Cryogenics 103 (2019) 102950

Contents lists available at ScienceDirect

Cryogenics
journal homepage: www.elsevier.com/locate/cryogenics

Research paper

Hydraulic analysis of liquid nitrogen flow through concentric annulus with T

corrugations for High Temperature Superconducting power cable
Ipsita Das , V.V Rao
⁎

Cryogenic Engineering Centre, Indian Institute of Technology, Kharagpur 721302, India

ARTICLE INFO ABSTRACT

Keywords: A hydraulic analysis for turbulent flow of subcooled liquid nitrogen through an annulus within the HTS cable is
HTS cable numerically investigated in terms of friction factor and flow pattern. The control volume for the investigation is
Annular flow confined to the annular space between outer smooth surface of cable core and inner surface of corrugated
Subcooled liquid nitrogen cryopipe. The corrugation geometries i.e pitch and depth affect the friction factor and thus the pressure drop and
Friction factor
pumping power of liquid nitrogen. Nine different geometries of corrugated cryostats with various combinations
CFD
of pitches (6, 10 and 14 mm) and depths (5, 7 and 9 mm) are considered and compared with smooth (non-
Corrugation pitch and depth
corrugated) pipe. The results are useful in selecting the cryostat configuration and the associated fluid pumping
system for cooling HTS power cables.

1. Introduction of corrugated flexible cryostat on frictional heat load needs a realistic

evaluation under various operating and geometric conditions. In open
One of the major concerns in recent days is the conservation of literature, there is some debate on how the corrugations in the inner
energy and the quest to develop any alternative solution to the existing cryostat wall of HTS cable affect the flow pattern and friction factor of
system, with minimum losses. High Temperature Superconducting LN2. Fuchino et al. [14] measured the pressure drop across a 10 m
(HTS) power cables can address this issue very well, as it has the ability corrugated pipe with different flow rates to evaluate the friction factor.
to transmit huge electrical power in compact size with minimum Joule The friction factor in corrugated pipes is larger than that in smooth
loss. Since the discovery of high temperature superconductivity [1], pipes. It is also reported that, the pressure drop in corrugated pipe is
many researchers have been working on different applications in power four times as high as that of smooth pipe [15]. Li et al.[16] have ex-
sector including HTS power cable. Moreover, various projects [2–9] on perimentally measured the flow resistance for both smooth and corru-
HTS cables worldwide have been successfully tested to introduce the gated pipes. They have reported that the friction factor in corrugated
technology into real grid operation. HTS cables require to operate pipes is three times that of smooth pipe. Chevtchenko et al. [17] have
below a certain temperature (critical temperature, Tc) in order to considered different cryostat sections (three straight and two corru-
maintain superconductivity. Subcooled liquid nitrogen (LN2) is mostly gated) with a dummy cable for their experiment to investigate the flow
used as coolant in HTS cable. Further, researchers have also adopted behavior of LN2 through different sections. Few researchers have nu-
some other possible alternatives as coolants. Demko and Hassenzahl merically analyzed the effect of corrugations on the friction factor of
[10] investigated with liquid air and Pamidi et al. [11] have used cold LN2 in case of corrugated pipes [18,19] and also analyzed the LN2 flow
helium gas, Kephart et al. [12] have conducted experiment with gas- behavior within the cryostat depending on the positioning of cable core
eous cryogen (air and nitrogen). Ashworth and Reagor [13] have used [20,21]. Ivanov et al. [22] have conducted an experiment considering
‘Distributed Cooling’ method for HTS cable. Further, various heat loads smooth cryo-pipe with cable in it, to measure the pressure drop along
like internal heat generation due to AC loss, heat-in-leak from the the length of pipe. Further, they have extended their research to finalize
surrounding and frictional heat loss are inherent in a HTS cable. The one optimum set of cryostats [23] from a combination of different sets
frictional heating is mainly dependent on flow velocity and is not so of cryostats (one for cable and another for LN2return line). Many re-
significant at small flow rates and in short length cables. However, with searchers [24,25] have developed correlations for friction factor con-
increase in cable length, frictional heat load increases, which is one of sidering the results available from experimental and numerical analysis.
the major issues of concern in operation of HTS cable. Hence, the effect Zajaczkowski et al. [26] have calculated the pressure drop along the

⁎
Corresponding author.
E-mail address: ipsitadas88@iitkgp.ac.in (I. Das).

https://doi.org/10.1016/j.cryogenics.2019.05.010
Received 11 December 2018; Received in revised form 21 May 2019; Accepted 28 May 2019
Available online 04 June 2019
0011-2275/ © 2019 Elsevier Ltd. All rights reserved.
I. Das and V.V. Rao Cryogenics 103 (2019) 102950

