ARIES Program
Public Information Site


ARIES Project Meeting Minutes

23-24 January 2012

UCSD, San Diego, CA

Documented by L. Waganer

Organization ARIES Project Team Members
Boeing Waganer, Weaver (phone-in)
FNTC Malang
General Atomics St.John, Turnbull
Georgia Tech -
INL Humrickhouse, Merrill (phone-in)
LLNL Rensink, Rognlien
PPPL Kessel
UCSD Carlson, Hossenini, Najmabadi, Tillack, Wang
U of T, Knoxville -
UW-Mad Blanchard, El-Guebaly

Ref: Agenda and Presentation Links: Meeting Agenda


Welcome/Agenda - Mark Tillack welcomed the ARIES team to UCSD and explained the facilities of Building Eng Building U-I. Les Waganer reviewed the day and a half agenda (see above agenda link) and the topics to be presented.

Next meeting - The next meeting will be held in about three months, date TBD. Farrokh Najmabadi and Mark Tillack will decide the date and time for the next conference call to finalize the initial design configuration.

Plans and General Scope

ARIES Project Update and Near-Term Objectives - Farrokh Najmabadi stressed the importance of completing the current study to define four unique designs within the next 18 months. We must work closely together to close the integration data loop for the two physics cases and the two physical configurations that will integrate with each other to achieve the four cases. We also need to update the systems code with its underlying precepts. For the immediate future, we will concentrate on one data set and finalize it. It is important to be able to address the transient conditions to make sure the case is technically reasonable to the fusion community. The new VV design is a critical part of the overall solution. The cases must also be maintainable, so the design configuration and maintenance approach must be credible.

ARIES Task Results (Engineering)

Updating the SCLL Design and ASC Documentation - Lane Carlson said that he had been updating the ARIES System Code (ASC) with new modeling definitions and documentation of the ASC inputs, internal coding and text/graphics outputs. A part of the new definitions is the improved modeling of the SiC design cases, including a new radial build, plasma surface area, and power balance. All of the new ASC features and documentation is contained within the new ASC home page. The ASC documentation is complete and easy to navigate. Lane has also completed all of the action items from the last ARIES meeting. The ASC modifications will continue to evolve to match the physics and engineering changes. He is ready to implement the new SCLL design changes. The filtering capability of the code will help the team to determine the choice of the four design points from the myriad of possible solutions.

Demonstration of 3-D Mockup of Divertor Design - Lane Carlson showed a 3D thermo-plastic model of a divertor module that could be easily taken apart for visualization of the assembly process and the intricate internal details that had been created with the 3D printer.

Progress on Coolant Routing and MHD - Mark Tillack summarized his progress on action items from the previous meeting as well as new action items he added. He illustrated the ARIES-AT magnetic field strength along the planned coolant flow paths. He noted there are large poloidal fields along the flow paths even outside the TF coils. Mark showed where flow stagnation can occur in the curved FW ducts. The heat transfer was previously approximated with laminar slug flow. The new profiles that include the stagnation effect result in both 50% higher pressure drop as well as increased wall temperatures. He evaluated the temperature increase in the internal blanket ribs as being less than 40°C above the bulk coolant temperature (note that the current design does not employ ribs). Mark emphasized that 3D MHD flow effects are very complicated and highly correlated to channel geometry. He showed several empirical models for representative flow channels. In addition to MHD, there are inertia, gravity and wall shear forces. The 3D MHD effects dominate.

To help reduce the 3D effects, Mark offered some design strategies. He discussed his visit at Karlsruhe Institute of Technology to confer on MHD design solutions, and the plans for future collaborations. In summary, Mark thought the MHD pressure drop could be acceptable (around 1 MPa), but the results are uncertain due to limited modeling and experiments. MHD and thermal hydraulics analyses for the SCLL blanket will be finalized in the near future, and boundary temperature conditions will be defined for thermal stress analysis.

Refinement of the Power Core Configuration for the ARIES-ACT SA - Xueren Wang highlighted his action items from the previous project meeting. He compared the previous power core to the current design that incorporates several improvements including a straight inner wall and blanket, reduced divertor target area, single curvature of outboard blankets, embedded tungsten shells and helium-cooled HT shield. He noted the manifolds and access pipes are in work. Xueren explained a possible fabrication procedure for the inboard and outboard blanket with vertical and kink shells. Joints for the electrical connections are provided at the back of the HT shield. Xueren outlined the design considerations for the forces on the blanket elements. He then showed the features and stress results of the blanket sector comprised of a different number of concentric tubes. The side wall needs to be thickened to accommodate the hydrostatic pressure of the LiPb coolant and the MHD pressure drop.

