Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

Heat Mass Transfer (2014) 50:559–572

DOI 10.1007/s00231-013-1250-8

ORIGINAL

Numerical study on heat transfer and resistance characteristics


of supercritical water inside internally-ribbed tube
Zhenxing Zhao • Xiangyu Wang • Defu Che

Received: 26 August 2012 / Accepted: 17 October 2013 / Published online: 6 November 2013
Ó Springer-Verlag Berlin Heidelberg 2013

Abstract The effects of structural parameters for inter-


Greek letters
nally-ribbed tube on heat transfer and flow characteristics
a Spiral rising angle (°)
of supercritical water were studied numerically. The results
b Volume expansion coefficient (1/K)
show that the heat transfer and pressure loss increases with
e Turbulent energy dissipation
the increase of mass flow or heat flux. The Heat transfer
k Thermal conductivity (W/m K)
and resistance coefficients of supercritical water increase
l Dynamic viscosity (Pa s)
with the spiral rising angle decrease or rib height increase,
sw Shear stress at the wall (N/m2)
while rib width has a weak influence on heat transfer and
q Density of fluid (kg/m3)
pressure drop.
Subscripts
List of symbols 0 Inlet conditions
b Rib width (mm) max Maximum
cp Specific heat transfer (J/kg K) min Minimum
d Tube diameter (m) out Outside
G Mass flux (kg/m2s) w Wall condition
g Acceleration of gravity (m/s2)
H Rib height (mm)
hb Bulk enthalpy (kJ/kg) 1 Introduction
k Turbulent kinetic energy (m2/s2)
N Head number In supercritical boilers, vertical water-wall has found wider
Nu Nusselt number application than spirally coiled water wall. Since inter-
Pr Prandtl number nally-ribbed tube has simple structure and can be easily
q Wall heat flux (W/m2) manufactured and installed, vertical internally-ribbed tubes
Re Reynolds number are commonly used in water wall of boiler, as well as in
S Thread pitch (mm) heat exchanger in nuclear reactor and other heat transfer
T Temperature (K) equipment. However, the working fluid in the supercritical
x Axial distance (m) vertical internally-ribbed tube water-wall boiler signifi-
y? Non-dimension distance from tube wall cantly differs from that in the subcritical boiler in both heat
transfer and resistance characteristics [1–3]. So it has
become necessary to study the heat transfer and resistance
characteristics of the supercritical water in internally-rib-
Z. Zhao  X. Wang  D. Che (&) bed tube.
State Key Laboratory of Multiphase Flow in Power Engineering, In 1958, Dickinson and Weich [4] studied the heat
School of Energy and Power Engineering, Xi’an Jiaotong
University, Xi’an 710049, China transfer characteristics of supercritical water and found the
e-mail: dfche@mail.xjtu.edu.cn heat transfer promotion phenomenon. Since then, the

