ARIES Program
Public Information Site
 
 
 
 

 

ARIES-IFE and -CS Project Meeting Minutes

3- 4 September 2003

Georgia Technology

Atlanta, Georgia

Documented by L. Waganer


Attendees:
Organization ARIES Compact Stellarator ARIES IFE
ANL - -
BNL - -
Boeing Waganer Waganer
DOE - -
GA Turbull Goodin, Petzoldt
GT Abdel-Khalik, Durbin, Sadowsik, Yoda Abdel-Khalik, Durbin, Yoda
INL(was INEEL) Merrill  
LANL   -
LBNL   Logan, Yu
LLNL   Abbott, Meier
MIT Bromberg  
MRC   Rose
NRL - -
NYU Garabedian  
ORNL Lyon  
PPPL Ku  
RPI Steiner  
SNL   Olson
TSI - -
UCB   Debonnel
UCSD Malang, Mau, Najmabadi, Raffray, Sze Malang, Mau, Najmabadi, Raffray, Sze
U of Mont - -
UW-Mad El-Guebaly El-Guebaly

Ref: Agenda and Presentation Links: Project Meeting

Administrative

Welcome - Said Abdel-Khalik welcomed the ARIES Team (both days) to the Georgia Technology’s new Conference Center and Hotel. He and Georgia Tech provided an excellent meeting room with ample snacks and refreshments to the team. He also hosted a tour through the Georgia Tech Mechanical Engineering Laboratory facilities to observe the IFE-related hydrodynamic experiments he and his students are conducting.

Update of the ARIES Design Study Budget - Farrokh Najmabadi noted that the IFE work is to be completed by September 2003. This will be the last meeting of the IFE group. The arrangements for the IFE study documentation will be discussed in the IFE session following.

The funding for next year remains to be finalized. Farrokh is working with DOE Fusion Sciences to increase the tentative Presidential budget allotment, but the amount still remains less than the current budget. The present year evaluated the existing CS physics and engineering basis, the simulation code capability, and possible parameter space for CS reactor configurations. Next year, the team will evaluate more self-consistent and detailed strawman configurations.

Arrangements for Meetings and Conference Calls - Princeton Plasma Physics Laboratory will host the next ARIES-CS meeting on December 3-5 (full days for the 3rd and 4th, with tentatively a half day on Friday 12/5). The intent is to invite Stellarator physics and engineering experts to interact with the ARIES team. The scope and goals of the ARIES investigation along with a presentation of our current assessment of the two strawman concepts. We would also hope the invited experts would enlighten the team to the most current thinking regarding stellarator confinement, equilibrium, and research direction.

To reach a defendable configuration and assessment for both the new strawmen configurations, the ARIES team has to have close and timely coordination. To that end, we would like to hold monthly conference calls. Les Waganer will provide the phone-in numbers, dates, and times separately.

ARIES CS Assessment

Configuration Development and Optimization

NonLinear Stability and Coils for (Compact Stellarator) Fusion Reactors – Paul Garabedian discussed the application of the NSTAB code to show the LHD stellarator is linearly unstable, however it remains nonlinearly stable at levels of beta below those experimentally achieved. Ballooning theory predictions are more pessimistic than experimental results. This is also observed on the W7-AS experiment. Paul is trying to understand these findings and examine plasma and coil configurations that would lead to a simpler and easier way to build and operate a Compact Stellarator reactor. He showed the ballooning criteria as applied to LHD with b values exceeding experimental values. He examined the flux surfaces for a bi-furcated LHD equilibrium at b = 0.025 and R = 3.6. For triangular pressure profiles, the flux surfaces indicated solutions that are linearly unstable, but nonlinearly stable. Poincaré maps of the flux surface indicate that at b = 0.04, the solution is nonlinearly unstable, but stable at b = 0.032. A closer examination suggests b = 0.035 might be the predicted ballooning stability limit. The experiment has found a stable solution exists at b = 0.033. Similar solutions were shown for the MHH2 experiment at beta = 0.04. For the wall stabilized PG3 (simulated LI383) configuration, when operated as a hybrid, yields a b of 0.05 with the rotational transform of 0.55 to 0.80.

The remainder of Paul’s talk involved coil configurations and maintenance access between coils. An asymmetric view of a configuration with an aspect ratio of 4.5 showed 6 coils, comprising one field period of three that was adjusted for ample access. However this produced significant field ripple. The MHH2 coil set was also examined with an aspect ratio of 3.5 and two-field periods. This coil set was closely packed at the inner bore. Some reshaping and rotation might be possible to improve space between coils. The outer portions of these coils are relatively vertical with ample space between coils.

