ARIES Project Meeting Minutes
1-2 July 2002
General Atomics, San Diego
Documented by L. Waganer
Welcome - Ron Petzoldt welcomed the ARIES team to GA and San Diego and pointed out the meeting arrangements at GA.
Program Status - Farrokh Najmabadi noted that the program is at a transition point changing from the thin liquid film wall to the thick wall protection concept. This will be the meeting where we establish our groundrules for comparison and the methodology for the thick-wall assessment. We intend to complete the thick-wall comparison by the end of this fiscal year. The project also needs to complete the documentation on the technical results, primarily the thin-wall concept. Also the thick-wall technical results will be due shortly.
Next year, the project will conduct studies on both IFE and MFE. On IFE, work will continue on the critical areas of wall protection schemes and the HI driver systems. On the MFE side, work will commence on a 3-year compact stellarator study. The first year will be devoted to optimizing physics and engineering tools. The second year effort will be to explore the design parameter space to determine optimal regions. The third year effort will concentrate on developing and defining a point design.
The budget has not been defined yet for next year. Accomplishment of the planned tasks will require some additional budget, which is being addressed by program management and the DOE.
Next Meeting/Conference Call – The next project meeting, tentatively at PPPL in September for about 2 ˝ days, will be devoted to both IFE (~1 day) and MFE (~1 ˝ day for briefing on compact stellarator work). [Subsequently, the meeting date was established as October 2-4 at PPPL.]
The next project conference call was selected to be 20 August. Les Waganer will notify the team of the conference call number.
ARIES-IFE Study Presentations
Systems Analysis and Integration
ARIES-IFE Systems Issues - Ron Miller noted that he has been incorporating the direct drive target fabrication costs supplied by General Atomics. He is continuing to improve the modeling of the gain curves for both the KrF and HI targets. He is using the iDecide influence code to model the system uncertainties.
There was a question raised during the last meeting that the pumping power for the thick wall concept is thought to scale with repetition rate and Ron is investigating the effect. Ralph Moir and other HYLIFE personnel have researched and documented a scaling law for this effect. Optimizing the COE would tend to minimize the pumping power as pumping power has a negative influence on the COE. The pumps considered are around 80% efficient.
Ron has run the iDecide model with both 45% and 55% gross thermal efficiencies. These runs optimized to repetition rates of 5 and 5.8 Hz, respectively. It was mentioned HYLIFE estimates a gross thermal efficiency of 43% with its design baseline. The temperature of the FLiBe coolant is limited by the vapor pressure in the HYLIFE design. Ron is using recirculating power fractions of 10 to 14% and will try to determine if this value is correct over the ranges of repetition rates being considered.
Updated Heavy Ion Driver Parameters for Snowmass Point Design - Wayne Meier said he was working to update the HI driver parameters for the July 2002 Snowmass summer study. He described the key HI parameters being used to characterize the driver and the interface requirements with the target chamber. There are 112 beams (56/side) within a maximum half angle of 24° (to centerline of outermost beam). Each beam has a half angle of 15 mradians to achieve the minimum spot size at the target (2.3 mm for foot beam and 2.0 mm for main beam). The beams from the accelerator are mapped on to a square grid and focused on the target. The first level protection for the first wall and beam final focus magnets is an orthogonal set of cylindrical liquid metal jets arranged to allow a clear path for the beams, yet block most of the neutrons and x-rays and divert some of the debris. Subsequent protections schemes were discussed by other presenters.
Wayne showed that the driver energy of 6.4 MJ and the repetition rate of 5.9 Hz are near the optimum COE operating points. Moreover, the COE is insensitive to moderate changes in these energy and repetition parameters in this region of parameter space, which allows design flexibility.
The HIF VNL is defining a self-consistent point design for a power plant. The thick liquid wall is the basis for the chamber protection with evaluations on activation, radiation damage, and thermal heating of final focus beam arrays. The beam final focus design configuration meets the target spot size. R&D continues to examine means to reduce driver cost as well as other factors to reduce the overall COE.
Chamber and Final Optics Nuclear Analysis
Updated Shielding Analysis for Final Focus Magnets - One of the enabling engineering features is the ability to shield the final focus magnets. The large number of beams compressed into two tight bundles, aggravate the difficulty of shielding the beamlines and the final focus magnets. Shielding within the beamline has to be very thin (~5 cm) to accommodate the magnets. The total dose to the magnets is limited to 100 MGy, which is dominated by gamma ray radiation. This suggests the use of inorganic insulators with high temperature superconductors (HTS). A conservative limit for neutron fluence for HTS may be double that of Nb3Sn (1019 n/cm2). The large number (112) of beams has contributed to the shielding difficulty by more than an order of magnitude as compared to prior studies.
Shielding analysis efforts during the past four years have increased the magnet lifetime from around 1 year to greater than 30 FPY for a self-consistent design. There is strong peaking of neutron flux at the center of the beam array due to cross talk between beams.