123
560 Heat Mass Transfer (2014) 50:559–572

research on supercritical water continued. Goldman [5] region. Cheng et al. [24] investigated the heat transfer of
proposed the pseudo-boiling heat transfer mechanism and supercritical water in various flow channels using CFX-5.6
confirmed that the heat transfer deterioration (HTD) phe- and found that the grid structure has no effect on heat
nomenon may occur around the pseudo-critical tempera- transfer for the j-e type models, as long as the dimen-
ture. Shitsman [6–8] found that there would appear no sionless distance y? of the first mesh to the wall is smaller
HTD phenomenon at quite large mass flow rates. By than 40.
comparing the surface heat transfer coefficient of super- For flow characteristics of supercritical water in the
critical water in horizontal tube and that in vertical tube, internally-ribbed tube, both experimental and numerical
Virkrev et al. [9] revealed that the influence of heat flux on studies are insufficient due to the complex flow field inside
surface heat transfer coefficient is very weak when the the tube and the various geometric parameters of inter-
temperature is far away from the pseudo-critical tempera- nally-ribbed tubes. Ackerman [25] revealed that the resis-
ture. Yamagata and Nishikawa [10] systematically studied tance coefficient in the internally-ribbed tube was 25 %
the influence of operating parameters on the heat transfer higher than that in the plain tube. However, the experi-
characteristic of supercritical water in horizontal and ver- mental results of Kolher et al. [26] found that the resistance
tical tubes. They found that the maximum surface heat coefficient in the internally-ribbed tube was two times that
transfer coefficient occurred at the temperature that is a of the similar plain tube. Chen [27] carried out experi-
little lower than the pseudo-critical temperature, and the mental study on the supercritical water heat transfer char-
temperature difference between the internal wall tempera- acteristics in plain and internally-ribbed tubes, proposing
ture and the bulk flow temperature significantly increased the empirical formulas for calculating the surface heat
with increased heat flux when the bulk fluid temperature is transfer coefficient and pressure drop of special internally-
above the pseudo-critical temperature. Stewart et al. [11] ribbed tubes.
carried out experiments on the influence of the pressure From the literature, it can be found that numerous
fluctuation on the supercritical water heat transfer charac- investigations have been carried out on the supercritical
teristics, and revealed that the maximum wall temperature water heat transfer and resistance characteristics. Never-
appeared where the pressure varied most severely while the theless, there are two obvious deficiencies: firstly, there are
maximum surface heat transfer coefficient occurred in the much more studies on plain tube than internally-ribbed
steady pressure zone. tube, and the studies are mainly on heat transfer charac-
Over the past few years, there is a good deal of exper- teristics of supercritical water, less studies have been
imental evidences for both local enhancement and local completed on resistance characteristics; secondly, all the
deterioration in turbulent mixed convective heat transfer of previous studies ignore the influence of geometric param-
supercritical fluids [12–20]. The majority of these experi- eters of internally-ribbed tube on heat transfer and resis-
ments on supercritical water or CO2 flowing in uniformly tance characteristics. In addition, in order to make more
heated vertical tubes can provide the data for wall tem- perfect research on heat transfer and resistance character-
perature and heat transfer coefficient. Results from these istics of supercritical water in the internally-ribbed tube,
early experiments showed that turbulent heat transfer of the effect of geometry should be fully investigated. Fig-
supercritical fluid is very complicated, especially for ure 1 illustrates the geometrical configuration and key
upward flow. Kurganov and Kaptilnyi [21] provided parameters of the internally-ribbed tube. This paper is
experimental data on flow structure, heat transfer coeffi- aimed at clarifying the effect of geometric parameter on the
cient and hydraulic drag of supercritical pressure CO2 heat transfer and resistance characteristics of internally-
heated in a vertical tube flowing upward and downward at ribbed tube by numerical study so as to provide a basis for
very high Reynolds numbers. They also found that the the optimization design of the internally-ribbed tube
development of the M-shaped velocity profile in upward employed in large capacity utility boilers.
flows favors the enhancement of heat transfer.
Numerical studies on supercritical water have also been
widely carried out in recent years. Koshizuka et al. [22] 2 Model and method
carried out two-dimensional numerical study on supercrit-
ical water heat transfer characteristics in a plain tube 2.1 Governing equations and turbulence models
adopting the standard j-e model, and the simulation results
agree quite well with experimental data. Lei [23] studied After supercritical water enters the internally-ribbed tube,
the heat transfer characteristics of supercritical water in the turbulent flow and heat transfer develop simultaneously in
inclined tube using RNG j-e model, revealing that there the tube. The flow in a vertical internally-ribbed tube is
may appear secondary flow due to the significant density assumed to be steady, single-phase, 3D flow. The govern-
variation of supercritical water in the large specific heat ing equations describing the flow and heat transfer are the