Paul did not have engineering data on this coil set, but Long-Poe Ku offered to work with Paul to generate the required engineering data for the systems code analysis to provide a second strawman coil set and plasma configuration.

Recent Progress in Configuration Development for Compact Stellarator Reactors – Long Poe Ku has examined the coil aspect ratio (specifically the Bmax and coil complexity), alternative coil concepts, and assessment of alpha heating load on the first wall. He worked primarily with the NCSX-like plasmas with three-field periods and 18 coils. This configuration with an aspect ratio of 4.5, R=8.3 m, and B = 6.5 Tesla, should produce around 2 GWth power.

The distance from the plasma surface to coil midpoint should be 1.2 m and a delta between coils should be 0.88 m. Long Poe discussed the implications of adjusting the primary machine parameters. The key parameters are coil current density, field intensity, power level, machine major radius, and plasma minor radius. A smaller aspect ratio system will imply kinkier coils and more machine engineering difficulties.

Long Poe Ku’s approach is to assume the Bnorm (coil) = -Bnorm (plasma pressure). For discrete coil configurations, he put a limit on the average of [Bnorm(coil) +Bnorm (plasma pressure)]/Bnorm(plasma pressure) < 0.5% and the maximum of [Bnorm(coil) +Bnorm (plasma pressure)]/Bnorm(plasma pressure) < 2.0%. Other constraints were delta coil to plasma distance of 1.2 m for the nominal machine sizes and the Dmin(coil-coil) = or > 0.85 m. Long Poe then described his analysis and decision flow process to incorporate the constraints and weights, initial coil parameters, and equilibrium plasma data to calculate and optimize the Bnorm. He found that 6 coils per field period (three-field cases) is the minimum number of coils. He examined plasmas with an aspect ratio of 6.8 and 5.9. These traded the value of coil to plasma distance and coil-to-coil distances. He also invoked a 1/R dependence to determine if this will help. He concluded that for these constraints, the optimum coil aspect ratio is around 6.0. The coil design previously provided is not optimal regarding the coil aspect ratio.

Long Poe Ku also examined 10 coils per period for three-field period machines with an aspect ratio of 4.5. The coils were quite twisted and closely packed. Some islands exist and may require saddle coils or wavy PF coils.

He also examined the heat load on the first wall due to escaping a particles. He showed a plot of the a particles that was clustered near theta of –1 with some diagonal structure evident. He estimated the Pa would be around 400 MW for a 2 GW thermal power machine with a total plasma surface area around 800 m2. The peak surface heat flux due to as is around 15 MW/m2.

Long Poe Ku’s next actions are to examine good flux surfaces, evaluate higher b conditions, trade the values of B, b, A and R in cost and systems space, integrate separate studies into a consistent and coherent design configuration.

Recent Results on CS Reactor Optimization – Jim Lyon first explained his analysis to characterize the plasma and coil to assist in establishing the methodology for reactor optimization. He defined the general equations that define the plasma configuration, modular coil parameters, and plasma to coil distance. One of the key parameters, B0/Bmax depends on AD, d, and k (where D is the distance from the plasma to mid coil, d is the coil depth, k is the coil aspect ratio, and AD is the <R>/dmin). He presented some parametric plots of these variables for two different plasma/coil candidates, excluding Paul Garabedian’s two-period case.

Jim generated some POPCON plots to support TK Mau for RF heating system analyses. He assumed some (plasma) profiles to bound the solutions. He showed POPCON plots that were bounded on the left hand side by n < 2nSudo and b < some b limit on the right hand upper regions. The bounding by the b upper limit was thought to be not a valid constraint as the b limit has not been defined yet and stellarators can operate at high b values.

Jim outline his new optimization approach to minimize the <R> with assumed constraints and determine the minimum HISS-95 that would satisfy the total power, neutron wall load, b limit, and still be on a thermally stable branch of the ignition curve. Again it was noted that the b limit is still unknown and it is likely the optimum machine would seek higher b solutions. It was also noted that the smallest machine (e.g. highest wall loading) might not be the optimum solution because of the more frequent change out of the wall.

With the developed database, Jim examined the sensitivities to differing input parameter assumptions. The smaller the distance from the plasma to coil surface, t, lowers the cost of the power core (subject to constraints and implications of wall loading). Jim also said that reducing the Bmax would produce a better reactor with higher b. Again this finding was challenged as past ARIES designs always proved to be improved at higher field cases.