Nomenclature k corrugation depth (mm)

L length of pipe (mm)
A cross-sectional area of the flow (mm2) ΔP pressure drop (pa/m)
Di inner diameter of annulus (mm) Pw Wetted perimeter (mm)
dh hydraulic diameter(4A/Pw) (mm) p corrugation pitch (mm)
Do outer diameter of annulus (mm) v flow velocity (m/s)
Dm mean diameter (mm) ρ density of subcooled LN2(kg/m3)
fi friction factor of smooth wall τw wall shear stress (Pa)
fo friction factor of corrugated wall µ viscosity (Pa·s)
f friction factor of annulus.

HTS cable using the friction factor obtained from Colebrook–- 2.2. Operating range of LN2
White–Barr (CWB) formula [27] for inner pipe and friction factor ob-
tained from numerical result of Lee at al. [18] for outer annular duct. The subcooled LN2 is considered as coolant in HTS cable as its heat
However, an extensive study of literature shows that there is still some capacity is quite high and also to avoid two phase flow. The operating
uncertainty about the fact that how the corrugation dimensions affect the temperature range can be increased by increasing the pressure (Fig. 1).
flow behavior. The present analysis is focused on the effects of various However, the increase in operating temperature reduces the critical
realistic corrugated geometries (pitches and depths) on flow pattern and current. Thus the operating range of LN2 should be decided considering
friction factor in the fully developed turbulent flow region. The friction the electromagnetic properties of HTS tapes and thermodynamic
factor obtained in case of corrugated cryostat is further compared with the properties of LN2.
same of smooth cryostat. The calculations are made with constant prop-
erties of LN2 in subcooled region. The analysis is performed numerically 3. Flow model
using CFD fluent. The outcome of the present investigation has practical
significance, as the friction factor plays an important role in calculating the In the present work, it is considered that the cable core is housed
pressure drop and pumping power of LN2 flow in case of long distance within a flexible corrugated cryostat along its centerline, as shown
cooling of HTS power cables requiring intermittent re-cooling. Fig. 2(a). Subcooled LN2 flows through the annular space between outer
surface of cable core (smooth wall) and inner surface of corrugated
2. HTS cable geometry and operating range of LN2 cryostat (corrugated wall).The inner and outer diameter of annulus
considered for the present analysis is mentioned in Table 1 [21]. The
2.1. HTS cable geometry annular section of the cable is modeled with three different pitches and
three different depths (Table 1). Further, the analysis is also made for
A cold dielectric (CD) HTS cable is made out of copper former, HTS smooth cryostat without any corrugations.
conductor layer, dielectric (PPLP) layer, HTS and copper shield layers. The flow velocity of LN2 is assumed to be within 0.18 to 0.34 m/s
Further, all these layers are arranged co-axially around the former and and the flow is turbulent with corresponding Reynolds number range
are housed within a double walled cryostat for cooling the HTS cable. 9582-30,632. The hydraulic diameters of the annulus having depths of
The cable cryostat is mostly made up of corrugated pipes to absorb the 5, 7 and 9 mm are 18.5, 20.5 and 22.5 mm respectively. The length of
thermal shrinkage, bending stress and to ensure flexibility. cable for the present analysis (0.8 m) is calculated by taking the sum of
entrance length (0.23 m), length of fully developed turbulent flow
(0.5 m) and outlet length (0.07 m). Further, the directions of LN2 flow
and heat loads into the annulus are also shown in Fig. 2(i)–(ii).

3.1. Corrugated annular section geometry

The schematic of cable annulus, investigated in the present analysis

is shown in Fig. 2. The outer wall of annulus (i.e inner surface of cor-
rugated cryostat) consists of a series of corrugations with pitch “p” and
depth “k” as shown in Fig. 2(i). Nine different geometric configurations
are studied, with three different pitches and three different depths at
different Reynolds numbers.
In order to calculate Reynolds number (Re) for the annulus, hy-
draulic diameter is calculated by:
4 2 (Di ) 2}
4A 4 {(Do )
dh = = = (Do Di )
Pw Do + Di (1)

vDh
Re =
Fig. 1. Operating range of LN2 for HTS cable operation. µ (2)

Table 1
Geometric specifications of cable.
Annulus Diameter (mm) Length (mm) Pitch (mm) Depth (mm)