123
Heat Mass Transfer (2014) 50:559–572 561

Fig. 1 The geometrical


configuration of internally-
ribbed tube

conservation equations for mass, momentum and energy, C2e are constants (1.42 and 1.68 correspondingly); Gk
which are developed from conservation laws of physics and represents the generation of turbulence kinetic energy due
can be summarized in the time-averaged form as follows. to the mean velocity gradients; the influence of the buoy-
Continuity equation: ancy on e is determined by means of constant C3e; Gb is the
o turbulent kinetic energy generation due to buoyancy [32].
ðqui Þ ¼ 0 ð1Þ
oxi
2.2 Boundary conditions
Momentum equation:
o  osij op For the inlet boundary condition, the uniform profiles for
qui uj ¼  þ qgi ð2Þ all the properties are as follows:
oxj oxj oxi
Gi ¼ G0 ; T ¼ T0 ; k ¼ k0 ; e ¼ e0 : ð6Þ
Energy equation:
   The turbulent kinetic energy, k0, and the turbulent
o cp lt oT oui 
kþ  quj Cp T þ sij eff ¼ 0 ð3Þ kinetic energy dissipation, e0, in the inlet section are
oxj Prt oxj oxj
approximated from the turbulent intensity, I, and a
where k is the thermal conductivity and Prt is the turbulent turbulent characteristic length, L, as follows:
Prandtl number. There are very few acknowledged exper-
3 k3=2
imental data for Prt of supercritical pressure fluids. The k0 ¼ ðu0 IÞ2 ; e0 ¼ Cl3=4 : ð7Þ
2 L
general method is to use the constant turbulent Prandtl
number, which has originally been assigned for constant At the outlet, the diffusion fluxes for all variables in exit
property fluids, for the supercritical fluids of highly vari- direction are set to be zero.
able property. Based on available empirical and direct In addition, uniform wall heat flux condition is widely
numerical simulations data, the constant 0.85 is used in this used and validated by the previous experiments [15, 20, 25,
study [28]. 33], so no slip and constant heat flux boundary conditions
For the turbulent flow and heat transfer simulation, the are applied on the outside wall of tube as follows:
RNG j-e model and the enhanced wall function are ui ¼ 0; q ¼ qw : ð8Þ
adopted in this study since its performance has been found
to be substantially better than that of the other j-e models 2.3 Computational details
and j-w for heat transfer and flow of supercritical water
[24, 29–31]. The turbulence models can be expressed as The governing equations for flow and heat transfer in heat
follows: exchangers were solved in the Cartesian coordinate system
Turbulent kinetic energy equation: with a control volume finite difference method. The SIM-
  PLE algorithm was used to resolve the coupling between
o o ok
ðui kqÞ ¼ ðak leff Þ þ Gk þ Gb  qe ð4Þ velocity and pressure. The QUICK scheme was used to
oxi oxj oxj
discretize the momentum and energy equations. Consid-
Turbulent dissipation rate equation: ering the numerical stability of supercritical fluid, the
  second order upwind scheme was used for other transport
o o oe e
ðui eqÞ ¼ ðae leff Þ þ C1e ðC3e Gb þ Gk Þ equations [34]. The numerical computation is considered to
oxi oxj oxj k
be convergent when the residual summed over all the
e2
 C2e q  Re ð5Þ computational nodes at the nth iteration is no more than
k
10-6.
where leff stand for the effective viscosity; rk and re rep- For accurate and efficient computation as well as ease of
resent the turbulent Prandtl numbers for k and e; C1e and grid generation for this complex configuration, the

123
562 Heat Mass Transfer (2014) 50:559–572

Fig. 2 Local unstructured grids on internally-ribbed tube

Fig. 3 Inner wall temperature versus bulk flow enthalpy


Table 1 A28 9 6 mm four-head internally-ribbed tube geometric
parameter
supercritical pressure. The geometric parameters of inter-
Parameter Size
nally-ribbed tube are shown in Table 1. The heated length
Outer diameter dout/mm 28 of the test section was 500 mm, and the unheated length
Max inner diameter dmax/mm 16.0 upstream of the test section was 150 mm. The fluid tem-
Wall thickness/mm 6.0 perature was measured by 10 pairs of NiCr-NiSi sheathed
Head number N 4 thermocouples installed in the test section. The 30 ther-
Rib width b/mm 9 mocouples were welded onto the tube outer surface to
Rib height H/mm 0.81 measure the outside wall temperature. The experimental
Thread pitch S/mm 21 conditions simulated are as follows: the pressure at the inlet
Lead l/mm 82 of the test section is 26 MPa; the mass flux of the water is
Spiral rising angle a/° 61.15° 600 kg/m2s; the heat flux on the inside wall of the tube is
500 kW/m2.
Figure 3 shows the inner wall temperature variation
computational region was composed of unstructured hybrid with the bulk flow enthalpy in the vertically upward
grids. The final mesh of present work was obtained after internally-ribbed tube. It can be seen that the numerical
several grid adaptions based on y? in order to resolve the results agree well with experimental results when the
heat transfer coefficient accurately. In order to obtain the enthalpy is lower than 1,900 kJ/kg, while the numerical
satisfactory solutions for supercritical water in internally- results are higher than the experimental results when the
ribbed tube, the grid independence is carried out in the enthalpy is above 1,900 kJ/kg. This phenomenon is mainly
analysis by adopting different amounts of grid cells. When due to that high enthalpy water is in the large heat region,
the computational grids consist of about 1.10 million vol- where the near-wall fluid physical properties change dra-
umes, no visible grid influence on the heat transfer coef- matically with increased temperature, leading to a complex
ficient and resistance coefficient was observed. The local flow field, which can not be simulated perfectly. Although
refined grid topology of internally-ribbed tube is shown in both the enhanced wall function and the near wall grid
Fig. 2. refinement are used to calculate the near wall flow field,
there still exists certain deviation. However, the numerical
results agree well with the experimental data on the whole,
3 Model validation and parameters considered the relative deviation between simulated prediction and
experimental data is smaller than 5 %. The above com-
3.1 Model validation parison reveals that both the model and the grid in this
paper are valid and reliable.
The experiments for verifying the present numerical work
were carried out in the high-temperature and high-pressure 3.2 Parameters considered
flow test loop at Xi’an Jiaotong University. Experimental
results were obtained with water flowing upwards in uni- In designing an internally-ribbed tube, the main geometric
formly heated vertical internally-ribbed tube under parameters of the internally-ribbed tube include spiral