Jim informed the group that his systems code is in the process of being updated to include current mathematics libraries and operating system. The current coil and plasma geometries are being added for all options considered. Les Waganer will provide current cost and material models for incorporation and Leslie Bromberg’s coil model for physics, engineering, and unit costs. Jim explained the methodology for the reactor optimization methodology, inputs, and outputs.

Ideal MHD Beta Limits for ARIES-CS – Alan Turnbull acknowledged several other contributors to this effort, especially for the TERPSICHORE porting to GA Linux workstations. Alan obtained equilibrium data from Mike Zarnstorff for the stability analysis, monitored high mode stability, increased b, or estimated plasma parameter changes for stability. Then he modified the VMEC equilibrium data by iterating stability analyses with TERPSICHORE and converged on final equilibrium when the b limit is reached. He then reconstructed free boundary equilibrium using PIES by modifying coils and plasma parameters to reproduce a reasonable set of free boundary nested surfaces.

Alan showed the CS equilibrium and stability analysis results from the GA version of the VMEC code that achieved an <b> - 4.1% and an aspect ratio of 4.47. He used the three-field case scaled from NCSX equilibrium data. The TERPSICHORE preprocessor was modified to accept input from the latest version of VMEC. The results are essentially equal to those of PPPL. Stable results of the scaled case was restricted to a range of moderately-placed external conformal conducting walls placed between 1.7 and 2.7 times the minor radius of the plasma. Outside this range, TERPSICHORE failed in the vacuum calculation.

Alan showed a range of cases with increasing b with a resulting increase in magnetic axis. He also examined the equilibria for ideal stability over a range of intermediate wall positions. He concluded a b limit of 6% could be achieved for an intermediate wall ~ twice the plasma minor radius, however continuing studies should confirm this result.

As the radius of the wall was less than 1.7 of the plasma radius, the wall approaches the plasma surface at a point on the inboard side at one toroidal plane and the vacuum calculation fails due to a logrithmetic singularity as this is a well-known problem with the TERPSICHORE code.

Placement of the wall for stability with maximum b values will be a difficult issue to resolve due to the complex compact stellarator shape. Researchers around the world are working on this issue and postulating theories to be tested. Also a stellarator equilibrium fitting code is being developed as well as a linear resistive MHD code (Spector 3D). Alan will be working on obtaining a free-boundary equilibrium from PIES and an iteration of the equilibrium and stability results reported herein.

Discussion of the CS Optimization Process and Action Items - Farrokh Najmabadi grouped the team’s activities into five distinct work areas that have been addressed to varying degrees to the present and then discussed what topics need to be addressed in more detail before the meeting in December 2003. The table below is intended to capture the essence of the summary of past work in preparation for the work to be accomplished in Action Item List that follows.

Description
Fixed Boundary Equilibrium
Filament Coil Representation
Realistic Coil Representation
System Code Evaluation
Engineering Analyses and Assessment
Who
Ku and Garabedian
Ku
Bromberg
Lyon and Waganer
Engineers
Work to date
Several Concepts incl NCSX-like with A=5.89 and MHH2-like
NCSX-like with A=5.89
Conductor selection; construction approach
Parametric search on physics parameters
Radial build scoping; preliminary maintenance approach assessment
Actions before next meeting
See Action Item List Below

CS Configuration Development Action Items

Responsibility
Due Date
Actions To Be Accomplished
Waganer 9/22/03 Update fabrication cost for power plant materials and fabricated densities
Ku, Garabedian 9/22/03 Determine Bmax/BO, coil positions, and coil currents for two-field period MHH2 case. Plasma shape, filament positions, and coil currents to be send to UCSD for CAD drawing.
Wang 10/15/03 Generate CAD drawings (dimensions updated per Lyon’s parametric survey)
Lyon 9/30/03 Redo survey of system parameters (sensitivity to t, d, k) done for GT meeting and generate tables of system parameters for device operating at maximum b for both NCSX and MHH2 cases.
Lyon 9/30/03 Send list of required engineering constants and variables for his studies.
Raffray, El-Guebaly 10/15/03 Provide best current estimates of these parameters
Bromberg 9/30/03 Define coil engineering data: Jmax, J/B, Cost/B, coil structure, coil case thickness, winding pack composition, materials, weights, stresses, thicknesses, as well as costing algorithms with documentation
Najmabadi 10/3/03 Post data on ARIES Web site the three-field period A=5.89 case and two-field period MHH2 case
Wang, Malang 10/8/03 Refine/verify clearances for maintenance approach.e.g. Can shaving minimal magnet intrusion provide substantial increase of the allowable replaceable unit thickness?
Raffray, El-Guebaly, Wang, Malang 10/8/03 Define/refine FW, blanket, shield, vacuum vessel, and support structure as well as maintenance scheme for both CS concept approaches (based on J. Lyon’s optimized reactor parameters)
Lyon 11/15/03 Provide cost-optimization studies of the three-field and two-field period cases.

