CHATS on Applied Superconductivity 2011

12 Oct 2011, 08:00 → 14 Oct 2011, 18:00 Europe/Zurich

Salle Andersson (CERN)

Workshop at CERN, Geneva, Switzerland October 12 to October 14, 2011

Slides CHATS-AS_2011_1st_circular.pdf TOUR_for_CHATS_final.pdf

Wednesday 12 October

08:15 → 12:30 Session 1 - Heat Transfer: Heat Transfer

Session 1 - Heat Transfer

08:30 Introduction

Speaker: Frédérick Bordry (CERN)

08:45 Modeling of a horizontal circulation open loop in two-phase helium

Two-phase circulation loop are often used to cool large superconducting magnets since it has the advantage to eliminate the use of pressurization system such as pumps which dissipate power on the fluid and require a costly maintenance and operation at these temperatures. Most of these loops are designed for vertically oriented system to take advantage of the pressure head to create a stable flow. The superconducting magnet of the R3B spectrometer is horizontally oriented and will be cooled with a circulation loop with quasi-horizontal heat exchanger tubes. To assist the design of such cooling scheme, a horizontal loop has been tested experimentally and a predictive numerical model has been developed. In this paper we present the thermo-hydraulics modeling of such a flow configuration, and the comparison with experimental results.

Speaker: Bertrand Baudouy (CEA Saclay)

Slides CHATS_Baudouy.pdf

09:15 Heat Transfer through Cable Insulation of Nb-Ti Superconducting Magnets Operating in He II

The operation of superconducting magnets with Nb-Ti conductors in helium II allows profiting of superfluidity to efficiently transfer heat from cable to helium bath. In wrapped cable insulations, like the ones used for the main magnets of the Large Hadron Collider particle accelerator, part of heat is transferred through the bulk of the dielectric insulation and part through micro-channels between the insulation layers. In this paper, available experimental data of polyimide insulation schemes are analyzed in terms of the relevant wrapping parameters, to compare their heat transfer efficiency. The path of the heat fluxes in the superconducting coil cross-section is determined, and the corresponding steady-state temperature distribution is evaluated.

Speaker: Mr Pier Paolo Granieri (CERN, Geneva & EPFL Lausanne, Switzerland)

Slides Granieri_CHATS-2011_Heat_transfer_Cable_Insulation_Nb-Ti_SC_Magnets_He_II_v2.pdf

09:45 Numerical study of the thermal behavior of a Nb3Sn high field magnet in He II

The High Field Magnet (HFM) project, within the European project EuCARD, aims at constructing an Nb3Sn high field accelerator magnet, Fresca 2, to serve as a test bed for future high field magnets and to upgrade the vertical CERN cable test facility. The Fresca 2 block coil type magnet will be operated at 1.9 K or 4.2 K and is designed to produce about 13.5 T. To study the thermal behavior of the magnet in He II, i.e. to calculate the temperature margin and the evolution of temperature due to quench within the magnet a simplified 3D numerical model was developed. The model is derived from the original two-fluid model and consisted of a conventional continuity equation, a modified momentum equation for the total fluid and an energy equation including the Gorter-Mellink internal convection term. The paper presents the numerical model, the assumptions taken for the calculations and several results of the simulation for the quench heating and temperature distributions due to several cases of heat loads.

Speaker: Dr Slawomir Pietrowicz (CEA)

Slides Modeling_of_magnet.pdf Modeling_of_magnet.pptx

10:15 Coffee Break

10:30 Implementation of the superfluid helium phase transition using 3D FE modeling: simulations of hot spot burnout, recovery and phase front movement

In the thermal design of certain superconducting accelerator magnets, the emphasis is on the use of superfluid helium as a coolant. The very large effective thermal conductivity of helium below the lambda transition helps to extract heat from the coil windings during steady state and transient heat depositions. The geometry and size of the helium channels have a strong influence on the temperature distribution in the magnet. To better understand the behavior of superfluid helium penetrating the magnet structure, 3D FE simulations can give valuable insight. Not only the helium bulk behavior is of interest, but especially the strong non-linear behavior at the interface between solid and superfluid (Kapitza conductance) is important from an engineering point of view, since relatively large temperature jumps may occur here. This work shows how superfluid helium behavior in magnet windings can be simulated using COMSOL Multiphysics. The code is validated with experimental results taken from literature. These include burnout (exceeding the critical heat flux), recovery and subcooling of long channels filled with normal liquid helium (phase front movement). Special attention will be given to the various solid-liquid and solid-solid interfaces, which can be found in accelerator magnets.

Speaker: Mr Erwin Bielert (CERN)

Slides CHATS_Bielert.pdf

11:00 Thermo-hydraulic simulation of the ITER PF coil joints based on their coupling losses calculated with JackPot-AC

The paper describes the results from the new thermo-hydraulic model for cable-in-conduit conductor (CICC) joints, which has been specifically developed to analyse the temperature margin of ITER PF coil joints. The heat generation is calculated with the electromagnetic model JackPot-AC, which can calculate the coupling losses in such a CICC joint on a strand-level. Because of this high-level of detail in JackPot’s output, the temperature is calculated for the six final-stage sub-cables (petals) individually, instead of assuming only one homogeneous cross-sectional temperature. This allows the simulation of localised power input from the copper sole into the petals only if they are in contact with each other. The geometry calculated with the JackPot-AC joint model is used to determine this, as well as the wetted perimeter of the sole and the contact area between the petals and the sole. The cross-sectional temperature in the helium, sub-cable strand bundles and copper sole is assumed homogeneous; however they can all vary along the length of the cable. The helium paths in the petals are thermally connected to each other through the petal wraps, which is at the same time the only thermal exchange that can take place between the petals. The results show indeed a small variation of cross-sectional temperatures among the petals in an ITER plasma scenario, but due to the strong dependence of the critical current of NbTi strands -which are used in the PF conductors- on temperature, this can have a large impact on the temperature margin of the joints.