123
Heat Mass Transfer (2014) 50:559–572 563

Table 2 The structural parameters of internally-ribbed tubes and plain tube


Parameter Plain tube Tube 1 Tube 2 Tube 3 Tube 4 Tube 5 Tube 6 Tube 7

Outer diameter dout/mm 14 14 14 14 14 14 14 14


Max inner diameter dmax/mm 10 10 10 10 10 10 10 10
Wall thickness/mm 2 2 2 2 2 2 2 2
Head number N 0 4 4 4 4 4 4 4
Rib width b/mm 0 4 4 4 2 6 8 4
Rib height H/mm 0 0.5 0.5 0.5 0.5 0.5 0.5 1
Thread pitch S/mm 0 13.6 7.85 29.3 13.6 13.6 13.6 13.6
Lead l/mm 0 54.4 31.4 117.2 54.4 54.4 54.4 54.4
Spiral rising angle a (°) 0 60 45 75 60 60 60 60

Table 3 The mass flow and heat flux for numerical simulation 1. It is found that the tangential velocity increases with the
G rise. The flow resistance in the channel is proportional to
Working condition Parameter
the square of fluid velocity, and the axial flow resistance in
Mass flow G (kg/m2s) 500, 700, 900, 1,100, 1,300, 1,500 channel increases with the G rise, thus, there is more tan-
Heat flux qw (kW/m2) 50, 150, 250, 350, 450 gential flow through the helical grooves in the tube.
Obviously, the higher the tangential velocity is, the more
rising angle, rib height and rib width. 7 internally-ribbed remarkable the heat transfer enhancement grows. However,
tubes and 1 same diameter plain tube are investigated in when q = 450 kW/m2 and G = 500 kg/m2s, Nu is con-
this paper, and the detailed structural parameters are shown siderably below that it should be. It can be explained as
in Table 2. Due to the significant variation of operational follows: smaller G leads to less tangential flow which
condition in the supercritical boiler, wide mass flow and would bring a thicker fluid boundary layer near the wall to
heat flux ranges are designed in numerical simulations to weaken heat transfer.
cover the conditions of a practical boiler. Details are pre-
sented in Table 3. 4.1.2 Influence of heat flux on heat transfer

Figure 4d–f show the influence of heat flux on heat


4 Results and discussion transfer for Tube 1, Tube 2 and Tube 3. Nu variation
tendency presents certain linearity with increased heat
4.1 Heat transfer characteristics flux. Similar results have been obtained for other inter-
nally-ribbed tubes. It can be seen from Fig. 5a, d and e
When mass flux is maintained constant, HTD may happen that the tangential velocity fields for different q were very
and the surface heat transfer coefficient may decrease similar, but the region of the tangential flow increases
sharply with further increase of heat flux. In the present with q rise. This could decrease the wake region in the
study, mass fluxes are all greater than 500 kg/m2s and heat helical groove to enhance the local heat transfer. In
fluxes are all less than 450 kW/m2, so no deterioration addition, for proper mass flux, the temperature gradient
phenomenon appears [10]. between inner wall and bulk fluid and the turbulence
generation rate increase with increased heat flux, resulting
4.1.1 Influence of mass flow on heat transfer in the increase of surface heat transfer coefficient. In the
present study, all the mass flux values are higher than the
The simulated results for Tube 1, Tube 2 and Tube 3 are mass flux that would happen HTD, so the surface heat
presented in Fig. 4a–c. Obviously, for internally-ribbed transfer coefficient increases with increased heat flux.
tubes, Nu increases monotonously with Re and shows good When the mass flux is higher than 900 kg/m2s, as long as
linearity. The results for other internally-ribbed tubes are the heat flux is over 450 kW/m2, it seems that Nu is
also similar to those of the above 3 tubes. With increased varied with a higher magnitude. This is so called heat
mass flow, both axial and circumferential velocities transfer promotion phenomenon. It is because that the
increase significantly, and the generation and development specific heat of supercritical water is larger and its heat
rate of flow turbulence become higher, therefore the heat absorption capacity is also greater in the large specific
transfer is enhanced. Figure 5a–c show the comparison of heat region. When the heat flux is high enough, more heat
tangential velocity fields for different G in the case of Tube will be transferred to the supercritical water.