Engineering Development and Optimization

Radial Build Definition for Li/FS and LiPb/FS System (Concepts) – Laila El-Guebaly presented the initial physics and engineering parameters for the L.P. Ku three-field period CS strawman concept. She then highlighted the nuclear design requirements and radiation limits to be satisfied for an acceptable design concept, along with the underlying design assumptions.

Five candidate blanket concepts were presented for consideration. The structural material is ferritic steel for four concepts and SiC for the remaining as compared to vanadium for the SPPS design approach. One differentiating factor was the location of the vacuum vessel, either inside or outside of the magnet. She concluded it was better to have the vacuum vessel inside the magnets as this approach yields a more compact design with much smaller distances of D. It is also better to have a water-cooled vacuum vessel (VV). Laila showed a radial build of the proposed blanket and shield approach of LiPb/FS/He blanket, FS shield, and water-cooled FS internal vacuum vessel. The characteristic distance (D) from the plasma to the mid point of the winding pack is ~1.43 m for the blanket zone and 1.12 m for the shield only zones. She also presented data for LiPb/FS/He helium-cooled external VV and a Li/FS/He helium-cooled external VV concept. The characteristic distances, D, were presented for all five concepts, with the Flibe/FS/Be being the thinnest blanket and radial build.

A similar assessment for the shield-only concepts yielded results that did not show a concept with a strong advantage with the conclusion that the vacuum vessel location has a negligible impact on the Dmin where a shield-only is required. This Dmin condition occurs twice each field period or six times in the power core and covers approximately 8% of the surface area for the strawman concept. A tungsten-carbide (WC) shield only approach has benefits from a more compact design, but there are engineering problems.

Laila's conclusions are:

  • Radial standoff is a powerful parameter that strongly influences power core design and COE
  • The minimum standoff is so influential that efficient, but expensive, shields should be used:
    • WC shield
    • Liquid breeder to cool high temperature shield (no He)
    • Water to cool low temperature shield (no He)(water is a good shielding coolant)
  • Internal VV is better, but other factors may prevail
  • Non-uniform blanket, covering ~ 98% of FW area, should meet breeding requirements with margin for all proposed blanket concepts

Power Core Maintenance Scheme for Compact Stellarators Based On Sector Replacement – Siegfried Malang said there are three schemes for maintaining CS reactors:

  • Replacement of blanket modules with articulated booms through several maintenance ports (discussed in Jan 03 meeting)
  • Replacement of modules through a port between each modular coil
  • Replacement of modular blanket/coil sectors (ala, ASRA6C study and the SPPS study)
The latter was the focus of this presentation.

Siegfried noted there are three dominate forces acting on the coils: large centering forces pulling the coils toward the center (need a strong bucking cylinder), out-of-plane forces between neighboring coils (need strong inter-coil structure), and the weight of the cold coil system (strong, but thermally resistive supports). Removal of whole field periods are preferred as there appears to be less interference between neighboring coils at the field period parting lines. It is desirable to move a whole field period without disassembling the coils to avoid disassembly of coils, thermal insulation, and intercoil structure. Because of the difficulty to efficiently transfer large loads between warm and cold structures, it was decided the entire support structure should be cold. Separate cryostats for each field period are recommended. Inner and outer cryostats are used in conjunction with a large bucking cylinder inside the torus. These individual cryostats are enclosed in a common, segmented and removable vacuum vessel. The coils for each period are wound on a supporting tube with grooves for the coils. Supporting structures for the cold coil cases are thermally resistive to minimize heat losses. Heat transfer tubes would be located between supports at the bottom of the reactor.

As in the prior ARIES designs, blanket and shield regions would be designed as zones with appropriate lifetimes to minimize replacement costs and time. The refurbishment of the blankets and hot shields will be accomplished in hot cells or designated maintenance areas with suitable radiation controls. Siegfried explained how the internal components would be designed, supported, and maintained. More detail was provided on the coolant plumbing design and cooling parameters for the helium and lithium. Siegfried amplified the design implications relating to the challenging CS-unique shield-only areas. Safety concerns involving the use of water because of its better shielding properties arise when used in proximity with lithium or, to a lesser extent, lithium lead. Heat transfer systems must be designed to minimize the likelihood or consequence of coolant spills. A passive conductive or conductive cooling system for the low temperature shield is mandatory.