Speaker: Ezra van Lanen (University of Twente)

Slides CHATS2011_Vanlanen_Thermo.pdf CHATS2011_Vanlanen_Thermo.ppt

11:30 Parametric analysis of pressure drop and heat transfer in the meander-flow heat exchanger of HTS current leads for fusion applications

The Karlsruhe Institute of Technology is responsible for the design, construction and testing of the high temperature superconductor (HTS) current leads (CL) for the stellarator Wendelstein 7-X and for the tokamak JT-60SA. The HTS CL consists basically of an HTS module, carrying the current in the temperature range from 4.5 K to 60 K in series with a Cu bar provided with transverse circular fins, working as a heat exchanger (HX), covering the range from 60 K to 300 K. The HTS module is cooled by heat conduction, whereas the Cu bar and the fins are actively cooled by helium. The HX is of the meander-flow type. Helium flows between the fins and is forced to a cross flow with respect to the central bar. An important issue in the operation of the HTS CL is the optimization of the cooling power consumption, i.e. the minimization of the helium mass flow rate inside the HX that allows a stable operation in steady state conditions. For this purpose, the knowledge of the helium thermal-hydraulics in such meander geometry becomes rather important. In this paper the procedure introduced and validated in [1-2] is extended to a systematic computational fluid dynamics (CFD) analysis of the helium thermal-fluid dynamics inside this kind of HX. The aim is to obtain a complete picture of the flow regime in a wide range of operative conditions and the dependence of pressure drop and heat transfer on the geometrical parameters of the HX. In the paper it is shown that the He flow inside the HX changes from laminar to turbulent and that the variation of geometrical parameters affects differently the pressure drop and the heat transfer process inside the HX. [1] L. Savoldi Richard et al., IEEE Trans. on Appl. Supercond., Vol. 20, No. 3 (2010), pp.1733-1736 [2] E. Rizzo et al., presented at SOFT-21, Porto, PT, accepted for publication in Fus. Eng. Des.

Speaker: Mr Enrico Rizzo (Karlsruhe Institute of Technology)

Slides Rizzo_CHATS_AS_def.pdf Rizzo_CHATS_AS_def.ppt

12:00 Parametric study for the cooling of HTS current leads

The analysis of cooling of a binary HTS 20 kA current lead (CL) operating between 4.5 and 300 K has been carried out. Assuming that the HTS module is conduction cooled, two cooling options for the copper heat exchanger (HEX) part of the CL have been considered, i.e. (1) cooling with a single flow of gaseous helium and (2) cooling with two flows of gaseous helium. The ideal refrigerator power required to cool a whole HTS current lead has been calculated for both cooling scenarios and different values of input parameters and the thermodynamic optimization for both cooling options have been performed. In addition, a preliminary analysis of the 3rd cooling option, in which both HTS and HEX parts of the CL are gas cooled, has been carried out. The obtained results indicate that the cooling options (2) and (3) cannot provide significant savings of the refrigerator power, as compared to the classical option (1). However, options (2) and (3) can provide other advantages over the option (1). It has been observed that at the same helium inlet temperature the temperature at the warm end of the HTS part, and the resulting number of HTS tapes, can be reduced in the option (2) with respect to the option (1). The option (3) appears to be recommendable in cases when high heat load in the HTS part due to AC losses, which cannot be effectively removed by heat conduction only, is expected.

Speaker: Dr Monika Lewandowska (Institute of Physics; Faculty of Mechanical Engineering and Mechatronics; West Pomeranian University of Technology, Szczecin, Poland)

Slides HTS-CLs_CHATS-2011.pdf HTS-CLs_CHATS-2011.ppt

12:30 → 14:00 Lunch Break

14:00 → 14:30 Session 1 - Heat Transfer

Session 1 - Heat Transfer

14:00 Modeling of heat transfer in an accelerator superconducting coil with a ceramic insulation in supercritical helium

One of the key issues in the operation of a superconducting particle accelerator is the temperature margin of the magnets exposed to heat losses. The temperature margins of the superconducting magnet coils are mainly determined by the heat transfer through the electrical insulation wrapped around the Rutherford-type cables. We have tested the thermal properties of a mock-up representing an accelerator magnet coil with a porous ceramic material in stagnant supercritical helium in steady state condition. In this paper we present the 3D thermal modeling of the experimental mock-up to identify the main heat transfer process and the influence of the geometrical parameters and thermal properties.