123
564 Heat Mass Transfer (2014) 50:559–572

Fig. 4 Influence of mass flow (a–c) and heat flux (d–f) on heat transfer

4.1.3 Influence of spiral rising angle on heat transfer effect, the significant change of supercritical water turbu-
lence intensity has completed in unheated section. As the
Figure 6a, b show the influence of spiral rising angle on spiral rising angle decreases, the bulk fluid turbulence
heat transfer in internally-ribbed tubes. Due to the entrance intensity increases dramatically and the fluid tangential

123
Heat Mass Transfer (2014) 50:559–572 565

tangential velocity is and the more widely the tangential


velocity changes. For instance, under the same axial
velocity, the greatest tangential velocity is 0.7 m/s when
the spiral rising angle is 45°, while the value is only 0.4 m/
s when the spiral rising angle is 60°.

4.1.4 Influence of rib width on heat transfer

Figure 6c, d show the influence of rib width on heat transfer


in internally-ribbed tubes. When the rib width is varied from
2 to 8 mm, insignificant effect on heat transfer can be found.
It is because that heat transfer is the energy exchange
between the wall and the bulk fluid as well as the generation
and development process of the turbulence, but the change of
rib width does not exert a significant influence on the velocity
distribution in tube. Therefore the change of rib width does
not have significant influence on the heat transfer of super-
critical water. For different rib widths, the Nu variation with
Re performs obvious similarity in wide heat flux and mass
flow ranges. So, it is unnecessary to take primary consider-
ation on rib width during the design of internally-ribbed tube.
Figure 7b shows the turbulence intensity variation in the
axial direction for different rib widths when q = 450 kW/
m2 and G = 500 kg/m2s. It can be found that the tendency
and the absolute value of turbulence intensity are almost
identical. In addition, Fig. 8c, d show the bulk fluid tan-
gential velocity distributions for Tube 4 and Tube 6. It can
Fig. 5 Influence of mass flow (a–c) and heat flux (a, e, f) on
tangential velocity fields be seen that, similar to the results of the density distribu-
tion, the distributions of the fluid tangential velocity are
velocity becomes higher, resulting in the enhancement of also almost identical even if the rib widths are different. So,
heat transfer. it can be concluded that rib width has a very weak influence
Under the working condition of q = 450 kW/m2 and on the heat transfer of supercritical water.
G = 500 kg/m2s, the bulk fluid turbulence intensity vari-
ation in the axial direction is shown in Fig. 7a. It can be 4.1.5 Influence of rib height on heat transfer
seen that the smaller the spiral rising angle is, the greater
the bulk fluid turbulence intensity is. The energy exchange Figure 6e, f show the influence of rib height on heat
between wall and bulk fluid becomes more intense, leading transfer. As the rib height increases from 0.5 to 1.0 mm,
to the increase of surface heat transfer coefficient. How- the surface heat transfer coefficient increases to certain
ever, the greater turbulence intensity may cause greater extent, indicating that the increase of rib height rise can
energy loss. Therefore, the spiral rising angle should be give rise to the heat transfer enhancement for the same
further optimized according to the heat transfer and resis- mass flow and heat flux.
tance characteristics of the internally-ribbed tube. Figure 7c shows the turbulence intensity variation in the
Figure 8a, b show the comparison of tangential velocity axial direction with different rib heights under the working
fields between Tube 1 and Tube 2 at q = 450 kW/m2 and condition of q = 450 kW/m2 and G = 500 kg/m2s. There
G = 500 kg/m2s. As the spiral rising angle decreases, the is an obvious difference between rib heights of 0.5 and
tangential velocity of the fluid in the internally-ribbed tube 1.0 mm, illustrating that the change of rib height can lead
increases significantly. The relative velocity between the to the change of turbulence intensity of supercritical water.
bulk fluid and the near-wall fluid may increase, which may The higher the rib height is, the greater the turbulence
reduce the thickness of the near-wall laminar boundary intensity is and the stronger the mass and energy exchange
layer. Moreover, the centrifugal force induced by the spiral between the wall and the bulk fluid is. Therefore, the heat
flow can also strengthen the mass and energy exchange transfer is enhanced accordingly.
between the fluid near wall and the bulk fluid. Obviously, Figure 8a, e show the tangential velocity distributions in
the smaller the spiral rising angle is, the greater the Tube 1 and Tube 7. For different rib heights, the tangential

123
566 Heat Mass Transfer (2014) 50:559–572

Fig. 6 Influence of spiral rising angle (a, b), rib width (c, d) and rib height (e, f) on heat transfer

velocities are quite different with each other. Moreover, the height gets higher, resulting in much more significant tur-
tangential velocity increases with increased rib height. The bulence increase. In addition, the density distribution tends
flow inside the internally-ribbed tube can be recognized as to be more uniform as the rib height increase from 0.5 to
the resultant of the axial fluid flow over the rib and the 1.0 mm, indicating that the rib height increase can effi-
circumferential fluid flow along the spiral channel. Both the ciently reduce the density gradient and improve the heat
two kinds of flows will become much stronger as the rib transfer.