Siegfried then examined the feasibility of removing replacement blanket units consisting of the first wall, breeding zone, and a high-temperature structural ring that are surrounded by high temperature shield and low temperature shield. Could these units be removed out the ends a complete field period sector? He constructed representative cross-sections at the 0°, 30°, and 60° locations with projected overlays to determine if the outer units could be withdrawn and then successively withdraw the inner units without interference with the high temperature shields throughout the sector. He concluded that a delta distance of 1.2 m would be insufficient for removal of the blanket units. If the FW/blanket/support is reduced to 0.82 m, extraction is possible. Movement on a circular rail centered on the torus axis will not work, but a circular arc, slightly offset, would work with the thinner blanket units.

Initial Assessment of Maintenance Scheme for two-field Period Configuration – René Raffray summarized the three maintenance schemes to be considered and evaluated during the first year of the ARIES-CS study:

  • Sector replacement including disassembly of a modular coil system (see Malang's presentation
  • Replacement of blanket modules through maintenance ports (initial assessment for two-field period configuration)
  • Replacement of blanket modules through small number of maintenance ports using articulated boom (presented May 2003)

Other scoping studies involved possible blanket/shield/divertor configurations compatible with the maintenance schemes and machine geometry:

  • Self-cooled liquid metal blanket (LiPb) (presented May 2003)
  • He-cooled liquid breeder blanket with FS (proposed at this meeting)
  • Flibe-cooled ferritic steel blanket (presented at January 2003 meeting)

The coil configurations are also being evaluated with regard to material selection and thickness, radius of curvature, shape, space, and shielding requirements.

René summarized the key issues with the modular maintenance replacement scheme with selected ports and an articulated boom. He contrasted that with the plasma shape and coil shapes for the two-field period configuration proposed by Paul Garabedian, which has considerably simpler coils and more room between modular coils. He quantified the port dimensions for the NCSX-like configurations for two radii machines and the two-field period configuration with the latter having much improved port sizes. Further study is needed for physics and plasma configuration parameters, coil size and shape confirmation, benefits of larger blanket modules, more detail on the maintenance scheme, and a solid-CAD definition of the plasma shape over the field period.

He also posed several safety questions on the number of protection barriers, helium leak into the steam cycle, acceptability of a water-cooled shield/vacuum vessel with a LiPb blanket, suitability of an external vacuum vessel, and the possible material interactions.

Stellarator Magnet Conductors – Leslie Bromberg illustrated the beneficial properties of YBCO High Temperature Superconductor (HLTS) as compared to LTS superconductors. At low fields, YBCO theoretically has a great advantage in current density. In the US, YBCO is the only HTS being pursued, although Japan and EU continue with others. The cost is predicted to be eventually in the range of $10-20/kA, which would be competitive with copper. LANL has developed a process of coating a Ni-superalloy with IBAD-textured MgO. Leslie showed a continuous process for YBCO to produce 4-mm wires. American Superconductor (AMSC) and ORNL are developing and analyzing various thicknesses of YBCO sample conductors and the current density decreases as the thickness increases. The stability margin is much larger for HTS than LTS. Leslie summarized the current design properties of YBCO as of July 2003. He also showed how the stellarator could be fabricated in the grooves in shells as Siegfried Malang is proposing.

To decrease the cost of fabrication of LTS magnets, in low-field regions, NbTi LTS could be used with Nb3Sn in high field regions. However, winding such magnets would be difficult and costly. The solution is to use HTS at low temperature. Leslie then showed a LTS design of the conductor proposed for the CS reactor. In more careful analysis, the current density should be around 100 MA/m2. Leslie showed a cost comparison of NbTi, Nb3Sn, and YBCO and provided magnet design algorithms.

Safety –Related Design Issues – Brad Merrill asked the audience how many safety barriers are needed in a commercial fusion reactor? The ITER design adopted a Defense-in-Depth safety concept with multiple passive and active safety measures. A passive measure would be the use of physical confinement barriers, where as an active would be an isolation valve or pressure relief valve. The ultimate measure of success is as many barriers as required to assure that releases are well below the site limits during Design Basis Accidents. A schematic of the ITER confinement barriers were shown. ITER adopted two confinement barriers; a primary barrier with a failure probability of <10-3 per challenge and a leak rate of 20%/day at a pressure of 200 kPa. Specific pressures and leak rates were given.

