recommended structural material for a fusion reactor.
Other materials were investigated. If a 300 series SS is considered as a near-term base line for the design, it is recommended that Ti-modified 316SS (PCA) be considered instead of 304SS for the first wall tubes, connecting bars, and rings connecting to the back wall. Austenitic steel is not considered for future reactors and is not included in the MFE materials program R&D efforts. Oxide-dispersion strengthened ferritic steel (ODS FS) and SiCf/SiC provide the possibility of higher temperature operation and much better power plant performance, thus they are potentially attractive candidates.
Relevant One-Dimensional Simulations of Prompt Physics in Thick (Liquid) Wall Design - Don Haynes said he is using the BUCKY code to determine the prompt physics in the thick liquid wall chamber concept. BUCKY analyzes the superheated liquid by conserving energy by redistributing the energy. Varying the plasma radius emphasized the importance of the plasma vapor. There is more vaporized material contained in the 9-m chamber than the 5-m radius chamber, although the vapor ratio is less than the ratio of surface areas.
BUCKY does treat explosive boiling, in a non-explosive sense. Don explained that energy is conserved, and after each time step, there is no super-heated liquid. However, it does not analyze the effect of exploding bubbles spewing liquid drops into the chamber. Don also noted that at a time of around 1 microsecond, the shock wave from the blowoff from the surface and the wave from the center meet and interact, at which time the simulation should be turned over to another higher order code.
Aerosol Production in Lead-protected and FLiBe-protected Chambers - René Raffray presented Phil Sharp’s results on aerosol production in IFE chambers. René explained some of the improvements to the code, namely the wall thermal response is now incorporated including penetrating x-rays. Surface vaporization and condensation are now solved consistently. FLiBe bulk properties are added to look at nucleation, however dissociation and condensation of component species have not been included. Aerosol properties (mass concentration and particle size distribution) in four regions of the 6.5-m chamber were determined. Liquid wall simulations were conducted for the 458-MJ ID target with only the 115 MJ of x-ray flux depositing prompt energy on the wall surface. The simulations below showed the velocities of both lead and FLiBe at times ranging from 100 microsec to 250 ms for all radial positions. Plots were also shown for the total vapor and aerosol masses, regional mass concentrations, and particle sizes. Results indicate that FLiBe gives a smaller mass concentration near the chamber center (10 mg/m3 versus 20 mg/m3 for lead). The distribution of the aerosol size is more limited.
Parameter Design Window for Thin Liquid Walls - René Raffray reported on his and M. Zaghloul’s work on characterizing the liquid-metal cooled 6.5-meter chamber. René showed a graph depicting photon energy deposition density profile with both film and explosive boiling for FLiBe. This graph shows the sensible energy with uniform vapor pressure following photon passage in chamber and including evaporated FLiBe from the liquid film. There are three regions shown: evaporation, explosive boiling, and two-phase region. From the results, René said there were appreciable number and size of aerosol particles present after 0.25 seconds (when the next target is injected). He said that the preliminary constraints for target tracking are based on 90% of beam tracking and HID based on stripping with an integrated line density of 1 mTorr for Neutralized Ballistic Transport (NBT). He concluded that aerosol formation could be a key issue and the driver and target constraints need to be more accurately defined.
Scoping Liquid Wall Mechanical Response to Thermal Shock - René Raffray explained the proposed 2003 ARIES effort to investigate the liquid wall ablation as an aerosol source term. One part of the effort would be the integrated effect of liquid wall thermal and mechanical response to x-ray energy deposition that would provide bounding estimates of the aerosol source term. The second part would be to determine liquid wall mechanical response to rapid x-rays energy deposition, specifically to determine the spall strength of materials as compared to anticipated IFE shocks, fracture or spall time scale, and droplets size and distribution, for a variety of wall materials (e.g., Pb, FLiBe and Li).