123
Heat Mass Transfer (2014) 50:559–572 567

Fig. 8 Influence of spiral rising angle (a, b), rib width (c, d) and rib
height (a, e) on tangential velocity fields

to the enhancement of heat transfer. Hence, the increase of


rib height can generate the heat transfer enhancement
effect.

4.2 Resistance characteristics in the internally-ribbed


tube

Pressure drop is another main parameter in heat exchanger


design, and it is characterized by the resistance coefficient
f here.

4.2.1 Influence of mass flux on resistance characteristics

Figure 9 shows the resistance coefficient variation with the


mass flux for the internally-ribbed tube with various
structural parameters. It can be seen that the resistance
Fig. 7 Influence of spiral rising angle (a), rib width (b) and rib height coefficient decreases with the increase of Re. For all the
(c) on turbulence intensity
internally-ribbed tubes, the decrease rate of resistance
coefficient with Re are fast at first and slow subsequently,
The maximum tangential velocity increases from 0.4 to and it approaches a constant values as Re increases, indi-
1.55 m/s as the rib height changes from 0.5 to 1.0 mm. cating that the supercritical water flow has entered the so
Because the tangential velocity can significantly influence called self-modeling region. In addition, for different
the fluid turbulence intensity and temperature distribution internally-ribbed tube structural parameters and heat fluxes,
[34], the increase of tangential velocity can be beneficial there are different asymptotic coefficients, revealing that

123
568 Heat Mass Transfer (2014) 50:559–572

Fig. 9 Influence of mass flow on resistance characteristics

the influences of structural parameter and heat flux on working condition of G = 500 kg/m2s. When
2 2
resistance are quite significant. q = 450 kW/m and G = 500 kg/m s, the boundary layer
near the inner wall is fairly thick. The energy exchange
4.2.2 Influence of heat flux on resistance characteristics between the wall and the bulk fluid will be inhibited,
resulting in weak turbulence generation and development,
Figure 10 shows the resistance coefficient variation with and reduced shear stress as well. Hence, the resistance
heat flux under various structural parameters. It can be seen coefficient decreases.
that resistance coefficient increases with increased heat flux
for all the internally-ribbed tubes. It is because that the 4.2.3 Influence of spiral rising angle on resistance
increase of heat flux leads to the increase of temperature characteristics
gradient and the turbulence dissipation rate of supercritical
water, which is consistent with Wang’s experimental Figure 11a, b show the influence of spiral rising angle on
results [35]. resistance coefficient. For all the heat fluxes and mass
For a given internally-ribbed tube, the increment of fluxes, the resistance coefficient increases with reduced
resistance coefficient with heat flux differs, depending on spiral rising angle. With increased mass flux, the flow in
mass flux. When G [ 1,100 kg/m2s, the increment from tubes will enters the self-modeling region finally. How-
q = 50 kW/m2 to q = 450 kW/m2 is great, approaching ever, the higher the heat flux is, the greater the corre-
50 %. When G B 1,100 kg/m2s, this increment is much sponding Re needed to reach the self-modeling region is.
greater, approaching 100 %. In addition, it can be seen For instance, the bulk fluid enters the self-modeling
from Fig. 10 that the increase rate of resistance coefficient region when Re [ 140,000 for q = 150 kW/m2 and
gets faster and faster with increasing heat flux except the q = 250 kW/m2, whereas the bulk fluid does not enter

123
Heat Mass Transfer (2014) 50:559–572 569

Fig. 10 Influence of heat flux on resistance characteristics

the self-modeling region until Re [ 220,000 for obvious variation of resistance for given heat flux and mass
q = 350 kW/m2 and q = 450 kW/m2. For high heat flux flux. As illustrated in Fig. 8c, d, the change of rib width
and low mass flux, the boundary layer near the inner does not give rise to a significant change of flow field.
wall may be fairly thick, resulting in weak turbulence Therefore, rib width has no considerable effect on the
and reduced shear stress. The tiny change of inner wall resistance coefficient as well.
structure can lead to the change of fluid layer structure,
which will dramatically affect the velocity gradient 4.2.5 Influence of rib height on resistance characteristics
between wall and bulk fluid as well as the near-wall flow
sheer stress, leading to the change of resistance coeffi- Figure 11e, f show the influence of rib height on the
cient. Besides, under all working conditions, the resis- resistance coefficient. It can be seen that the resistance
tance coefficients of internally-ribbed tubes are 2–5 times coefficient for 1.0 mm rib height is greater than that for
that of the plain tube. 0.5 mm rib height over wide heat flux and mass flux ran-
ges. As shown in Fig. 7c, the supercritical water turbulence
4.2.4 Influence of rib width on resistance characteristics intensity value for 1.0 mm rib height is much greater than
that for 0.5 mm rib height, which may result from the
Figure 11c, d show the influence of rib width on the dissipation of the fluid energy. Therefore, higher resistance
resistance coefficient. Obviously, the resistance variation coefficient is induced. As illustrated in Fig. 8a, e, the tan-
with Re are essentially coincident with each other, gential velocity increases with increased rib height, greater
revealing that the change of rib width does not lead to resistance coefficient will be induced as well.