The EU Power Plant Conceptual Study uses one barrier for in-vessel LOCAs and two barriers for ex-vessel LOCAs. For the ARIES-CS, Brad recommended two barriers, as was the case for ITER and ARIES-AT and –RS. Brad then analyzed several accident scenarios that must be addressed:

  • Helium-cooled module be designed to take secondary cycle steam pressure (more details are necessary)
  • Water-cooled shield in combination with a LiPb blanket (maybe not)
  • Safety issues with an external vacuum vessel (possible mitigation solutions)
Two confinement barriers are recommended. For a helium-cooled module with a steam system, the steam pressure must be addressed as a safety issue. Additional safety investigation of the LiPb blanket needs to be completed. A passive decay heat removal system must be provided.

Updated Concerns on the Material Interactions
– Dai-Kai Sze discussed the material interactions and compatibility issues, especially the temperature-driven corrosion rate between structural materials and coolant/breeding materials that should be ~10 mg/m2-h. This corrosion problem is not well documented and the existing experimental results do not support the conventional operational window. Lithium is compatible with vanadium up to high temperatures (600ºC to 700ºC per the BCSS study), although some data would disagree. ANL had measured the dissolution rates of vanadium in lithium in a forced convection loop and found the nitrogen concentration to be 20 to 100 wppm. Corrosion rates of vanadium alloys were also provided. Corrosion work by Russians was also shown. Dai-Kai questioned the validity of the database.

Dai Kai noted that the community has suggested that LiPb and SiC are compatible up to 800ºC or higher. A reference experiment for this claim was a rotating SiC disk at 800ºC in heated LiPb. However, it was not possible to separate the LiPb from the SiC sample, thus no information is available on the corrosion. ORNL recently did a SiC-LiPb compatibility test at 800ºC and 1000ºC. They are also having difficulty in removing LiPb from the Si coupons to determine the extent of corrosion.

ARIES IFE Assessment

General IFE Overview

Perspectives on Recent Progress and Future Directions for IFE – Grant Logan summarized the ARIES Contributions to Heavy Ion Fusion (HIF) Inertial Fusion Energy (IFE), especially clarifying the important driver-target-chamber interface issues. These include determining viable operating windows for chamber pressure, beam propagation, and pulse rate for the three leading chamber protection approaches. ARIES also updated the materials development needs as well as exploring the final focus shielding and manufacturing/delivery of IFE targets.

Grant examined the remaining HIF-IFE challenges of achieving a faster development path, contributing to High Energy Physics, and advancing the target technologies. These challenges will be explored in a new FESAC panel in 2004.

Grant then discussed the technologies that will help advance HIF-IFE: multi-beam quad-array linac, faster/cheaper HIF development pathway, new targets to be tested on “Z” at Sandia, new focusing schemes for higher peak currents and fewer beams on larger target beam spots, and new modular-solenoid driver with a vortex cusp chamber. These larger focal spot targets might enable a single beam IRE to validate the driver module and a variety of integrated experiments with the promise of a faster development pathway. Grant touched on a variety of research and development needs that remain and key areas that will be explored in the near future to help advance IFE.

Chamber/Beam Physics, Beam Transport, and Chamber Clearing

Self-Pinched Transport Modeling – Craig Olson noted the impetus to develop the self-pinched beam transport (SPT) is to enable the simplest, technically attractive, and least expensive chamber option that is applicable to all chamber concepts. Craig reviewed the SPT areas of research noted in the Workshop on Final Transport of Heavy Beams at the San Damino Retreat in February 13-15, 2001, which were:

  • Self-pinched radial equilibria
  • Beam conditions for transport in both 1 torr and 10-100 mtorr gas atmospheres
  • Beam conditions for transport in preformed plasma channels
  • Creation of isolated self-consistent plasmoid equilibria (no chamber concepts proposed)

Craig reviewed the results of the first SPT trumpet experiment on Gamble II at NRL that demonstrated onset of pinching as predicted by IPROP. Additional experiments at higher currents (>1 kA) on IRE with possible nearer term experiments. He is hoping for a small equilibrium beam radius for channel-like transport with high current and small emittance. Driver-scale simulations indicate self-pinched transport equilibrium conditions.

New analytical modeling of trumpet equilibria indicates neutralization is needed for self-pinched transport. This neutralization can be obtained by space charge neutralization, axial trapping of electrons, or providing a trumpet to pull electrons into the beam channel. Craig then explained how the trumpet is envisioned to work.