Inner diameter (smooth) (Di) 85 800 – –

Outer diameter (corrugated) (Do) 98.5 800 6, 10, 14 5, 7, 9

2
I. Das and V.V. Rao Cryogenics 103 (2019) 102950

Table 2 4. CFD model

Geometric specifications of corrugated cryostat used in numerical calculations
(p, k, dh). The LN2 flow within the annulus of HTS cable is considered as in-
Geometry sets Configuration p/dh k/dh k/p compressible and turbulent. The CFD method helps us to obtain wall
shear stress of the present complicated geometries numerically. The
Set-1 p1k1 0.333 0.278 0.833 velocity distributions of LN2 flow are investigated using FLUENT®. As
p1k2 0.3 0.35 1.167
per the model shown in Fig. 2(a), an axisymmetric CFD model is pre-
p1k3 0.273 0.409 1.5
Set-2 p2k1 0.555 0.278 0.5 pared for the analysis. The details of geometric specifications, used in
p2k2 0.5 0.35 0.7 CFD formulation are mentioned in Table 2. Simulation is carried out at
p2k3 0.455 0.409 0.9 five different Reynolds numbers within the ranges for each (p and k)
Set-3 p3k1 0.777 0.278 0.357
combination, in order to reflect its effect on friction factor and velocity
p3k2 0.7 0.35 0.5
p3k3 0.636 0.409 0.642
distribution. The numerical mesh is used in the analysis, is shown in
Fig. 2(c). For turbulent flow analysis, one has to properly take the

Fig. 2. The simplified model of annulus. (i) (a) Schematic of HTS cable (annular section with corrugated cryostat), (b) Corrugation showing Pitch and depth, (c)
Meshed model and zoomed portion showing the details of meshing. (ii) HTS cable (annular section with smooth cryostat).

The range of Reynolds number for different corrugation depths are Table 3
(9582–18,100), (13,268–25,062), (14,742–27,847), (16,217–30,632) Details of Meshing, Boundary condition and CFD formulation.
for smooth cryostat and corrugated cryostat with 5, 7 and 9 mm depths Meshing Details Boundary conditions
respectively.
The annular section has inner diameter Di and outer diameter Do. Number of nodes: 379,181 Inlet: Velocity inlet (0.18–0.34 m/s)
Number of elements: 371,716 Outlet: Outflow
For the annulus with smooth cryopipe (Fig. 2ii) the outer diameter is Do
Type of element: Quadrilateral Wall: Inner wall and Corrugated wall with
and for corrugated cryopipe it is Dm. Table 2 shows the dimensions of Minimum size of the element: uniform heat load of 0.4 W/m and 1.4 W/m
the aforementioned nine geometries; pitch (p1, p2, p3) and depth (k1, k2, 4.0128e-4 respectively [21]
k3). CFD Formulation
Pressure based Navier Stokes, 2D axis-symmetric, steady flow, Turbulent realizable k-ε
Model with enhanced wall treatment
3.2. Theoretical analysis

For a long length HTS power cable, pressure drop is an important

aspect in estimating various parameters such as length of cable, cryogen
pumping power. The pressure drop of subcooled LN2 flow is explained
using Darcy-Weisbach equation.
fL v 2
P=
2dh (3)

In order to calculate pressure drop along the length of cable, friction

factor has to be evaluated. In case of fully developed flow, the pressure
force balances the wall shear force. Hence after balancing both the
forces the friction factor can be expressed as
8 w
f=
v2 (4)

The friction factor in the annular space is estimated by considering

the existence of two types of wall surfaces (i.e. smooth surface of cable
outer and the corrugated surface of cryostat’s inner wall). Hence the
friction factor is evaluated as the sum of the contributions from both the
walls, which are proportional to their wetted perimeters and can be
expressed as [28]
fi Di + fo Do
f= Fig. 3. Validation of the present model.
Di + Do (5)

3
I. Das and V.V. Rao Cryogenics 103 (2019) 102950

viscous sublayer (the near wall laminar flow) into account. The grids from the wall shear stress using Eq. (4). The details of the mesh for-
are stretched to cluster grids near the wall and are adjusted for every mulation and boundary conditions are mentioned in Table 3.
Reynolds number in order to ensure the dimensionless distance, y+
within 5. A mesh sensitive test is conducted for all nine different in- 5. Results and discussion
vestigated geometries. The best suitable meshing is considered for the
analysis as per the mesh convergence. However, the number of ele- The wall shear stress is obtained from the CFD simulation along the
ments in the meshed model differs due to change in geometries. The length of annular section, in the fully developed turbulent flow region.
tabulated number of elements and nodes in Table 3 are for the axis- The values of wall shear stresses for all pitch and depth combinations at
symmetric meshed model shown in Fig. 2(c), taken for the p = 10 mm different Reynolds’s numbers are obtained for both smooth (cable side)
and k = 7 mm configuration. During simulation, the “velocity inlet” and corrugated (cryostat side) surfaces of the annular section through
condition is used for inlet boundary condition and “outflow” boundary which subcooled LN2 is flowing. Moreover, the friction factor is cal-
condition for the outlet. The numerical friction factor is then derived culated using average shear stress in fully developed flow region. The

Fig. 4. Friction factors as a function of flow rate at varied pitch with fixed depths (a, b, c) and at varied, depths with fixed pitch (d, e, f).