123
570 Heat Mass Transfer (2014) 50:559–572

Fig. 11 Influence of spiral rising angle (a, b), rib width (c, d) and rib height (e, f) on resistance characteristics

5 Conclusions resistance characteristics under supercritical pressure, fol-


lowing conclusions can be drawn:
Numerical simulations on turbulent mixed convection in
1. When mass flux is greater than 500 kg/m2s and heat
internally-ribbed tube have been performed to investigate
flux is less than 450 kW/m2, the heat transfer in
the influence of geometric parameters on heat transfer and

123
Heat Mass Transfer (2014) 50:559–572 571

internally-ribbed tube is enhanced with increased mass Wissenschaftliche Berichte, FZKA 6609, Institute für Kern- und
flow under a fixed heat flux. Moreover, the heat Energietechnik
14. Jackson J (2006) Studies of buoyancy-influenced turbulent flow
transfer is also enhanced with increased heat flux. and heat transfer in vertical passages. In: Keynote Lecture at the
2. The fluid tangential velocity and main flow turbulent International Heat Transfer Conference, IHTC13, Sidney,
intensity increase significantly with reduced spiral Australia
rising angle or increased rib height, leading to the 15. Jackson J (2001) Some striking features of heat transfer with
fluids at pressures and temperatures near the critical point. In:
increase of heat transfer coefficient, and rib width has a Keynote paper for international conference on energy conversion
weak influence on the internally-ribbed tube heat and application, ICECA, Wuhan, China
transfer. 16. Hall WB, Jackson J (1969) Laminarization of a turbulent pipe
3. The heat flux increase promotes the radial temperature flow by buoyancy forces. ASME Paper 69-HT-55
17. Kim J, Jeon H, Yoo J, Lee J (2005) Experimental study on heat
gradient and the turbulent dissipation rate of super- transfer characteristics of turbulent supercritical flow in vertical
critical water, so the resistance coefficient increases circular/non-circular tubes. In: Proceedings of the 11th NU-
accordingly. Under a certain mass flux, the resistance RETH-11, Avignon, France, October 2
coefficient of internally-ribbed tube increases with 18. Pis’menny E, Razumovskiy V, Maevskiy E, Koloskov A, Pioro I
(2006) Heat transfer to supercritical water in gaseous state or
reduced spiral rising angle or increased rib height. affected by mixed convection in vertical tubes The ASME
However, there appears no obvious influence of rib Foundation, Inc., Three Park Avenue, New York, NY
width on the resistance characteristic. 10016-5990 (United States)
19. Jiang PX, Zhang Y, Shi RF (2008) Experimental and numerical
investigation of convection heat transfer of CO2 at supercritical
pressures in a vertical mini-tube. Int J Heat Mass Transf
51:3052–3056
References 20. Wang J, Li H, Guo B, Yu S, Zhang Y, Chen T (2010) Investi-
gation on the mechanism of abnormal heat transfer of supercrit-
1. Asinari P (2005) Numerical prediction of turbulent convective ical pressure water in vertically—upward tubes in the large
heat transfer in mini/micro-channels for carbon dioxide at specific heat region. In: 6th International symposium on multi-
supercritical pressure. Int J Heat Mass Transf 48:3864–3879 phase flow, heat mass transfer and energy conversion. AIP con-
2. Bae JH, Yoo JY, Choi H (2005) Direct numerical simulation of ference proceedings vol 1207, pp 531–536
turbulent supercritical water cooled flow channels. Phys Fluids 21. Kurganov VA, Kaptilnyi AG (1993) Flow structure and turbulent
17:1–24 transport of a supercritical pressure fluid in a vertical heated tube
3. Kim WS, He S, Jackson JD (2008) Assessment of predictions of under the conditions of mixed convection. Experimental data. Int
turbulent models of mixed convection by comparison with DNS. J Heat Mass Transf 36:3383–3392
Int J Heat Mass Transf 51:1293–1312 22. Koshizuka S, Takano N, Oka Y (1995) Numerical analysis of
4. Dickinson N, Weich C (1958) Heat transfer to supercritical water. deterioration phenomena in heat transfer to supercritical water.