Because the beams arranged so compactly, the beam-to-structure clearance is very important. This clearance has recently been reduced from 5 mm to 1 mm. The beam-to-FLiBe cross jets distance is held constant at 5 mm. Liquid vortex shielding is being considered as the inner bore shielding approach. Three shielding options were evaluated that achieved lifetimes somewhat less than life of plant for the insulator and superconductor. Additional work is underway to further optimize the magnet lifetime, beam neutralization, HTS conductors, waste disposal, and recirculating power.
Update on Self-Pinched Transport Simulations - David Rose has been improving the modeling and the self-pinch transport efficiency. This work includes multiple plasma ion species, beam temperature profile and examination of the super pinch mode. Equilibrium results studies included electron and ion evolution in the presence of a rigid beam (high charge state). He showed plasma density profiles in the presence of a beam and observed that the radial ion velocities were consistent with channel hydro limits. Contours of rB q show some diffusion of the return current on short time scales. Finite-mass beam ions (Pb+65) and radial temperature profiles suggest reduced evaporation losses.
3-D Simulations of a Diversion in a Beam Port with a Vertical Magnetic Field - David Rose reviewed prior results that a plasma can be blown off the chamber wall and expand into the beam port. Dale Welch’s 2-D simulation results showed confinement of drifting, low-b plasmas. David reviewed a sample long wavelength, low-b plasma penetration model from the literature (see his presentation for the reference). His 3-D simulation used a drift velocity of 9 cm/µs (90 km/s). His results indicated both the 2-D and 3-D calculations predicted the plasmas were diverted by the moderate strength field (1 kG). The 3-D results indicted deeper particle penetration. Therefore, the magnetic field region should be extended further into the beam port. More modeling would help refine the recommended design approach.
Coupling of Assisted Pinch Transport (APT) Ion Beam to Hybrid Target - Dale Welch described how a laser-channel is formed ahead of the beam to assist the beam propagation to the target. His IPROP calculations show that 43 kA of beam current is required to contain a beam at ˝ cm radius. With a 50 kA discharge, he estimates an 87% energy transport efficiency.
Calculations with the hybrid target were completed and the simulations showed two distinct beams with the halo beam being the weaker. The halo is formed by the interaction of the rippled self-fields and the beam. The self-fields tend to degrade the transport efficiency. Dale expects a beam efficiency of 85% within 5 mm for a nominal beam and discharge, whereas an ideal case would suggest and efficiency of 94%. The efficiency falls for currents below 50 kA and more slowly with beam divergence (<150 mradian). On the other hand, transport efficiency is insensitive to the choice of beam ion.
Operating Window for Channel-Assisted Pinch Propagation - Stephen Neff explained there are limits on the plasma operating window for chamber pressure, beam length, and density reduction because of the beam energy loss in the channel, channel clearance by j x B force, and to prevent breakdown to the walls. Simulation results were shown for 5 torr xenon with and without thermal conduction applied. An analytical channel model was defined that could vary the plasma conductivity. Stephen concluded that his simulations indicated a good operating window, a channel clearance simulation is needed, and the gas rarefaction is a critical factor.
Fluid Dynamics in a HYLIFE Heavy-Ion Beam Line - Christophe Debonnel noted the HYLIFE design evolution toward a larger number and more tightly packed beamlines for the HI design basis. These factors have complicated the design with thinner and less shielding that results in higher fluence to the final focus magnets. Strategies are being evaluated to prevent or minimize target debris deposition in the beamline tubes. Within the chamber, multiple crossed liquid jets would help shield the beams. Liquid FLiNaBe surfaces inside the beam tubes (vortices) are being evaluated to determine how they would be formed, how they would function and their effectiveness. Christophe explained the formation of the stable vortex fluid flow. Experiments indicate a stable film thickness with an error of 2% of the pipe radius and a standard deviation of 1%. To help reduce debris ingress, transverse magnetic fields would help divert the ionized debris.
Driver Requirements for FLiBe Gas and Liquid Parameters - S. Yu and C. Olson were not able to attend and make their presentation.
Liquid Chamber Analyses Results (Principally Thick Liquid Walls)
Numerical and Experimental Studies of Thin-Liquid Film Wall Protection Schemes - Said Abdel-Khalik summarized the numerical simulation results of porous wetted walls he presented at the prior ARIES meeting at UW, Madison. Since that meeting, the GT group has added the effects of evaporation and condensation at the interface to the simulation. Additionally, they assessed the minimum film thickness prior to drop detachment to determine if the surface would be exposed (zero film thickness). With the evaporation and condensation effects included, Said presented the results for the postulated initial film surface conditions (sine, random, and saddle). The random surface is probably more likely, but the sine and saddle conditions were more severe for droplet formation. Combinations of film thicknesses and injection velocities could achieve conditions where the surface film thickness would go to zero during a droplet formation. However, the nondimensional solutions for the considered fluids would define suitable sets of system parameters to eliminate that occurrence.
The effects of evaporation and condensation were compared to a no net mass change (no evaporation or condensation) case for several sets of fluid conditions. Evaporation and condensation only slightly changed the time to form the droplet, but the ejected mass and film thicknesses had measurable changes.