4
I. Das and V.V. Rao Cryogenics 103 (2019) 102950

Fig. 5. Streamlines from CFD for all nine geometries at a particular flow velocity of 0.22 m/s (with Re = 16217.14, 18019.04, 19820.951 for k = 5, 7 and 9 mm
respectively). All p and k are in mm.

friction factors for both the surfaces are calculated using Eqs. (4) and 6. Conclusions
(5).
The numerical formulation considered for the present simulation The friction factor and the flow pattern of subcooled liquid nitrogen
model is first validated with the results obtained from Ivanov et al. [23] flowing within the annulus of HTS cable are investigated numerically
as shown in Fig. 3. The simulated result of pressure drop shows very using RANS turbulence model, for nine different corrugation geome-
good agreements with the results of Ivanov et al. [23]. The pressure tries. From these investigations, it can be concluded that smaller k/p
drop is estimated from the friction factor using Darcy-Weisbach for- ratios (k/p ≤ 0.5) are preferred for getting lesser friction factor and
mulation (Eq. (3)). Thereafter, validated numerical formulation is used thereby lesser LN2 pumping power to maintain thermally stable op-
for current nine corrugated models. eration of HTS cable. The HTS cable with a pitch of 10 mm and depth of
All six plots in Fig. 4 correspond to the values of friction factor at 5 mm shows the minimum pressure drop of 76.57 Pa/m for 28.69 L/
three different pitches (p1, p2, p3) and three depths (k1, k2, k3) within min. Further the CFD analysis gives an insight into flow visualization
the Reynolds number range. It can be observed from Fig. 4(a)–(c) that with breaking of vortices.
the friction factor tends to decrease with the increase in corrugation This study can provide an useful guideline to the engineers and
pitch at constant corrugation depth. Similarly, Fig. 4(d)–(f) show that manufactures of flexible HTS cable while selecting the depth and pitch
friction factor increases as corrugation depth increases with constant of corrugated pipes and designing the flexible cryostat with minimum
pitch. All six plots of Fig. 4 also show the friction factor variation of pressure drop of liquid nitrogen.
smooth cryostat with flow rates. The values of friction factor in smooth
cryostat are smaller than that of corrugated cryostats. Further, the Declaration of Competing Interest
friction factors are low at larger Reynolds numbers as compared to
those for smaller Reynolds numbers. The authors declared that there is no conflict of interest.
The flow patterns of LN2 inside the corrugations of the annulus, for
all combinations of pitches and depths are shown in Fig. 5. The results References
correspond to a particular flow velocity of 0.22 m/s. The obtained flow
patterns vary with respect to k/p ratio. In smaller k/p ratios (Fig. 5(a), [1] Bednorz JG, Müller KA. Possible high TC superconductivity in the BaLaCuO system.
(d)–(i)), a single vortex tends to fill up the whole corrugation space, Z Physik B 1986;64:189–93.
[2] Takigawa H, Yumura H, Masuda T, Watanabe M, Ashibe Y, Itoh H, et al. The in-
whereas for higher k/p ratios (Fig. 5(b) and (c)), two sets of vortices are stallation and test results for Albany HTS cable project. Physica C
observed within the corrugation space. 2007;463–467:1127–31.
[3] Mukoyama S, Maruyama S, Yagi M, Yagi Y, Ishii N, Sato O, et al. Development of
500 m HTS power cable in Super-ACE project. Cryogenics 2005;45:11–5.