ASME 80:745–751 Int J Heat Mass Transf 38:3077–3084
5. Goldman K (1961) Heat transfer to supercritical water at 5000 psi 23. Lei XL, Li HX, Yu SQ, Ren DL (2010) Numerical simulation on
flowing at high mass flow rates through round rubes. In: Inter- heterogeneous heat transfer in water at supercritical pressures in
national Developments in Heat Transfer, Part III, January 8–12. inclined upward tubes. Chin J Comput Phys 27:217–228
ASME, University of Colorado, Boulder, USA, Paper 66. 24. Cheng X, Kuang B, Yang YH (2007) Numerical analysis of heat
pp 561–568 transfer in supercritical water cooled flow channels. Nucl Eng
6. Shitsman M (1962) Investigation of heat transfer during the Des 237:240–252
cooling of water in the near-critical region. Teploenergetika 25. Ackermann JW (1970) Pseudoboiling heat transfer to supercriti-
(Therm Eng) 9:83–86 cal pressure water in smooth and ribbed tubes. J Heat Transfer
7. Shitsman M (1959) Heat transfer to water oxygen and carbon 92:490–497
dioxide in the approximately critical range. Teploenergetiky 26. Kolher W, Kastner W (1986) Heat transfer and pressure loss on
10:185–193 rifled tubes. In: 8th International heat transfer conference, San
8. Shitsman M (1966) The effect of natural convection on temper- Francisco, pp 2861–2865
ature conditions in horizontal tubes at supercritical pressure. 27. Chen T (2003) Experimental study on heat transfer characteristic
Therm Eng 13:69–75 of supercritical boiler. In: Chinese society of engineering ther-
9. Virkrev Y, Barulin Y, Konkov A (1967) A study of heat transfer mophysics multiphase flow conference, Beijing, pp 530–536
in vertical tubes at supercritical pressure. Therm Eng 14:116–119 28. Kays WM, Crawford ME, Weigand B (1993) Convective heat
10. Yamagata K, Nishikawa K (1972) Forced convection heat and mass transfer. McGraw-Hill, New York
transfer to supercritical water flowing in tubes. Int J Heat Mass 29. Yang J, Ishiwatari Y, Liu J, Yoo J (2007) Numerical investi-
Transf 25:2575–2593 gation of heat transfer in upward flows of supercritical water in
11. Stewart E, Stewart P, Watson A (1973) Thermo-acoustic oscil- circular tubes and tight fuel rod bundles. Nucl Eng Des 237:
lations in forced convection heat transfer to supercritical pressure 420–430
water. Int J Heat Mass Transf 16:257–270 30. Kim SH, Kim YI, Bae Y, Cho B (2004) Numerical simulation of
12. Pioro IL, Duffey RB (2005) Experimental heat transfer in the vertical upward flow of water in a heated tube at supercritical
supercritical water flowing inside channels. Nucl Eng Des pressure. In: Paper 4047, Proceeding of 2004 international con-
235:2407–2430 gress on advances in nuclear power plants (ICAPP), 13–17 June
13. Cheng X, Schulenberg T (2001) Heat transfer at supercritical 2004, Pittsburgh, USA
pressures: literature review and application to an HPLWR, Report 31. Mahdi M, Majid B (2011) The effect of the low Reynolds number
Forschungszentrum Karlsruhe, Technik und Umwelt, k-e turbulence models on simulation of the enhanced and

123
572 Heat Mass Transfer (2014) 50:559–572

deteriorated convective heat transfer to the supercritical fluid 34. Zhao Z, Wang X, Che D, Cao Z (2011) Numerical studies on flow
flows. Heat Mass Transf 47:609–619 and heat transfer in membrane helical-coil heat exchanger and
32. Choudhury D (1993) Introduction to the renormalization group membrane serpentine-tube heat exchanger. Int Commun Heat
method and turbulence modeling. Fluent Incorporated Technical Mass Transfer 38:1189–1194
Memorandum TM-107 35. Wang J, Li H, Guo B, Yu S, Zhang Y, Chen T (2009) Investi-
33. Shiralkar B, Griffith P (1970) The effect of swirl, inlet conditions, gation of forced convection heat transfer of supercritical pressure
flow direction, and tube diameter on the heat transfer to fluids at water in a vertically upward internally ribbed tube. Nucl Eng Des
supercritical pressure. J Heat Transfer 92:465–471 239:1956–1964

123

You might also like