Design Windows for Thin and Thick Liquid Protection Schemes - Minami Yoda explained that most of Georgia Tech’s examination of liquid protection pertains to the “Wetted Wall” (porous surface and forced flow) and thick liquid wall concepts. For the thin film, can a stable liquid film be maintained over the entire surface, will it dry out, and can it be reestablished before the next cycle? One of the key questions is the time it takes for a drop to form on an inverted surface and when will the drop detach? GT has developed a set of general nondimensional charts that apply to a variety of candidate coolants and operating conditions. Stability of the thin liquid film imposes a lower bound on the cycle repetition rate from droplet formation and a lower bound on liquid injection velocity to maintain minimum film thicknesses over the cavity. A wetting surface increases the distance the film adheres to the surface and reduces the number of injection and extraction locations. Cylindrical dams around port openings have been found to be incompatible with effective forced film protection.
HYLIFE-II uses jets or sheets to form the moving pocket for the target and a lattice of stationary jets to protect the beam and target entry ports. Major questions relate to the ability to form smooth prototypical turbulent liquid sheets? Can they join and separate cleanly without creating aerosols and droplets? Can the flow be reestablished before the next shot? To answer these questions, GT established representative liquid sheet parameters and measured surface ripple and standard deviation with fluorescing water in air over a range of Reynolds numbers. They concluded typical HYLIFE-II jets would have an average surface ripple of < 5 mm, which is sufficient for magnet shielding requirements. Their measurements and analysis-generated data was useful for shielding calculations.
Irradiation effects on insulating materials for magnets for IFE and/or MFE - Leslie Bromberg explained superconducting insulators can be damaged by e-beams, gammas, and neutrons and is mainly due to power absorption per unit mass (rads and Grays). He then showed the radiation effects of various insulators due to neutron fluence and electron beams. He noted that gas evolution could determine the insulator degradation. He outlined an insulator irradiation program to determine the performance of candidate insulators in an irradiating environment.
Action Items from ARIES-IFE Meeting - René Raffray assembled a list of IFE action items from the above discussions and from among the teammates. They are summarized below:
ARIES-Compact Stellarator Study Presentations
Recent Progress in Configuration Development for Compact Stellarator Reactors - Long-Poe Ku reviewed the status presented at the October 2002 ARIES meeting regarding the prospects of the NCSX as a reactor. The fast ion confinement in NCSX is not as good as desired and the alpha losses could amount to > 25%, thus it might be necessary to use a configuration figure of merit search to improve the particle confinement characteristics. Other considerations include:
Long-Po proposed and discussed several figures of merit for quasi-axisymmetry and alpha confinement, namely:
He then said we should test the effectiveness of these five penalty (utility) functions, or figures of merit, by examining the configurations that we currently have. For each of FOM, plots were shown that evaluated the many configurations over a range of aspect ratios, iotas, and field periods. After examining the results, it was concluded that to achieve an alpha optimized QA configuration the most effective procedure would be to minimize the residual Bm, and n (field periods) to reach the QA regime and minimize alpha losses using a cutoff of small loss fraction and a limited number of toroidal transits.
Using the direct initial orbit optimization, a new class of three-field period configurations were developed. Comparison of properties indicate that there may be an optimal aspect ratio for CS reactors where alpha losses may be minimized while plasma volume is maximized (or reactor size is minimized for a given power.)
Power Plant Study of Compact Stellarators - Paul Garabedian pointed out that although the experimental LHD stellarator is linearly unstable, it remains nonlinearly stable at a beta of 3.2%. This shows promise for reactor regimes and opens up the range of candidate configurations. Linear stability analysis is insufficient to determine stability. He then showed several configurations to stress this point. With care, magnetic islands can be removed by adjusting coil configurations.
Exploring ECRF Heating on CS Reactors - TK Mau suggested it is more advisable to use RF heating rather than neutral particle beam for bulk heating and plasma initiation in compact stellerators. It can also provide plasma profile control and localized current drive to offset bootstrap current for improved physics performance. It has the advantage of a compact design with minimal-sized openings in the first wall.
Currently, power sources and hardened components are being evaluated. There is a significant database for Electron Cyclotron Heating and Current Drive (ECH/ECCD) for tokamaks, stellarators, and other magnetic confinement systems. TK defined a CS/QA example for analysis. He noted the Q-mode and the X-mode are strongly absorbed at the plasma saddle point. The X-mode and O-mode are also attractive operating conditions.