Although there is additional work to be accomplished, he thought the self-pinched transport would be the most attractive beam transport mechanism that is applicable to all chamber protection concepts. It has been demonstrated and new methods (ala, trumpet) may improve the mechanism.

Simulations of Neutralized Drift Compression – David Rose presented results that he and Dale Welch had accomplished. They examined the neutralized plasma drift region into which 10-80 ion beams enter for combining and compression at optimal transport conditions. The primary concern for a driver system is both electrostatic and magnetic self-fields. They examined the problem with simulations of a driver beam injected into a preformed-plasma with a solenoidal field but no compression. Net currents of 6-7 kA were predicted for the beam without an applied Bz field. Net currents decrease with increasing Bz and skin depths to the cyclotron radius. David noted the beam transport improves with plasma density as long as wc > wp. The electrostatic potential drops with the plasma density. The transport is most ballistic with large plasma density and applied solenoidal field. He concluded the neutralized drift compression requires an axial compression of 1000 with beam current compression to 300 A. This is confirmed by a full Lsp simulation. There is no significant growth in emittance and only a small growth in longitudinal energy spread. Simulations show a robust compression in a neutralizing plasma.

Gas Transport and Control: Challenges, Opportunities, and Plans for the Future – Christophe Debonnel reviewed the challenges and opportunities for gas transport and control up the beam lines. The steady-state chamber atmosphere density is set by the nature and temperature of the liquid jets and driven by target and beam transport requirements, whereas the high-pressure impulse and density rise after the target explosion is determined by the target yield/spectra and the chamber protection approach. The gas density in the beam tube needs to be low enough to avoid degradation of the beam quality and prevent arcing between the wall and space-charged beam. A two-part solution is to design efficient target chamber protection structures and a beam protection approach with molten salt vortex and magnetic shutters to deflect the target debris. The TSUNAMI code has been upgraded to simulate and analyze the chamber and beamline gas dynamics after the target explosion. Time-wise density and pressure movies are generated to visualize the movement of the target debris and pressure pulses.

Christophe discussed the development of new radiation hydrodynamic TSUNAMI modules to model the target chamber and beamline tubes, including methodology, physical models, boundary conditions, and initial conditions. He has been modifying the TSUNAMI code to accommodate the physical models and processes. He is considering hydrodynamics in two dimensions, condensation/evaporation on liquid jets, real gas equation, radiation, fast ions energy deposition and ignoring neutrons, jet motions, and in-flight condensation. He explained the liquid heat transfer module and the radiation module, the boundary conditions, and initial conditions used. He then compared his results to predictions made by Susana Reyes, LLNL, with the ABLATOR code.Agreement between the two codes were good. A comparison with UW BUCKY code showed some differences, but Christophe thought the higher ablation depth was due to the use of hot vs. cold opacity data. He is planning on continuing to upgrade TUSUNAMI to be more user-friendly and accurate.

Liquid Wall Ablation under IFE Photon Energy Deposition at a Radius of 0.5 m – René Raffray reminded the team that he had examined and reported ablation effects of wetted walls at radii of 3.5 m and 6.5 m during the ARIES Town Meeting Chamber Dynamics Session. It was recommended at that time that he examine a small radius, 0.5 m, case typical of the HYLIFE-II chamber protection concept. The photon energy deposition density profile in a flibe curtain at 0.5 m radius was shown that indicated a zone of evaporated material, an explosively boiling region, and a heated region. The explosively boiling region is of the most interest as this region can eject significant quantities of liquid into the chamber atmosphere. He summarized the explosive boiling region thickness for all three radii for both flibe and lead.

A second mechanism for injecting liquid into the chamber atmosphere is the mechanical response due to the induced shock. The shock propagates through the liquid layer and can:

  • Reflect off a solid back wall and propagate back through the layer and spall at the front free surface
  • Reflect off a non-perfectly still back wall and less-intense wave is reflected back to the front face and spall a lesser amount
  • If there is no solid surface, the rear surface may spall

René illustrated the spall strengths of the candidate materials and the key parameters from several prior studies and their liquid protection materials. He then showed the pressure waves for the three conditions mentioned above. All this data (explosive boiling thickness, spallation thickness, and total fragmented thickness) was compared for flibe and lead at the three candidate radii. It was noted the spallation thickness is not perceived to be a problem for HYLIFE-II as the jets at 0.5 m are not a solid jet, rather they are a combination of smaller individual jets that will absorb any spallation of the first set by succeeding jets.