Speaker: Slawomir Pietrowicz (CEA Saclay)

Slides ceramic_insulation_1.pdf ceramic_insulation_1.pptx

14:30 → 17:45 Session 2 - Stability and Quench: Stability and Quench

Session 2 - Stability and Quench

14:30 A novel model for Minimum Quench Energy calculation of impregnated Nb3Sn cables and verification on real conductors

In light of the ongoing conductor research at CERN for the planned LHC upgrades, minimum quench energy (MQE) of the latest Nb3Sn conductors is investigated using spot heaters. These conductors are used in high field accelerator magnets, which are prone to training quenches emanating from small energy depositions (wire movement, epoxy cracks) in the conductor during ramps. To improve the understanding of the thermal stability mechanisms, the experimental results are compared to a novel numerical model of an impregnated Rutherford cable. The model can then be used to extrapolate the values of the MQE to ranges that are not measurable due to practical limitations. Additionally, the model can be used to predict the optimal set of parameters to attain maximum stability, decreasing the number of training quenches. This paper will describe the numerical model and the concepts behind it. It is a transient simulation, taking into account the heat and current flows and the local properties of the conductor depending on the temperature. Aside from an external heat pulse to simulate the energy deposition, the model assumes the cable to be perfectly adiabatic. The size of the heat pulse is varied over several simulation runs to find the transition point where the conductor can recover or not. The results of the model are then compared to actual MQE measurements and will be used to map the thermal stability margin in the cross section of a coil that is build with this cable.

Speaker: Mr Willem Michiel De Rapper (CERN)

Slides CHATS_2011_Oral_WMdeRapper_Final.pps CHATS_2011_Oral_WMdeRapper_Final.ppsx

15:00 Electro-thermal simulations of the 13kA shunted LHC joints

A defective bus connection between two dipole magnets was the primary cause of the 2008 September incident in the LHC. This led to the decision that all ten thousand dipole and quadropole bus joints will be improved to avoid the reoccurrence of such an event. The purpose of the present work is to support the design choices and quality control of the shunts. In order to achieve high reliability of the implemented solution, worst case conditions are assumed for the supporting simulations. Adiabatic thermal boundaries and conservative RRR values of the copper elements of the bus are used. Variation of the soldering and thus of the electrical and thermal contacts in the interconnect is the most important design parameter. Partial soldering to the bus is presumed, and the condition that the shunt needs to carry the full current of the circuit during a discharge of the circuit through a dump resistor is applied. The time constant, tau, of the discharge is determined by the inductance of the circuit and the value of the dump resistor. For the dipole circuits tau equals 100 seconds and for the quadropole circuits 30 seconds. Safe current as a function of the shunt design parameters is analyzed in this work. Also the influence of a partly soldered shunt is analyzed, where the percentage of soldered area is a function for the quality control. Safe operating current is defined by the condition that during the discharge of the circuit the temperature of the bus remains below 300K. For the calculations finite element analysis (FEA) is used, with a commercial software Comsol Multiphysics. Comparisons were also performed with a one dimensional model, called QP3.

Speaker: Daniel Molnar (CERN)

Slides Electro_Thermal_CHATS2011_.pptx

15:30 Stability Analysis of the Interconnection of the LHC Main Superconducting Bus Bars

The operation of the Large Hadron Collider (LHC) calls for a better understanding of the thermo-electric behavior of the 13 kA superconducting bus bars interconnecting its main magnets. A deep insight of the stability of the bus bars interconnections is required to ensure the protection of the accelerator against undesired effects of resistive transitions. This is especially important in case of defective interconnections which can jeopardize the operation of the whole LHC. This paper presents a numerical analysis of the interconnections between the main dipole and quadrupole magnets, aiming at determining the thermal stability and its dependence on the heat transfer towards the cooling helium bath, quality of manufacturing, operating conditions and protection system parameters. Defective interconnections featuring a lack of bonding among the superconducting cables and the copper stabilizer components are considered, as those present in the machine. The critical defect length limiting the maximum stable current for powering the magnets is evaluated. This computation is a contribution to the evaluation of the maximum allowed beams collision energy.

Speaker: Mr Pier Paolo Granieri (CERN, Geneva & EPFL Lausanne, Switzerland)

Slides Granieri-CHATS-2011-Stability_analysis_IC_LHC_main_SC_bus_bars.pdf Granieri-CHATS-2011-Stability_analysis_IC_LHC_main_SC_bus_bars.pptx

16:00 Coffee Break

16:15 Quench of ITER poloidal field coils: influence of some initiation parameters on thermo-hydraulic detection signals and main impact on cryogenic system

The 6 ITER Poloidal Field Coils are wound in double pancakes (DP) using two-in-hand large NbTi Cable-In-Conduit Conductors (CICC) with a central channel. All pancakes are cooled in parallel. The helium inlets are located at the innermost turns whereas the helium outlet (in the electrical joints) are locate at the outer radius. For a each PF coil winding, a simplified rectangular cross-section is assumed in the modeling, considering an integer number of turns in the radial (Nr) direction and an integer number of pancakes vertically (Nz). In case of a magnet quench, the primary quench detection system is classically based on voltage detection. In addition, a safety related secondary quench detection is required and this could rely on signals of thermo-hydraulic nature. A model based on the coupled GANDALF and FLOWER codes already partly presented in a previous papers for Central Solenoid and Toroidal Field Coils have been developed: This model represents one pancake of one PF coil modeled by Gandalf and the other pancakes and other PF coils modeled with Flower with compressible heated channels (as well as the inlet and outlet channel in the feeders). This study focuses on two PF Coils, namely PF3 and PF5: - PF3 has the longest hydraulic length and a relative high temperature margin (of nearly 2 K); there is very little difference in the temperature margin between the top, middle or bottom pancake. For the median pancake (with a conductor temperature of 5.1 K), one short heated zone of between 1 m and 10 m, at the middle of conductor, will be the most difficult to detect, except with the signal difference of mass flow rates (i.e. inlet – outlet) in the feeder at the Cold Termination Box (CTB), where the sensors are located. - PF5 has the higher magnetic field value along the conductor, and a high nuclear heat load (during plasma) on the top pancake (due also to the steady state support heat loads). The top pancake has then the lowest temperature margin, and for this case the study of a “big quench” is performed, that means with all the pancakes quenching at the same time, at the inlet and outlet of the coil. In this case the cryogenic consequences are presented, specially the pressure inside the conductor and upstream the relief valves. Results of some parametric studies are presented and the influence of the initiation parameters of the quench (specially quench energy, location and length of heat deposition,, and time duration). The possibility of a secondary thermohydraulical detection and the main impact on cryogenic system are analyzed and discussed.