5
I. Das and V.V. Rao Cryogenics 103 (2019) 102950

[4] Maguire JF, Schmidt F, Bratt S, Welsh TE, Yuan J. Installation and testing results of 2015;25(3):5401304.
long Island transmission level HTS cable. IEEE Trans Appl Supercond [17] Chevtchenko O, Zuijderduin R, Smit J, Willén D, Lentge H, Thidemann C, et al. Low
2009;19:1692–7. friction cryostat for HTS power cable of Dutch project. Phys Procedia
[5] Yang HS, Kim DL, Sohn SH, Lim JH, Choi HO, Choi YS, et al. Long term performance 2012;36:1309–12.
test of KEPCO HTS power cable. IEEE Trans Appl Supercond 2009;19:1782–4. [18] Lee CH, Kim CH, Kim KS, Kim DH, Kim IS. Performance of heat transfer and
[6] Yumura H, Ashibe Y, Ohya M, Itoh H, Watanabe M, Yatsuka K, et al. Test results of a pressure drop in superconducting cable. Cryogenics 2003;43:583–8.
30-m HTS cable pre-demonstration system in Yokohama project. Physica C [19] Dondapati RS, Rao VV. Pressure drop and heat transfer analysis of long length
2010;470:1558–62. cooled HTS cables. IEEE Trans Appl Supercond 2013;23(3):5400604.
[7] Xin Y, Hou B, Bi Y, Xi H, Zhang Y, Ren A, et al. Introduction of China’s first live grid [20] Sasaki A, Hamabe M, Famakinwa T, Yamaguchi S, Radovinsky A, Okumura H, et al.
installed HTS power cable system. IEEE Trans Appl Supercond 2005;15:1814–7. Cryogenic fluid dynamics for DC superconducting power transmission line. IEEE
[8] Yagi M, Mukoyama S, Amemiya N, Ishiyama A, Wang X, Aoki Y, et al. Progress of Trans Appl Supercond 2007;17:1748–51.
275 kV–3 kA YBCO HTS cable. Physica C 2011;471:1274–8. [21] Maruyama O, Ohkuma T, Izumi T, Shiohara Y. Numerical calculation of advection
[9] Shabagin E, Heidt C, Strauß S, Grohmann S. Modeling of 3D temperature profiles heat transfer and fluid flow dynamics. IEEE Trans Appl Supercond
and pressure drop in concentric three-phase HTS power cables. Cryogenics 2015;25(3):5401505.
2017;81:24–32. [22] Ivanov Y, Watanabe H, Hamabe M, Kawahara T, Sun J, Yamaguchi S. Design study
[10] Demko J, Hassenzahl W. Liquid air as a coolant for thermal management of long- of LN2 circulation in a long SC power transmission lines. Phys Procedia
length HTS cable systems. AIP Conf Proc 2012;1434(1):273–82. 2012;36:1372–7.
[11] Pamidi S, Kim CH, Kim JH, Crook D, Dale S. Cryogenic helium gas circulation [23] Ivanov YV, Romashov MA, Bemert SE, Sytnikov VE. Choice of flexible cryostat for
system for advanced characterization of superconducting cables and other devices. 2.5 km DC HTS cable to be laid in St. Petersburg. AIP Conference Proceedings, vol.
Cryogenics 2012;52:315–20. 1573. 2014. p. 887.
[12] Kephart JT, Fitzpatrick BK, Ferrara P, Pyryt M, Miller J, Nadig C. Flow resistance [24] Koh D, Yeom H, Hong Y, Lee K. Performance tests of HTS power cable cooling
experiments in HTS cable form. IEEE Trans Appl Supercond 2013;23(3):5401304. system. IEEE Trans Appl Supercond 2004;14(2):1747.
[13] Ashworth SP, Reagor DW. A novel cooling scheme for superconducting power [25] Sasaki A, Ivanov Y, Yamaguchi S. LN2 circulation in cryopipes of superconducting
cables. Cryogenics 2011;51:161–7. power transmission line. Cryogenics 2011;51(9):471–6.
[14] Fuchino S, Tamada N, Ishii I, Higuchi N. Hydraulic characteristics in super- [26] Zajaczkowski B, Giesbers AJM, Holtrust M, Haenen E, Heijer R. Feasibility of inline
conducting power transmission cable. Physica C 2001;354:125–8. cooling in long distance HTS power line. Cryogenics 2011;51(4):180–6.
[15] Weisend II JG, Van Sciver SW. Pressure drop from flow of cryogens in corrugated [27] Barr DIH. Two additional methods of direct solution of the Colebrook-White
bellows. Cryogenics 1990;30:935–41. function. Proc Inst Civ Eng 1976;61:489–97.
[16] Li Z, Li Y, Liu W, Zhu J, Qiu M. Comparison of liquid nitrogen flow resistance in [28] Ivanov YV, Watanabe H, Hamabe M, Sun J, Kawahara T, Yamaguchi S. Circulation
corrugated pipe with smooth pipe. IEEE Trans Appl Supercond pump power for 200 m cable experiment. Physica C 2011;471:1308–12.