A second area of investigation was the determination of a parametric design window for the performance of a high-velocity, inverted thin film from a flat wall plate. Dimensionless groups were formed to analyze the independent and dependent variables. It was observed the mean detachment distance is quite stable for the input conditions, but the distance will vary considerably (1%) at high frequency. The reported data is the average of a large number of observations.
The Froude number is the most important parameter for the detachment distance. The film width will initially widen (as much as 4-5 times initial width) and then narrow as the film thickens due to gravity and surface tension effects before detachment. The inclination angle is the most important to the film thickness in the far field.
Scoping Study for FLiBe Evaporation and Condensation - Since the documentation of FLiBe evaporation and condensation properties are vitally important to the validity of many aspects of the IFE chamber study, René Raffray is characterizing these important properties of this coolant. There is some difference of opinion (experimental and theoretical) regarding the vapor pressure data of FLiBe. René developed an analytical fit of the two sets of data. He also compiled a set of physical properties for Pb, FLiBe and Li. Based on these data, he determined the condensation flux and characteristic time as a function of FLiBe vapor and film conditions. The condensation time is sufficiently short to allow repetition rates up to 10 Hz, however the aerosol generation may be a bigger problem. The condition of high-energy deposition into a thin layer over a very short time may lead to the formation of aerosols.
René analyzed the vaporization from the free surface and concluded the vaporization is very high. Much higher heating rates (photon-like) free surface vaporization will not have time to occur. On the other hand, vaporization within the fluid film as heterogeneous nuclei is dependent on number of nuclei per unit mass but it very low for the heating rate. Phase explosion (explosive boiling) involves homogenous nucleation and high heating rates in the metastable liquid.
René combined these properties into a single chart showing the evaporated layer, the explosive boiling layer, and the 2-phase region. A table summarizes the three candidate coolant fluids: Pb, FLiBe, and Li vapor. The explosive boiling results obtained were compared to the ABLATOR results. A preliminary aerosol-operating window was presented showing a range of concentrations and particle sizes.
Thermodynamics of FLiBe System and Structural Material Recommendations - Dai-Kai Sze discussed the background on the existing vapor pressure documentation as provided by experimental data from ORNL and theoretical calculations from Olander and co-workers at UCB. (Raffray earlier showed these data in his presentation.) Comparing the two approaches, Dai-Kai would have greater confidence in the experimental data.
Dai Kai also reviewed the available choices of structural materials for the IFE chambers along with their positive and negative attributes. They are listed below:
HYLIFE-II has recently increased their shielding effectiveness and now feels 304 SS should be classified as a life of plant structural material. It is probably the lowest cost material.
Target Fabrication, Injection, and Tracking
Update on Target Fabrication Tasks - Dan Goodin provided some insight into the details that went into his calculated cost that a commercial NRL radiation preheat direct drive target would cost on the order of 16.6 cents/target. Both the capital cost of $97M for the target plant and the operating costs were annualized for the production of 0.5M usable targets per day. The costing was assumed to be consistent with the “n-th of a kind” facilities.
Fabrication of the indirect targets are a much more challenging problem. Dan was requested to estimate the cost of a commercial indirect drive target with the same groundrules as the direct drive target. The baseline target is the Tabak and Callahan target consisting of a hohlraum of many unique materials in complex configurations surrounding the cryogenic fuel shell. Processes for filling and cooling the inner capsule are being investigated. More detailed hohlraum requirements, cheaper materials and easier fabrication processes are being evaluated. Preliminary costs for manufacturing the shell, then filling and layering it have been estimated at about 11 cents each. This is very encouraging in that it leaves significant margin for the manufacture of the unique hohlraum components and assembly of the hohlraum with the shells. Evaluation of the cost of manufacturing the hohlraum itself is still underway.
Fabrication of the GA target injector is underway with 85% of this years equipment has been ordered. Preliminary tracking system optical test has been completed. The first detector housing is complete and being readied for tests. The software design specification and test plan are complete.
Previous data had indicated an unprotected radiation preheat target would not survive with high chamber pressure gas. Further analysis indicated that the previous data was optimistic for a bare target: gold surface reflectivity was too high, the average convective heat flux did not include condensation, and chamber gas temperature may be much higher than chamber wall (true background). Thus new target protection schemes were proposed and some preliminary analysis performed. A circular wake shield was analyzed to determine key parameters and range of expected performance improvement. One approach might be a heat barrier membrane coated with frozen gas and supported with a thin frame surrounding the direct drive target. Also a fast ignition cone might reduce the maximum heat flux.
Configuration Development for HIF Final Focus Superconducting Quadrupole Array - Phil Heitzenroeder defined the interface issues for the HIF final focus magnet arrays in the following areas:
PPPL has developed a conceptual design concept for the beamline. Critical details are being refined to determine the most difficult challenges. The large number of beams compressed into a small cone angle severely restricts the shielding and access. Rapid removal and repeatability of alignment are critical elements. Not all issues have been resolved yet.