Speaker: Sylvie Nicollet (CEA/IRFM Saint-Paul-lez-Durance 13108 France)

Slides 111012_CHATS_S.Nicollet_Pres.pdf 111012_CHATS_S.Nicollet_Pres.ppt

16:45 Integrated quench protection model for Nb3Sn high field accelerator magnets

Reliable quench protection of long high-field Nb3Sn accelerator magnets requires optimization of the present state-of-the-art protection heater technology. The heater geometry and materials must be designed to provide enough multiple normal zones along the winding to ensure fast and uniform dissipation of the stored magnetic energy. At the same time, voltages in the windings and heater temperatures must be kept at an acceptable low value. An integrated design tool is being developed for this purpose as part of the high-field Nb3Sn magnet modeling effort at LBNL. The model is based on numerical computation of thermal heat transport in the cables, which is used to analyze the coil temperature and voltage for different quench origins and heater layouts. In order to have a flexible multi-physics tool adaptable to various magnet and protection schemes, a modular approach is chosen where the coil geometry and magnetic field map are imported from external sources and coupled with the core program for the thermal and electrical computation. This initial stage of the model development relies on simplifying assumptions of the quench initiation and propagation. The principles of theory and development status of the code as well as benchmarking with existing quench computation codes regarding hot spot temperature and MIITs computation are presented.

Speaker: Ms Tiina Salmi (Lawrence Berkeley National Laboratory)

Slides Salmi_CHATS_seminar_Oct2011.pdf Salmi_CHATS_slides2_Oct2011.pptx

17:15 Integrated analysis of quench propagation in a system of magnetically coupled superconducting coils

The design of the quench protection system for a superconducting solenoid for laboratory applications is a practical matter that has often been addressed by using simplified models based on analytical estimates for the quench propagation velocities in the longitudinal (along the conductor) and transverse (from turn to turn) directions. This approach can become cumbersome, especially for coil configurations such as segmented magnets or multiple coils with inductive coupling. We demonstrate in this paper how to use direct modeling of the heat transfer and circuital equations to produce a model suitable for such configurations, but producing results based on propagation velocities which are consistent with the physics of the phenomena.  To this aim we use an equivalent thermal 3-D model, based on a 1-D longitudinal model coupled with a simplified 2-D transversal model to solve the heat exchange among solids and coolant, and a circuital description of the coupled system of magnets. The building blocks are the CryoSoft suite of dedicated codes for the analysis of superconducting coils, i.e. THEA, POWER and SUPERMAGNET, integrated by "ad hoc" problem-related routines.  Our application example is a system of three, layer-wound and bath-cooled coils, all using a NbTi conductor. Following a quench initiation, we demonstrate the calculation of the time evolution of currents, voltages and temperatures in all coils. The sensitivity of these results to input parameters such as the thermal conductivity of the insulation between turns and layers, the heat transfer coefficient to helium bath and the heat perturbation, is investigated in a parametric study. The same model can be easily adapted to study coils with different topologies (e.g. pancake winding), geometry (e.g. non-circular shapes) and conductors (e.g. force-flow cooled).

Speaker: Claudio Marinucci (EPFL-CRPP)

Slides CHATS2011.pdf CHATS2011.pptx

Thursday 13 October

08:30 → 09:30 Session 2 - Stability and Quench: Stability and Quench

Session 2 - Stability and Quench

08:30 Experimental simulation of helium pressure rise during a quench of a superconducting coil cooled by a superfluid helium bath

Experimental and numerical studies have been conducted with the aim of modeling pressure rises which occur in the helium, during quenches of the 11.7-T superconducting magnet named Iseult. Iseult is based on double pancake winding internally cooled by superfluid channels opening into a pressurized He II bath at 1.8 K. A scale mock-up has been built of 10 copper equivalent pancake slices and 7 helium channels per pancake. The heat produced by a quench of the Iseult magnet is simulated by electrical heaters put inside each copper plate. Cryogenic pressure and temperature sensors have been fitted in the helium channels and in the bath. Bath pressure measurements are given for various heating powers, various numbers of heated plates and various bath volumes. Different simple numerical models are proposed to simulate the pressure increases in helium. Comparisons with experimental measurements permit to identify the best model and the main physical mechanisms which drive the pressure rise during a quench.

Speaker: Dr Chantal MEURIS (CEA Saclay)

Slides Experimental simulation of helium pressure rise during a quench of a superconducting coil cooled by a superfluid helium bath

09:00 The current state of the art in comprehensive computer analysis of quench in pool-cooled superconducting multi-coil magnets at the NHMFL

The current state of the art in computer simulation of quench in pool-cooled superconducting magnets, particularly in those consisting of many coils/sections, is viewed. The discussion is mostly about up-to-date solenoid closely-packed epoxy-impregnated LTS high-current-density windings, albeit not only about them. The approach to computer modeling of a quench event is considered in comparison with the approaches by others: basic 2D and 3D models of the process of a normal zone propagation employed at present in the Applied Superconductivity area for quench simulation are reviewed and their practicality is discussed. Some examples of detailed and precise computer analysis of quench events in multi-coil superconducting magnets performed at the NHMFL in an effort to optimize the quench protection systems and the magnets’ designs themselves are presented. Methods to take appropriately into account the physical processes that govern and/or affect a superconducting magnet quench behavior (e.g., AC loss generation, the stored energy dumping on diode-controlled shunt resistances, etc.) are dealt with in the work, too, with due regard for time-varying distributions (maps) of the magnetic field and, if required, those of the strain over the coils.

