Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
I. I NTRODUCTION
Fig. 3. Side grid in the TF He inlet mock-up tested at CEA. The inlet pipe
carrying the coolant flow directly to the other grid is also visible.
In this paper, building on previous experience made on Fig. 4. Two limiting situations considered in this paper for the direct passage
the analysis of the He inlets for the ITER superconducting of the coolant from the inlet pipe (white circle) to the cable: unobstructed
Case A (top); largely plugged by the petal wrappingCase B (bottom).
central solenoid [4], we develop and apply a computational fluid Dark blue = wrapping; light colors = petals.
dynamics (CFD) model to compute the pressure drop versus
mass flow rate characteristic in the TF inlet mock-up recently
previous work showed that the two-layer k model was
tested at CEA Cadarache, France.
very good at reproducing the hydraulic characteristic [8], but
The results of the calculation will be validated against
that model is not available in OpenFoam; therefore, we chose
the available measurements. The thus validated computational
here the k RNG turbulence closure as an alternative, which
model will be finally used to compute the distribution of the
is known to be also suitable in cases of flow detachment/
coolant flow among the different petals, which was not mea-
reattachment, as typical of the central channel flow close to the
sured, as well as the sensitivity of the localized pressure drop at
spiral.
the inlet to the presence of the petal wrapping.
All simulations are performed imposing a given mass flow
rate at the inlet, and a given reference pressure (0 Pa gauge) at
II. TF H ELIUM I NLET D ESIGN G EOMETRY the outlet.
The TF cable is made of six petals twisted with a pitch of
420 mm around the central channel spiral (8 mm/10 mm IV. M ODEL G EOMETRY AND M ESH G ENERATION
ID/OD, 6 mm width, 33% open fraction). Each petal is partly
covered by a helical wrapping (0.1 mm thickness, 13.5 mm A. Cable
width, 50% open fraction). With respect to the petal helices, In the cable model, the azimuthal position of the petals
both the central channel spiral and the wrappings are wound in (phase of the cable) is chosen randomly at one end of the
the opposite direction. conductor, as well as the phase of the wrappings with respect to
The inlet sleeve is applied after suitably cutting away two the petals. The cable is finally inserted in the jacket to obtain
parts of the jacket, and substituting them with two perforated the CICC.
grids, see Fig. 2. The axial length of the computational domain needed for the
While in the actual ITER TF inlet, the cable wrapping will inlet simulations is Ldomain = 2 600 mm (centered around
be removed from the grid regions, it was left in place in the TF the inlet pipe axis). This was checked a posteriori, see below,
inlet mock-up analyzed in this paper, see Fig. 3. to be sufficient for the full development of the flow in the CICC.
TABLE I
N UMBER OF C ELLS (/106 ) IN THE D IFFERENT M ESHES B UILT FOR THE
TF S TRAIGHT C ONDUCTORa
A. Calibration of the Bundle Region Permeability The collection of the available experimental data in pseudo-
dimensionless form is reported in Fig. 8. The data from [2]
The hydraulic characteristic of the bundle region of a short
and [3] cover a range of relatively low Re , which is not fully
straight TF conductor sample was measured at CERN using
representative of the nominal flow conditions in the ITER TF
pressurized water [2] and keeping the central channel plugged.
coils, expected to operate at Re 2100 m, which is well
The parabolic fit of the experimental data (average pressure
covered, however, by the data from [9]. The error bar for the
gradient versus mass flow rate) can be used to derive the values
data set in [9] is not negligible, while the error bar for the other
of the permeability K (assumed isotropic at this stage) and of
two data sets [2] and [3] is unfortunately unknown.
the inertia constant J, using the following relation:
The simulations have been performed on the different meshes
p J using water as a fluid and under two different assumptions:
= m m2 . (1)
L KA KA2 i) isotropic porous medium;
4202405 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 24, NO. 3, JUNE 2014
ACKNOWLEDGMENT
We should like to gratefully acknowledge the compu-
tational resources provided by HPC@POLITO, a project
of Academic Computing within the Department of Con-
trol and Computer Engineering at the Politecnico di Torino
(http://www.hpc.polito.it), and by JuRoPA-Jlich Research on
Petaflop Architectures/HPC-FF-High Performance Comput-
ing for Fusion (http://www.fz-juelich.de/ias/jsc/EN/Expertise/
Super-computers\/JUROPA/JUROPA_node.html).
R EFERENCES
[1] C. Boyer, K. Hamada, A. Foussat, M. Lerest, N. Mitchell, P. Decool, and
F. Savary, Development of helium inlet structure for the ITER toroidal
field coil, presented at the 23rd Int. Conf. Magnet Technol., Boston, MA,
USA, Jul. 2003.
[2] N. Dalexandro and F. Savary, ITER TF He Inlet. Pressure drop mea-
surements and development of a manufacturing process, CERN, Geneva,
Switzerland, Rep. EDMS 123456 v1.0, 2012.
[3] K. Hamada, K. Kawano, N. Ebisawa, H. Nakajima, Y. Yano, and
T. Yamaguchi, Pressure drop characteristic of the ITER cable-in-conduit
conductor, in Proc. ICEC 24-ICMC 2012, 2012, pp. 559562.
[4] R. Zanino, N. Martovetsky, A. Pasquali, L. Savoldi Richard, and
D. Speziani, Computational thermal-hydraulic analysis of the helium in-
let options for the ITER central solenoid, IEEE Trans. Appl. Supercond.,
vol. 22, no. 3, p. 4902505, Jun. 2012.
[5] [Online]. Available: http://www.openfoam.org/
Fig. 12. Distribution of the mass flow rate (positive if directed to the right for [6] R. Zanino, S. Giors, and R. Mondino, CFD modeling of ITER cable-in-
x > 0, if directed to the left for x < 0) below the inlet and along the conductor: conduit super-conductors. Part VI: Heat and mass transfer between cable
central channel (dashed line), different petals (solid lines of different colors, region and central channel, Cryogenics, vol. 50, no. 3, pp. 158166,
color code explained in the sketch on top). Anisotropic porous medium. Mar. 2010.
[7] R. Zanino and S. Giors, CFD modeling of ITER cable-in-conduit super-
conductors. Part IV: Friction factor of the simplest (triplet) bundle, in
The computed distribution along the conductor of the flow Proc. 21th ICEC, 2007, vol. 1, pp. 201206.
repartition on the CICC cross section is reported in Fig. 12: [8] R. Zanino, S. Giors, and R. Mondino, CFD modeling of ITER cable-in-
conduit super-conductors. Part I: Friction in the central channel, in Proc.
the homogenization length among the petals results to be Adv. Cryogenics Eng., 2006, vol. 51, pp. 10091016.
200300 mm. The flow repartition between bundle region (B) [9] P. Decool, M. Tena, and H. Cloez, Pressure drop measurements on TF
and central channel (H), reaches after 400500 mm from the conductor and He Inlet, CEA Report 2013.
[10] R. Zanino and L. Savoldi Richard, A review of thermal-hydraulic is-
inlet the asymptotic value of 48% (B) and 52% (H). The two sues in ITER cable-in-conduit conductors, Cryogenics, vol. 46, no. 7/8,
geometries considered here for the inlet/cable interface (Case A pp. 541555, Jul./Aug. 2006.