Xueren explained the design considerations of the vacuum vessel - vertical motion of the control coils and clearance for removal and replacement of the power core sectors. At present, both water and helium coolants are being evaluated for the vacuum vessel. The tentative routing of the blanket coolant pipes are similar to those for the ARIES-AT design as being on the lower floor of the vacuum vessel inside the VV door. The VV door and port arrangement are being refined to eliminate the fixed wedges used in the ARIES-AT design while retaining the ability to remove the sectors. Future work will involve selection of the VV coolant and wall thicknesses and stress analyses

Preliminary Stress Analysis of the Vacuum Vessel - Hamed Hosseini, of UCSD, has analyzed the preliminary stresses of the ARIES-AT vacuum vessel fabricated from 316SS. This stress analysis was based on the existing CAD files, but the VV was represented as single-walled, 5- and 10-cm thick vessels. The primary load on the vessel is the atmospheric pressure on the outside and the vacuum inside the vessel. The original ARIES-AT design had not been analyzed for stresses, thus there is no design effort on localized stress concentrations or high stress regions. Hamed's preliminary stress results highlighted the presence of stress-concentration areas, such as sharp corners and transitions. The flat areas in the maintenance ducts exceeded the design stress, which would be candidates for localized strengthening and elimination of stress-concentration factors. Hamed said he would run the TMAP code to yield better results. Les Waganer suggested adding ribs to the high stress, flat panels to reduce the stresses

Fracture Studies for ARIES Plate-type Divertor Concept - Jake Blanchard showed the design concept of the tungsten, helium-cooled, flat plate divertor. At the last meeting, he had results with a crack positioned in the upper coolant region under the tile, centered in the channel. This time, he modeled the divertor under full-power, steady-state thermal loads. He described the stress intensities he used. The current proposed crack is located in the FS structure under the tile near the centerline of the coolant opening and oriented parallel with the coolant flow. The maximum stress intensities occur not during operation, but during cool-down due to plastic deformation. These intensities are lower when modeled in 3D. Jake's second analysis was in the FS structure on the centerline, at the coolant interface and with the crack running perpendicular to the coolant flow - this stress is less than the crack at the structure/tile interface. Creep may be included in the next analysis. It was also suggested to add the ELM transient loading. Jake needs fracture toughness data on tungsten, both as manufactured and after irradiation.

Recommended VV Design Load Criteria (Power Plant Relevant Design) - At the last ARIES meeting, Jake Blanchard presented the ITER vacuum vessel design loads. This meeting's presentation focused on his recommendations for the ARIES design loads. (The ARIES-ACT VV design is currently in a state of flux - double vs. single wall, 316-SS or 3Cr-3WV ferritic steel, choice of coolant (water or helium) and addition of stress considerations into the design requirements (added by LMW)). Jake is considering the anticipated load conditions of weight, pressure and thermal (both operational and baking). In addition, many off-normal conditions are also being considered. Combined loads will include normal, exceptional, hypothetical and test. He suggests using the ITER Structural Design Criteria pending a more detailed analysis. Off normal loads would consider seismic, disruptions, LOCA, LOFA and in-vessel pipe breaks. The combined loads would be considered after the single-load cases are completed. Non-specific cases will be skipped.

Impact of Water-Free Vacuum Vessel on ARIES-ACT-SiC Inboard Radial Build - Laila El-Guebaly mentioned the 2006 ARIES-AT inboard radial build that contained a double-wall VV component with FS structure, H20 cooling and WC shielding. There are several proposed changes being considered for the ARIES-ACT design that will negatively impact the radial build to achieve the necessary shielding, TBR and activation requirements. Laila noted the general impact of the changes including lower temperature TF coils, a redesigned skeleton ring with helium cooling (instead of LiPb) and FS structural material, eliminating VV water and WC and adding a thermal shield. Laila showed the radiation limits for the SiC concept components and then presented some radial builds and charts illustrating the impact of the design options. The team suggested that Laila analyze a thicker blanket (and thinner shield) to help find more radial build options, but not necessarily the thinnest one. Laila will conduct a trade off of VV, shield and blanket design options while maintaining the shielding, decay heat and activation criteria.

Breakdown of Elements Degrading TBR of ARIES-ACT-SiC Blanket - Laila El-Guebaly outlined the principle features of the ARIES-ACT SiC blanket design associated with aggressive physics. She showed the initial inboard and outboard radial build of the blankets and HT shields and summarized the generic TBR requirements to guide her analysis using the UW DAG-MC 3-D neutronics code. The 3-D model features were explained along with a step-wise analysis approach to arrive at the appropriate TBR. She noted some divertor geometric inconsistencies between the CAD model and the LLNL/PPPL model. She showed the progressive reduction of the theoretical TBR (~1.8) down to the more realistic one (~1.05) when the real geometry of the power core, including penetrations, is addressed. This analysis approach will be used to determine the likely TBR when the more definitive power core and plasma features are adopted, per her final slide showing her "to do" list and data needs.