Speaker: Dr Andrey Gavrilin (National High Magnetic Field Laboratory of Florida State University, USA)

09:30 → 10:30 Session 3: Electromagnetics, Models and Methods: Electromagnetics, Models and Methods

Session 3: Electromagnetics, Models and Methods

09:30 Speeding up the Simulation of Coupling Losses in ITER size CICC joints by using a Multi-Level Fast Multipole Method and GPU technology

The twin-box pancake joints for the PF coils of ITER will be constructed by connecting overlapping sections of the cable-in-conduit conductors with copper soles. For the analysis of the stability of these joints in a plasma scenario, it is necessary to know how its coupling losses respond under different orientations of the background field. For this purpose, a new model has been developed based on JackPot, which can calculate the coupling losses in such a joint at strand level. It calculates the mutually induced voltages across strand and copper sole sections with Biot-Savart type equations which, due to the discretisation of the system, results in many tens of billions of mutual couplings. The Multi-Level Fast Multipole Method (MLFMM) has been implemented to calculate these couplings in an acceptable time. This method separates the system in groups of current carrying elements that are close to each other, for which the couplings are calculated directly, and in groups that are far away, in which case an approximation is used. Larger number of elements can be put into one group if they are farther away, for additional calculation efficiency. Since this method is highly suitable for parallel computation techniques, the code has been implemented for use on a Graphics Processing Unit (GPU). The paper describes the implementation of the MLFMM in JackPot, and demonstrates the speedup that is acquired with it, at an acceptable tolerance. It also describes the results from coupling loss calculations on ITER PF pancake joints.

Speaker: Ezra van Lanen (University of Twente)

Slides CHATS2011_VanLanen_MLFMM.pdf CHATS2011_VanLanen_MLFMM.ppt

10:00 Development of a three-dimensional finite-element model for high-temperature superconductors based on the H-formulation

Finite-element models are a powerful and widely used tool for evaluating the ac losses of HTS tapes and wires as well as of assemblies such as cables and coils. The H-formulation, which uses the magnetic field components as state variables, has proved to be an efficient implementation to solve 2-D problems, involving infinitely long or axially-symmetric geometries. However, the simulation of certain applications requires a fully 3-D model. In this paper we report on the development of a 3-D model based on the H-formulation. We describe in detail the implementation of Maxwell’s equations, the imposition of current constraints and we discuss the issues related to meshing 3-D volumes. The obtained results are compared with analytical expressions based on the critical state model, if available, or with the results obtained with 2-D representations of 3-D problems

Speaker: Francesco Grilli (Karlsruhe Institute of Technology)

Slides grilli_chats_2011.ppt

10:30 → 10:45 Coffee break

10:45 → 12:00 Session 3: Electromagnetics, Models and Methods: Electromagnetics, Models and Methods

Session 3: Electromagnetics, Models and Methods

10:45 Usage of manifolds in electromagnetism and superconductor modelling

Many modelling software packages utilise predetermined coordinate systems and readily embedded metric. However, sometimes these coordinate systems are not practical for modelling or they present too restricted view of the real world. Manifolds and differential geometry offer a more general framework for modelling than coordinate systems can do. For computations we always need some charts, i.e. coordinate systems, of the manifold, but the charts need not be the ones already presented in textbooks. In this paper we consider new modelling possibilities that usage of manifolds offers in electromagnetic modelling of superconductors.

Speaker: Dr Antti Aleksis Stenvall (Tampere University of Technology)

Slides AnimationSymmetry.avi Stenvall-Chats-2011.pdf

11:15 How to model a large and complex apparatus with high precision - from

Today's particle detectors for high-energy physics are >10m in size and have <100 micrometer spatial resolution. This is for the detector elements which measure particle momenta by curvature in magnetic fields. Other detector elements do measurements by absorbing the energy of particles. Large size paired with fine resolution are due to the very high energy of the elementary particles observed in the detector. These particles emerge from the interactions happening in the center of the detector and which are of primary physics interest. We will have a look at how a particle detector such as ATLAS, built of many elements of complex shapes, is described in terms of relatively few geometrical elements which yet allow the path of particles through the detector to be reconstructed with full precision. The magnetic fields which bend the particle paths are a complication in reconstruction. Coordinate transformations along the path help. All this requires a large amount of computing software and apparatus. Finally the interactions of interest are analyzed, using the particles measured in the detector as well as simulation of the interactions, thus verifying or falsifying existing physics theories or better, developing fresh ideas.