Superconducting Magnets for Stellarators - Leslie Bromberg explained the topology of current experimental stellarator magnet systems. He discussed the insulation and the superconducting material and manufacturing issues. He discussed with the group the idea of using modular stellarator coils as well as ancillary TF and PF coils. The group generally thought it is better to only use stellarator coils on the reactor configuration. The TF and PF coils are provided in experiments to enhance the experimental operating space. The added complexity for the reactor is not desired.
Leslie discussed the modular coil construction types: shell, radial, and hybrid. His analysis indicated the shell was preferred, especially when the high temperature superconductor is used.
Issues in the Stability Analysis for the ARIES Compact Stellarator Design - Alan Turnbull stated there are several important issues involving evaluation of equilibrium tools for CS design studies:
Alan then described several equilibrium codes that exist and discussed their advantages and disadvantages. VMEC is the industry standard inverse fixed boundary equilibrium code. There is a choice of direct free boundary equilibrium codes available. Also a suite of linear global MHD stability codes is available for stellarators. TERPSICHORE and CAS-3D are state of the art free boundary, fully 3-D linear ideal MHD stability codes and are essentially equivalent. Spector-3D is the only existing linear global resistive MHD stability code. It is not clear localized MHD stability is relevant for stellarators. 3-D global extended MHD stability code tools are available, such as M3D and NIMROD.
Alan then summarized his thought on CS equilibrium and stability and their respective available codes.
Plans for Engineering Study of ARIES-CS - Rene Raffray said that the assessment of the CS as an option for a commercial power plant will help advance the physics and technology, address the attractiveness issues in the context of a power plant study, and identify the optimum CS configuration for the power plant.
Rene envisions the ARIES-CS as a three-year study with the first year to develop the plasma and coil configuration optimization tools. The second year would be devoted to exploring the configuration design space. The third year would be devoted to detailed system design and optimization within the design space identified. Rene then identified the hardware systems that would be highlighted in each year with the major activities within each year’s activities. Maintenance schemes are essential to be able to obtain the optimal system that is most attractive. Key systems were the first wall/blanket, shielding, coils, divertor and maintenance. Les Waganer and Mark Tillack pointed out that the vacuum vessel should also be equally important and carried as an integral system as all the blankets, shields, and divertors are structurally mounted to the vacuum vessel. The distance from the plasma to the midpoint of the coils is an important parameter that helps define the coils and the plasma performance.
Maintenance Concept for Modular Blankets in the Compact Stellarator Power Plants - Siegfried Malang discussed the key maintenance issues (goals) in previous stellarator studies:
For stellarators, the main issues for maintenance with an articulated boom are:
For the possible blanket types, Siegfried suggested:
As a starting point, Siegried suggested a FLiBe/FS blanket be used and maintained with an articulated boom removing blanket modules from the inside the blanket.
Minimum Radial Standoff Problem Definition and Needed Information - Laila El-Guebaly summarized the initial parameters for nuclear assessment. Laila illustrated the key elements inside the core: first wall, blanket, shield, vacuum vessel, magnet cryostat, and the winding pack. Laila outlined the radiological design requirements, such as TBR, dpa, helium production in the VV, and magnetic damage. She listed a host of potential breeding materials and suggested FLiBe with a Be multiplier results in a thin blanket (30 cm). A FLiBe blanket has a high shielding performance that helps reduce the radial build thickness. A low cost, steel-based shield could be 45-cm thick to help ensure re-weldability of the water-cooled vacuum vessel at any time. Additionally, certain FW areas could be shield-only zones that could offer a reduced thickness radial build.
Discussion of Compact Stellarator Reactor Requirements - Les Waganer reviewed a preliminary list of reactor design requirements that would help highlight those items that might be beneficial to examine to help direct and optimize the CS plant configuration. He then went over the list to elicit comments from the ARIES team. This list will serve as the basis for the CS design data book.