RHerzog

Simulationof the thermosyphon effect in dual channel CICC for ITER: 

Simulation of the thermosyphon effect in dual channel CICC for ITER Luca Bottura CERN Pierluigi Bruzzone Robert Herzog Claudio Marinucci EPFL CRPP

Outline: 

Outline Motivation for the study General features of the thermosyphon effect Experimental setup and measurement results Model, boundary conditions and setup of simulations Simulation results with key features Transfer of model to ITER geometry

Cooling of the ITER TF coils: 

Cooling of the ITER TF coils The ITER TF coils are planned to be made of Cable-in-Conduit Conductors (CICC), which are forced-flow cooled by supercritical liquid helium. In parts of the winding pack the helium flows clockwise, in the others counter-clockwise. In the vertical inner leg of the D-shaped coil there are thus conductors cooled by top-to-bottom helium flow. During ITER operation neutron radiation and AC losses generate heat in the winding pack; heat also flows in through the insulation and terminals. ~ 10 m He n n n n n n n n n n ~ 40 mm

Downward helium flow in a CICC with central channel: 

Downward helium flow in a CICC with central channel The central cooling channel serves to adapt the hydraulic impedance of the CICC to the desired cooling circuit. With realistic He mass flow rates the hydrostatic pressure is larger than the pressure drop caused by friction: g ρ dh > 2 f ρ v2 dh / Dh Downward He flow is therefore against the pressure gradient. Pressure variations over ~10 m length of conductor are much smaller than the operational pressure in the CICC.

Effect of heat on helium flow in CICC: 

Effect of heat on helium flow in CICC Heat creation and heat influx initially affect the ‘bundle’: Neutrons are primarily absorbed by the superconducting strands. AC losses generate heat in strands. Outside heat enters across the steel jacket (conduit). The heat leads to an increase of helium temperature and thus to a significant reduction of the helium density. T : 4.5 → 6.5 K  ρ : 148 → 120 g/l T bundle > T hole  ρ bundle < ρ hole Because of the hydraulic coupling p bundle = p hole. If ρ bundle is sufficiently small the hydrostatic pressure of the He in the bundle may be smaller than the pressure from the hole.  He flows upward in the bundle.  Thermosyphon effect

Thermosyphon effect with local heating: 

Thermosyphon effect with local heating In case of local heating substantial heat and helium flow occurs between bundle and hole. This is the configuration which we used with the TFAS2 sample to study the thermosyphon effect.

Results of the experiments on TFAS2 (1): 

Results of the experiments on TFAS2 (1)

Results of the experiments on TFAS2 (2): 

Results of the experiments on TFAS2 (2)

Simulations with THEA: 

Simulations with THEA THEA is a parametric program to simulate thermohydraulic effects in superconducting cables. The term “g ρ h” was recently added to the momentum equation to take account of gravity related effects, like the thermosyphon effect. For the simulation of TFAS2 we modelled the CICC conductor with: 2 ‘hydraulics’: the two coupled He channels ‘hole’ and ‘bundle’ 1 ‘thermal’: the sucon strands (heat capacity and heat conduction) For the He flow in the hole a constant friction factor ‘frictionfactor’ was used, for the bundle the Katheder model with the parameter ‘Dh’. Many parameters are determined by the geometry; among them the wetted perimeter of the strands and the wetted perimeter and the perforation between hole and bundle. Attempting to reproduce and understand the experiments we varied the three parameters frictionfactor (hole), Dh and HTC, the heat transfer coefficient between bundle and hole.

Boundary conditions: 

Boundary conditions We started the simulations varying the pressure drop between inlet and outlet to obtain the measured helium mass flow. In the interesting region around the onset of the thermosyphon effect this led to an instability: Eventually we modified THEA to run with imposed mass flows for bundle and hole at the inlet and constant pressure at the outlet. With constant inlet temperatures (4.5 K), realistic mass flow rates and heater power this led to consistently stable simulation runs.

Simulation results (1): 

Simulation results (1) Stepwise increase of heater power

Simulation results (2): 

Simulation results (2) Temperature profile bundle Temperature profile hole

Parameter variation: 

Parameter variation It is possible to reproduce the experimental data of the TFAS2 thermosyphon experiments. To obtain conditions for the thermosyphon effect the hole mass flow must be large! However, for the best fit the three parameters HTC (heat transfer coefficient hole-bundle), Dh (bundle) and friction factor (hole) assume values which are possibly unrealistic. Possible reasons are: An inadequate model for the heat (and mass) transfer between hole and bundle (see LB) Two He channels with internally homogeneous state variables may be insufficient

Implications for ITER: 

Implications for ITER 10 m vertical leg with 1 W/m heat load 5 bar He pressure, 8 g/s mass flow Temperature increase by about 0.4 K, but no thermosyphon effect. May be useful to study a complete TF coil and not only the innermost leg. Temperature profile bundle

Relax: 

Relax relax

SULTAN of Villigen (PSI): 

SULTAN of Villigen (PSI) SULTAN test station