Modular Solenoid with Assisted Pinch (Transport) – Simon Yu described a new modular solenoid approach that is an alternative to the Robust Point Design multiple beam accelerator driver. This new approach provides a shorter and less expensive development pathway to fusion energy because the driver is comprised of many similar modules and only one is required to demonstrate feasibility and proof of concept. The solenoidal driver with neutralized drift is compatible with the hybrid target and assisted-pinch chamber transport. Simon described the solenoidal transport experiment that would demonstrate the concept feasibility.

Chamber Engineering

Hydrodynamic Source Term: Boundary Layer Cutting and Flow Conditioning - Minami Yoda described the HYLIFE-II stationary and oscillating liquid jet concept used to protect the chamber walls and beamlines. These jets are a potential hydrodynamic source term for aerosols in the chamber environment. Therefore, Said and Minami agreed to analytically model and experimentally test the jets to validate the analysis and experimentally predict the source term. The source term arises from the turbulent breakup of the flow. Analytical prediction based on empirical correlations of a turbulent jet (no flow conditioning or boundary layer cutting) determined the source term to be 1300 kg/s, which is unacceptable for most, but not all ion beam transport schemes. The HYLIFE design is planning on flow conditioning and boundary layer cutting, so these factors were assessed experimentally by the HYLIFE team.

The Mechanical Engineering Laboratory in Georgia Tech set up an experimental flow loop to study the effects of the flow conditioning and boundary layer cutting. Flow conditioning was accomplished with a perforated plate, honeycomb, and fine mesh screens in conjunction with a 5th order polynomial contracting nozzle with an area contraction ratio of 3. Water was used as the working fluid for flows at similar Reynolds, Froude, and Weber numbers.

Independently, a boundary layer cutter was used on one side of the flow stream. The average surface ripple was reduced most with flow conditioning, but the boundary layer cutting also helped, with both being most beneficial. A mass collection experimental set up was used to compare the base case or standard flow conditioning and no boundary layer cutting with those including flow conditioning without a fine mesh screen (the element most likely to become blocked) and/or boundary layer cutting. Flow straightening and contracting nozzle significantly reduced ejected droplet mass by 3-5 orders of magnitude. Boundary layer cutting appears to eliminate droplet ejection for a well-conditioned jet, barring blockage of the fine mesh screens.

Also, tests of a flowing liquid around a cylindrical or shaped penetration were extended to examine a penetration surrounded by a thicker depressed wall zone to accommodate the slowing fluid around the penetration; the geometry used was a scaled-down version of that simulated in the APEX studies. It was hoped the depression would allow the fluid to flow around, rather than over the penetration. However, tests indicated, the flow continued to flow over the penetration, therefore the concept was not successful.

Target Fabrication, Injection, and Tracking

Heavy Ion Fusion Target Materials Selection and Removal of Tungsten Carbide from Flibe – Ron Petzoldt discussed a detailed evaluation of the hohlraum materials for the heavy ion target. Key parameters were determined, but Ron concluded the final selection would require a systems trade. Key elements considered were:

  • Fabricability
  • Target energetics
  • Materials separation
  • Radioactive inventory and handling
  • Materials compatibility
  • Injectability

Some of the target material combinations considered were: Au/Gd, Hg and Hg/Xe, PbHf, PbW, W, and PbHfXe. Tungsten was one of the more attractive materials from a manufacturing and waste considerations, but it posed a possible separation problem from Flibe. If fine W and WC particles were added to the Flibe, the seeded particles would increase the surface area and would collect the W or WC, thereby minimizing plating out elsewhere. Crossflow sintered metal filters are recommended to remove the WC particles. He concluded there were existing viable target materials that are compatible with Flibe. Tungsten is the leading candidate.

Summary of Thick Liquid Chamber Protection

ARIES IFE – Some Concluding Comments on Thick Liquid Walls – Wayne Meier stated the ARIES-IFE work and interaction with other OFES-funded effort on thick liquid chambers has been interesting and beneficial. Key findings for each were provided.

  • Reexamined structural material options
  • Determined economic impact of first wall replacement, if needed
  • Updated mechanical design and defined maintenance sequence
  • Facilitated workshop on liquid chamber dynamics
  • Produced new views on recycled target materials
  • Improved final focus magnet shielding designs

Summary of Issues, Results, Findings, and Recommended Research and Development – René summarized the documentation of the ARIES-IFE study. There will be a special issue of Fusion Science and Technology. Some papers have been returned for slight updates with the remaining critiques in a month or two. Hope to have the manuscript completed by December 2003. This issue will be devoted to ARIES-IFE results.