Nuclear Heating Profiles for ARIES-ACT SiC Design - Laila El-Guebaly displayed the key parameters for the 3-D nuclear heating analysis of the current ACT SiC design. The nuclear heating was computed for the inboard, outboard and divertor regions for all the inner power core components for both 1/32 and 1/16 of the power core. Following this, she provided more detailed data for the internal components. She ascertained the energy multiplication factor would be 1.126 for the current set of materials. She also determined the thermal power split absorbed and conveyed in the power core coolants to be 27% in the He and 73% in the LiPb coolant.

Nuclear Heating Profile for ARIES-ACT DCLL Design - Laila presented a parameter set and analysis results with new DCLL values similar to those presented for the SCLL design. The current DCLL concept produces about 2760 MW as compared to the SiC version producing 1907 MW of fusion power (note that all concepts are normalized to ~ 1000 MWe so material choices, thermal conversion efficiency, and pumping powers have a profound impact). As a result, most DCLL power core elements have approximately the same ratio of heating in the power core elements. However, the heating in the DCLL divertor is more than double the SCLL divertor because of the definition of the divertor shield. The energy multiplication is almost the same in both cases. The ratio of the He to LiPb power split is 46/54 as compared to the SiC case of 27/73, so the helium transfers much more power in the DCLL design.

Coolant Connections to the DCLL Blanket - Siegfried Malang showed the DCLL liquid metal coolant connections from the skeleton ring and manifold at the bottom of the power core to the external plumbing piping in both a side view and top view of the manifold in the lower skeleton structure. Care was taken to provide gentle bends with respect to the local magnetic fields to minimize LM MHD effects. The flow is divided into two or three manifold sections to assure equal flow division. He also devised an approach to transition from the skeleton ring to the blanket that enabled remote cutting and re-welding of the pipes. He further provided a sequenced set of actions to remove and replace the blanket sector.

ARIES Task Results (Physics)

MELCOR Model Development for ARIES-Safety Analysis - Paul Humrickhouse presented the status of the MELCOR modeling of the SiC and DCLL ARIES-ACT design concepts. Two separate models exist, one for the ARIES-AT SiC design and one for the DCLL design, and they now include the ARIES-CS VV and a cryostat based on the ARIES-AT design. INL is testing these models based on the long-term-station-blackout (LTSBO) accident, with decay heat removal by gas injection into the cryostat volume. He explained how the heat is transferred from the hot elements to the heat sink elements. He has not run a case without air (or a gas) as Laila suggested nor considered the FS-based skeleton ring/shield of the ARIES-ACT SiC design. Paul noted that the choice of multi-layered super-insulation instead of a low-emissivity thermal shield would likely reduce the effectiveness of decay heat removal by air injection into the cryostat. In the near future, he will improve the physical modeling of the latest radial build and decay heat distribution, include the VV coolant media, consider a thermal barrier and divide the model of the VV interior into several volume zones to better characterize the heat flow during an accident.

Update on Divertor Plasma Solutions - Marv Rensink presented the modeling he and Tom Rognlien have been doing of the divertor plasma region. The principle parameters that affect the plasma and hardware in the divertor region are the tilt angle of the divertor plates, core plasma boundary conditions, impurity radiation and particle pumping. Definitive answers are difficult because there may be multiple solutions, oscillatory behavior may exist and the model mesh sizes may be too large. Many UEDGE runs have been cataloged in a spreadsheet for further data analysis.

Neutral particle removal is determined by the albedo on the private dome. Private flux pumping can control transition from attached to detached plasmas, resulting in a significant particle throughput. The more favorable detached plasma regime occurs over a small range of albedo. Tilting the plates may be both favorable (increased wetted area, hence lower heat flux) and unfavorable (inward radial transport of particles may produce higher temperatures in the SOL and higher heat fluxes on the divertor plates). Tilting the plates can control attached/detached plasma solutions.

ACT2 Preliminary JSOLVER Equilibria - Chuck Kessel is using the JSOLVER to search for a suitable set of solutions for use on ARIES-ACT advanced physics cases. So far, he can find stable ballooning cases around ?N in the range of 2.8-2.9, but cannot find n = 1 cases without a (stabilizing?) wall. He intends to finish the JSOLVER analyses to identify ideal MHD stable configurations. He also wants to develop a free-boundary equilibrium for other analyses. He can also do a TSC time-dependent analysis for ACT-2 including both heating and CD. He may also conduct ELM and other disruption analyses for ACT-2.