Speaker: Hans Von Der Schmitt (Max-Planck-Institut fuer Physik (Werner-Heisenberg-Institut) (D)

Slides

• hvds-11-10-13-CHATS.pdf
• hvds-11-10-13-CHATS.pptx
• hvds-11-10-13-CHATS.pptx.pdf

12:00 → 12:15 Introduction to CERN Visits

Speaker: Mr Luca Bottura (CERN)

Slides Presentation_visitsCHATS-1.pdf Presentation_visitsCHATS-1.pptx

12:15 → 13:30 Lunch Break

13:30 → 18:30 Session 4: CERN Visits: CERN Visits

Session 4: CERN Visits

13:30 Visits

19:00 → 22:30 Banquet BRASSERIE DES HALLES DE L'ÎLE

BRASSERIE DES HALLES DE L'ÎLE
PLACE DE L'ÎLE 1
CH-1204 GENEVE
+41 22 311 08 88
buffets@brasseriedeshallesdelile.ch

Friday 14 October

08:30 → 09:30 Session 3: Electromagnetics, Models and Methods

08:30 Model for electromagnetic field analyses of assemblies of coated conductor considering their three-dimensional geometries

Numerical electromagnetic field analyses are powerful tools to study the ac loss characteristics of superconductors. If they are combined with thermal analyses, they are also useful to study the quench characteristics of superconductors. Here, we focus on (RE)BCO coated conductors. Electromagnetic field analyses in the cross-section of a coated conductor have been made widely to study its ac loss characteristics, and the focus of interest are shifting to the assemblies of coated conductors such as cables or coils. However, cables or coils are generally with three-dimensional geometries, and their analyses are much more complicated than those of single coated conductors. A peculiarity of coated conductors from the viewpoint of numerical analyses is its extremely thin superconductor layer: its cross-sectional aspect ratio is 2000~10000. This leads to a difficulty in numerical modelings. If we keep a moderate cross-sectional aspect ratio for each element, a large number of elements must be generated in just one cross-section of the superconductor layer; the number of elements in an entire three-dimensional model should be too large for practical calculation. One of the approaches against this problem is a thin strip approximation where the magnetic field component tangential to the superconductor layer and the current density normal to it are neglected. Models for electromagnetic field analyses of superconducting Roebel cables as well as power transmission cables have been developed based on this approximation. The results of analyses by using the models to calculate ac losses are also presented. This work was supported in part by Japan Science and Technology Agency under Strategic Promotion of Innovative Research and Development Program.

Speaker: Prof. Naoyuki Amemiya (Kyoto University)

Slides CHATS-AS-2011-AmemiyaRv3.pdf

09:00 Upscaling Method Applied to the Study of Superconducting Magnet Thermo-hydraulics

In the field of applied superconductivity, there is a growing wish to better understand the collective behaviour of the thermo-hydraulics of superconducting magnets. Depending on the specific item to be addressed, either 0-D, 1-D, 2-D or 3-D modelling may be needed. Because of the size of these magnets, in isolation or coupled together, it is already for computational reasons alone not possible to study this numerically without any simplification in the description of the geometry and the physics. The main idea of this study is to consider the interior of a superconducting magnet as a porous medium and to apply methods used in the field of porous media to obtain the equations that model the thermo-hydraulic behaviour of a superconducting magnet in different configurations (steady-state, beam losses, quench,...) with minimal compromises to the physics and geometry. Since the interior of a superconducting magnet is made of coils, collars and yoke filled with liquid helium with channels that interconnect the helium inside the magnet, an upscaling method should provide models that describe the thermo-hydraulic behaviour at the magnet scale and are suitable for numerical studies.

Speaker: Dr Herve Allain (CERN)

Slides CHATS2011_Herve_Allain.pdf CHATS2011_Herve_Allain.pptx

09:30 → 12:45 Session 5: System analysis and Multi-physics: System analysis and Multi-physics

09:30 Modeling and simulation of cryogenic processes using EcosimPro

Differential and Algebraic Equations (DAE) are especially adapted to model and simulate thermo-hydraulic systems. CERN has developed a dedicated library to model helium cryogenic plants handling usual cryogenic components using DAE with a commercial software called EcosimPro. The aim of such a library is to provide a simple way to model large-scale cryogenic systems and perform dynamic simulations in an acceptable timescale to assist both, control and operation teams, in the optimal commissioning and operation of the cryogenic plants.Moreover, EcosimPro allows users to develop easily models related to their specific components such as cryogenic transfer lines or superconducting magnets. During the last years, this library has been used to model several CERN cryogenic systems, namely the CMS cryoplant, the CERN central helium liquefier, a LHC 18 kW @ 4.5 K refrigerator and a LHC 2.5 kW @ 1.8 K cooling unit. The simulations have been used for different purposes i.e. operator training, virtual commissioning of control systems and control optimization. This paper presents in detail the modeling of some basic components and shows simulation results compared with experimental data obtained on different CERN cryogenic plants.

Speaker: Dr Benjamin Bradu (CERN, EN department)

Slides CHATS-Bradu.pdf CHATS-Bradu.ppt

10:00 Validation of the cryogenic circuit module in the 4C code against SHe closed loop data

The Cryogenic Circuit Conductor and Coil (4C) code has been recently developed for the simulation of thermal-hydraulic transients in the ITER magnets and cryogenic circuit [1]. The code validation was started against different types of transients ranging from (fast) safety discharge [2] to (slow) cool-down [3], but it did not include, so far, the cryogenic circuit module. Very recently, the Cryogenics Modelica library has been developed by the authors as the basis for the 4C cryogenic circuit module, and the 4C results have been compared with those of a publicly available demo of the Vincenta code, showing very good qualitative agreement [4]. In this paper we validate the newly developed cryogenic circuit module of 4C against data collected in the Helios SHe cooling loop, at CEA Grenoble, France [5]. The main components of the Helios loop are: a cold circulator; heat exchangers to a saturated helium bath equipped with a resistive heater; pipes simulating the cooling channels of the ITER TF magnet casing and equipped with resistive heaters; control and bypass valves [6]. The computed evolution of temperature, pressure and mass flow rate at different circuit locations will be compared with the measurements. [1] L. Savoldi Richard, et al., "The 4C Code for the Cryogenic Circuit Conductor and Coil modeling in ITER", Cryogenics 50 (2010) 167-176. [2] R. Zanino, et al., “Validation of the 4C Thermal-Hydraulic Code against 25 kA Safety Discharge in the ITER Toroidal Field Model Coil (TFMC)”, to appear in IEEE Trans. Appl. Supercond. (2011). [3] R. Bonifetto, et al., "Modeling of W7-X superconducting coil cool-down using the 4C code", to appear in Fus. Eng. Des. (2011). [4] R. Bonifetto, et al., “Dynamic modeling of a SHe closed loop with the 4C code”, presented at Cryogenic Engineering Conference, Spokane (WA) USA, June 2011. [5] R. Vallcorba, et al., “ITER cryogenic system validation tests at Helios facility”, presented at Cryogenic Engineering Conference, Spokane (WA) USA, June 2011. [6] C. Hoa, et al., “Forced flow SHe in a closed heat transfer loop submitted to pulsed heat loads”, presented at Cryogenic Engineering Conference, Spokane (WA) USA, June 2011.

Speaker: Roberto Zanino (Politecnico di Torino)

Slides HELIOS_Chats2011_presentation_FINAL.pdf

10:30 Coffee Break

10:45 THELMA Analyses of ITER NbTi Cable-in-Conduit Conductors

The THELMA code, developed for the analysis of coupled electromagnetic and thermal-hydraulic transients in superconducting Cable-in-Conduit Conductors (CICC) [1, 2], has been recently applied to the analysis of one of the two legs of the Poloidal Field Conductor Insert Full Size Joint Sample (PFCI-FSJS), a NbTi CICC short sample tested in the SULTAN facility. THELMA was able to reproduce the premature and sudden nature of the cable resistive transition at high transport current, due to the combined effect of current, field and temperature non-uniformities on the CICC cross section, together with the presence of voltage spikes precursors of the transition, caused by fast current redistribution [3]. The same computational tool is now applied to a more general study of the influence that different parameters have on the NbTi normal transition, namely: 1) the amplitude of the transport current, 2) the presence or absence of sub-cable wrappings, 3) the length of the high-field region and its distance from the joint, 4) the type of joint. The above issues have been addressed over the years in different experiments, including the PFCI-FSJS, the PFCI [4], and the hairpin samples tested more recently in SULTAN [5]. These experiments provide the database against which the THELMA results will be compared. [1] M. Ciotti, A. Nijhuis, P. L. Ribani, L. Savoldi Richard, and R. Zanino, “THELMA code electromagnetic model of ITER superconducting cables and application to the ENEA stability experiment,” Supercond. Sci. Technol., vol. 19, pp. 987–997, Oct. 2006. [2] F. Bellina, P. Bettini, and F. Trevisan, “Electromagnetic analysis of superconducting cables and joints in transient regime”, IEEE Trans. Appl. Supercond., vol. 14, pp. 1356–1359, June 2004. [3] R. Zanino, F. Bellina, P. L. Ribani and L. Savoldi Richard, “Analysis of Sudden Quench of an ITER Superconducting NbTi Full-Size Short Sample Using the THELMA Code”, Supercond. Sci. Technol., 2011, submitted. [4] R. Zanino, M. Bagnasco, D. Ciazynski, B. Lacroix, E.P.A. Van Lanen, S. Nicollet, A. Nijhuis, L. Savoldi Richard, C. Sborchia, A. Torre, A. Vostner, L. Zani, “EU Contribution to the Test and Analysis of the ITER Poloidal Field Conductor Insert (PFCI) and Central Solenoid Model Coil”, Supercond. Sci. Technol. 22 (2009) 085006 (11pp). [5] B. Liu, Y. Wu, F. Long, and S. Li, “Test results and analyses of conductor short samples for China first PF conductor”, Cryogenics, vol. 51, pp. 90–94, Feb. 2011.

Speaker: Prof. Fabrizio Bellina (Udine University, Dipartimento di Ingegneria Ingegneria Elettrica, Gestionale e Meccanica)

Slides CHATS2011_Bellina.pdf CHATS2011_Bellina.ppt

11:15 0-D thermo hydraulic approach for predicting pressure and temperature along HELIOS SHe closed loop under pulsed loads.