Disruption Specifications in ARIES - Chuck Kessel defined the predicted disruption specifications for the PFCs, namely the divertor heat, divertor particle, first wall heat and first wall particle. He considered the transient conditions, including disruptions, runaway electrons, fast confinement losses and fast alpha particles. Major concerns include energy deposition, EM loads and runaway electrons. For ARIES plasmas, a major disruption (MD) and a vertical displacement event (VDE) are most severe disruption events and will be addressed first. For disruptions, the time sequence includes energy loss in the pre-event phase, a fast thermal quench (~ 1 ms), plasma quench (~25 ms) and then runaway electron generation. Chuck then compared the features of the VDE and MD events. For the high power plasmas, such as ARIES, he is anticipating that the thermal quench energy release is similar to the plasma stored energy, unlike current experiments that have lower thermal quench energies than their plasmas. The thermal quench time is roughly correlated to plasma volume, but exact correlations are difficult to obtain. Typically, the energy release is roughly 25% in the rise phase and 75% in the decay phase. The high energy density in this energy release may cause PFC material damage. It is also important to examine the deposition footprint because in quenches the area may expand an order of magnitude. Chuck then provided some numerical examples for consideration. He said halo currents form in the latter stages of the current quench emanating from the cold plasma region to the power core conducting surfaces completing the current circuit.

How Do We (ARIES) Deal with the Power/Energy Fluxes Derived for ELMs, Disruptions and Other (Events) - Chuck Kessel stressed the distinction that ELMs are transient events, whereas disruptions are off-normal events. The expected ELMs are periodic, transient, short duration events and their heat flux is additive to the steady-state heat loads on the divertor. The PFC materials must withstand the additive steady-state heat loads and the cyclic power loads. Chuck presented a chart displaying the properties of CFC and W to withstand the cyclic heating scenario. He also provided some other background information for consideration and asked the question "How will ARIES contribute to this field?"

TSC Time-Dependent Free-Boundary Simulation of the ACT1 Plasma and Disruptions - Chuck Kessel presented the key parameters for one of the ARIES ACT1 aggressive physics operating points used for the current analysis, while suggesting that others may be better. The TSC code provides some insight into the predicted current, temperature and density temporal histories. Future work may include new parameter space investigations, corrections for the ASC systems code, examine other profile and transport models, MHD stability investigations, disruption simulations and examination of the ACT2 cases.

Chuck provided a sequence of plasma conditions during a VDE, showing the movement of the plasma and related VDE events. He also examined a major (or midplane) disruption (MD) event and its time, spatial and current histories. He suggested how to begin to mechanically model the disruption event parameters for the physical power core structural and EM analyses. He also suggested approaches to address the thermal modeling of the power core. Chuck outlined future work that could be done to further address these issues.

Progress on Optimization Results for Steady-State Plasma Solutions - Holger St. John discussed his progress in the ARIES-AT physics modeling. He reported on his parametric density scans over ranges of ßN pedestal values. He has done pedestal stability investigations using the EPED code. With a ßN of 5.7, he found that the pedestal pressure can be increased to 1.8 kPa with the density, ρped, equal to 0.90 [ρpedis dimensionless, normalized to 1.0 at the separatrix}. The EPED1 code can predict a pedestal height and width given a sufficient set of plasma physics input parameters. Holger outlined a set of transport grid requirements that can be satisfied by using higher order derivatives and non-uniform adaptive grids. He noted that the turbulent transport conditions are sensitive to the order of interpolation schemes and may need higher order methods for solutions. The Glf/Tglf solver uses Jacobian free methods to speed flux computations. Steady-state current drive solutions are found using Faraday's law. He assumes the inductive current is proportional to the inverse of resistivity. Holger discussed the density scaling he used. The base case parameters for ARIES were provided yielding the density and temperature profiles. He noted the large bootstrap current near the pedestal overdrives the plasma. Significant density peaking is not thought to be beneficial. The prior results were obtained with the Glf code. Holger also showed some electron and ion density and temperature results using the Tglf code.

He summarized that the Glf confinement suggested the ARIES ßn of 5 plasma scheme is acceptable from the physics perspective. The ARIES-AT type of plasma with ßn-ped = 1 will require a higher density than currently proposed and/or (the use of?) auxiliary heating. Initial pedestal stability results indicate that ßn-ped between 1.0 and 1.1 @ ρ = 0.9 is satisfactory. Good alignment of total current with bootstrap current may be possible, but on-axis current drive is needed.

Action Items

During the meeting, Les Waganer transcribed the action items and provided the list to Farrokh Najmabadi for documentation. Farrokh will refine the list and publish it as soon as possible. It is important to complete these action items so others on the team can accomplish their work in a timely manner.

The next conference call will be in a few weeks as determined by Farrokh Najmabadi and Mark Tillack. The next meeting will be in roughly 3 months, dependent on the progress of the team results.

Action Item list shown on the following page.