Cryogenic systems for future large superconducting tokamaks (e.g. JT-60SA or ITER) are expected to cope with large pulsed heat loads due to cycling plasma operation. Their superconducting magnets are cooled down with forced flow supercritical helium. An experimental set up named HELIOS [1] has been designed and realized in CEA Grenoble to study pulsed load effects on a forced flow supercritical helium loop. It has been scaled down to reproduce the cooling scheme of the JT-60SA Central Solenoid loop, and some improvements were also realized in order to be relevant of ITER magnets cooling loop by adding some by-pass valves and circuits. The experiment is used both to test various smoothing load methods and to benchmark codes. Concerning this latter point, a certain number of codes have already been applied. Among them, some can solve differential and algebraic equations in time and space with appropriate corresponding mesh (Vincenta [2], 4C[3]., …) whereas others have a 0-D space approach (C_PREST[4],…). This 0-D space approach is also (as far as we know) used for supercritical flow in EcosimPro[5], a code that CEA/SBT has recently adopted and will applied to HELIOS simulations. Finally, it is important to mention that problems involving only derivative in time can benefit from Laplace transform which improves the calculation time and is also fully adapted to control process. So the motivation of this paper is to verify to what extent a 0-D thermo-hydraulic model can well reproduce in space and time, the variations of pressure and temperature along a supercritical helium closed loop, subjected to pulsed heat loads. The 0-D model is presented with the motivations of the assumptions taken. A typical scenario of pulsed heat loads applied in HELIOS (already used for Vincenta benchmark [2]) has been tested and the resulting variations of pressure and temperature have been compared with experimental data. The results of the 0-D model demonstrate the relevance of such approach for predicting transient behaviours in response of pulsed heat loads in a closed loop. This simple approach is also a justification to use process modelling codes where dynamics of the cryogenic circuits can be simulated with cryogenics components. [1] Investigations of pulsed heat loads on a forced flow supercritical helium loop, Part A: HELIOS experimental set up, C. Hoa et al to be published in Cryogenics [2] Investigations of pulsed heat loads on a forced flow supercritical helium loop, Part B: Simulation of the cryogenic circuit, R. Vallcorba et al to be published in Cryogenics [3] Validation of the cryogenic circuit module in the 4C code against SHe closed loop data, R. Zanino to be presented at this workshop [4] Private communication [5] Modelling, simulation and control of CERN cryogenic systems, B. Bradu Thesis

Speaker: Bernard Rousset

Slides presentation-BRfin_CH.pdf presentation-BRfin_CH.ppt

11:45 Simulation of the Operation of the ITER Coils using a Domain

The ITER magnet system has unprecedented complexity and stored energy, and its reliable operation will depend critically on the knowledge of the operating margins, and the optimal use of the large installed cooling capacity. We have tailored the system codes "glued" within the Supermagnet suite to perform simulations of the flow, temperature and operating margin during a sequence of plasma cycles as planned for the ITER CS, PF and TF coil systems. In this paper we describe the building blocks of the models of each coil system, the models assembly, and the motivation for the approximations taken. The approach taken is that of a splitting of the physical domain of simulation in a number of sub-domains coupled through a communication manager, i.e. a domain decomposition method. Examples of simulations and comparison to previous results are shown to demonstrate the overall functionality. Finally, we show how the model allows for detailed inspection of very detailed features in space and time.

Speaker: Luca Bottura (CERN)

Slides LB_DDM_chats-as_XI.pdf LB_DDM_chats-as_XI.ppt

12:15 4C analysis of thermal-hydraulic transients in the KSTAR superconducting magnet system

The KSTAR tokamak [1] is operating since 2008 at the National Fusion Research Institute in Korea. KSTAR is equipped with a full superconducting magnet system including the central solenoid (CS), which is made of 4 symmetric pairs of coils PF1L/U-…-PF4L/U, the poloidal field coils PF5L/U-…-PF7L/U and the toroidal field coils TF1-...-TF16. The CS coils, the PF5 coils and the TF coils are pancake wound using Nb3Sn cable-in-conduit conductors with a square Incoloy 908 jacket, while PF6 and PF7 are made with NbTi and a square 316LN jacket. All coils are cooled with supercritical He in forced circulation at nominal 4.5 K and 5.5 bar inlet conditions. The 4C code [2], already validated against different types of thermal-hydraulic (TH) transients in different superconducting coils [3], [4], has been recently applied to the analysis of the TH transient due to AC losses in the PF1L/U coils associated to a trapezoidal current scenario [5], confirming some inconsistency between the expected [6] and the measured [7] maximum temperature rise in the winding. In order to clarify this issue, in this paper we will present a broader TH analysis of the PF coils, including the PF5, where the same strand was used as in PF1, as well as different current scenarios. The computed results will be compared with the experimental results obtained during the KSTAR campaigns. [1] Y. K. Oh, et al., “Commissioning and initial operation of KSTAR superconducting tokamak”, Fusion Engineering and Design 84, 344 (2009). [2] L. Savoldi Richard, et al., “The 4C Code for the Cryogenic Circuit Conductor and Coil modeling in ITER", Cryogenics 50, 167 (2010). [3] R. Zanino, et al., “Validation of the 4C Thermal-Hydraulic Code against 25 kA Safety Discharge in the ITER Toroidal Field Model Coil (TFMC)”, IEEE Trans. Appl. Supercond. (2011) in press. [4] R. Bonifetto, et al., “Modeling of W7-X superconducting coil cool-down using the 4C code”, Fus. Eng. Des. (2011) in press. [5] R. Zanino, et al., “Simulation of thermal-hydraulic transients in the KSTAR PF1 coil using the 4C code”, to be presented at ICOPS-SOFE 2011, June 26-30, 2011, Chicago, USA. [6] H. J. Lee, et al., “The AC Loss Measurement of the KSTAR PF1 Coils During the First Commissioning”, IEEE Transactions on Applied Superconductivity, 20, 551 (2010). [7] Y. M. Park, et al., “Analysis of the Helium Behavior Due to AC Losses in the KSTAR Superconducting Coils”, IEEE Transactions on Applied Superconductivity, 20, 530 (2010).

Speaker: Laura Savoldi Richard (politecnico di torino)

Slides KSTAR_presentation_chats_final.pdf

12:45 → 14:00 Lunch

14:00 → 16:00 Session 6: Close-out: Close-out

14:00 Close Out