This article was downloaded by: [Fondren Library, Rice University ]

On: 03 May 2013, At: 10:03

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

HVAC&R Research
Publication details, including instructions for authors and subscription information:
[Link]

Heat Transfer from Supercritical Carbon Dioxide in

Tube Flow: A Critical Review
a
Srinivas S. Pitla , Douglas M. Robinson , Eckhard A. Groll & Satish Ramadhyani
a
School of Mechanical Engineering, Purdue University, Indiana
Published online: 28 Feb 2011.

To cite this article: Srinivas S. Pitla , Douglas M. Robinson , Eckhard A. Groll & Satish Ramadhyani (1998): Heat Transfer from
Supercritical Carbon Dioxide in Tube Flow: A Critical Review, HVAC&R Research, 4:3, 281-301

To link to this article: [Link]

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: [Link]

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 TABLE OF CONTENTS

VOL. 4, NO. 3 HVAC&R RESEARCH JULY 1998

Heat Transfer from Supercritical Carbon Dioxide

in Tube Flow: A Critical Review
Srinivas S. Pitla Douglas M. Robinson
Student Member ASHRAE Student Member ASHRAE

Eckhard A. Groll, Ph.D. Satish Ramadhyani, Ph.D.

Member ASHRAE
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Since the discovery that CFCs and HCFCs destroy the ozone layer and cause global warming, the
need to regulate their use has been critical. In the ensuing search for new environmentally benign
refrigerants, carbon dioxide has been considered due to its excellent thermophysical properties.
This paper gives a review of the heat transfer and pressure drop characteristics of supercritical
carbon dioxide in tube flow. This information is necessary for designing the gas cooler of a car-
bon dioxide refrigeration system. A comparison of the different heat transfer correlations appli-
cable to cooling of supercritical carbon dioxide has also been made.

INTRODUCTION
With the growing awareness of the dual threats of ozone depletion and global warming, sig-
nificant research activity has been directed to the identification and development of environmen-
tally benign refrigerants. Carbon dioxide (R-744) has recently been subjected to scrutiny as a
refrigerant for automobile air conditioners. Since carbon dioxide is a constituent of the atmo-
sphere, it would be obtained from atmospheric air by fractionation and, thus, would have no
additional impact on global warming, except for the energy consumption associated with the
fractionation process. Carbon dioxide possesses a high specific heat, a high volumetric heat
capacity and, in general, excellent thermodynamic and transport properties—all desirable char-
acteristics of a refrigerant. Carbon dioxide was used as a refrigerant until the 1930s, but was
then replaced by the CFCs and HCFCs that offered lower absolute pressures and higher efficien-
cies in conventional vapor compression cycles. Recent research on carbon dioxide has focused
on the development of a transcritical cycle.
The process path of a transcritical carbon dioxide cycle is shown in Figure 1. The path
1-2-3-4h-1, as shown in the figure, shows compression (1-2), isobaric heat rejection (2-3), adia-
batic expansion (3-4h), and isobaric evaporation (4h-1). The temperature and pressure at the
critical point (CP) are 303.97 K and 7.353 MPa, respectively. Many of the recent investigations
of carbon dioxide as a refrigerant have included a thermodynamic cycle analysis, that yields
information such as the theoretical coefficient of performance (COP) and the required mass flow
rate. However, relatively few investigations have been performed to quantify local heat transfer
coefficients and pressure drop characteristics during heat rejection or heat absorption, and the
available information is both sparse and scattered.
The heat rejection process 2-3 from supercritical carbon dioxide is of particular interest. This
process occurs along an isobar that is close to the critical isobar. The heat exchanger in which
the process occurs is called a “gas cooler,” instead of a condenser, because the process occurs at

Srinivas S. Pitla and Douglas M. Robinson are graduate research assistants, Eckhard A. Groll is an assistant profes-
sor, and Satish Ramadhyani is a professor in the School of Mechanical Engineering, Purdue University, Indiana.

281
 VOLUME 4, NUMBER 3, JULY 1998 282
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Figure 1. A transcritical cycle of carbon dioxide

(Robinson and Groll 1998)

a supercritical pressure, and a phase change does not take place. However, the thermophysical
properties of the fluid do change drastically during the process. While a great deal of research
has been performed to determine the heat transfer and pressure drop characteristics during con-
densation of CFC, HCFC, and HFC refrigerants, comparatively few investigations have been
conducted on the in-tube heat transfer and pressure drop of a supercritical fluid. This paper pre-
sents a detailed literature survey on the in-tube heat transfer and pressure drop of supercritical
carbon dioxide.
This survey revealed that in the 1960s and 1970s several investigations were conducted on the
heat transfer of supercritical fluids, such as carbon dioxide and water, flowing in small diameter
tubes. Table 1 is a compilation of the literature discovered during this survey, and arranged in
chronological order. The entries in the various columns of the table serve to provide a synopsis
of the parameters covered in each investigation. This paper summarizes and highlights the out-
come of these investigations. It is organized under the following major topics:
• Thermophysical properties of carbon dioxide in the supercritical region
• Physical factors influencing in-tube forced convection heat transfer
– Deterioration/improvement of heat transfer in the supercritical region
– Effect of buoyancy-driven secondary flows on heat transfer in tube flow
• Heat transfer correlations for in-tube flow of supercritical carbon dioxide
• Coefficient of friction for in-tube flow of a supercritical fluid
• Numerical methods used in the calculation of heat transfer at supercritical pressure

THERMOPHYSICAL PROPERTIES OF CARBON DIOXIDE

The thermophysical properties of carbon dioxide are, of course, extremely important in
determining the heat transfer. These properties, i.e., the thermal conductivity, the viscosity,
the specific heat, and the density, display extremely large variations near the critical point, as
shown in Figures 2a, 2b, 3a, and 3b, respectively. The region in which these large variations
are observed is called the “pseudocritical region” by many authors.
 283 HVAC&R RESEARCH

Table 1. Heat Transfer of Fluids at Supercritical Pressures

Bulk Tube
Re Number Temp, Pressure, Exp/ Orienta-
Author(s) {Gr Number} °C MPa Fluid Theo tion
Petukhov et al. (1961) 5×104–3×105 4–50 9,10,11 CO2 E/T Hz H
Koppel and Smith (1961) 3×104–3×105 18–49 7.3–7.5 CO2 E Hz H
Shitsman (1963) ~5×104 — ≈23 water E H
Melik-Pashaev (1966) >5×104 — — CO2 T H
Krasnoshchekov and Protopopov (1966) 104–6×105 20–110 7.75,9.7 CO2 E Hz H
Shiralkar and Griffith (1969) 105–4.5×105 — 7.48–7.85 CO2 E/T Ver H
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Krasnoshchekov et al. (1970) 2.67×105 30–215 8–12 CO2 E Hz C

–8.35×105
Shiralkar and Griffith (1970) 9×104–1.65×106 — — CO2 E Ver H
Bourke et al. (1970) ~1.5×105 — 7.35 – 10.2 CO2 E Ver H
Tanaka et al. (1971) — ~31 — CO2 E Ver H
Bourke and Pulling (1971) — — — CO2 E Hz H
Shykov et al. (1971) — — — water T H
Yamagata et al. (1971) — 230–540 22.6–29.4 water E/T Hz Ve
Krasnoshchekov and Protopopov (1972) — — — CO2 T —
Krasnoshchekov et al. (1972) 6×105–1.2×106 22–45 10.32–10.75 CO2 E Ver H
{7.5×106–15×106}
Petukhov et al. (1972) 1.2×104–1.6×105 — 7.85,9.8 CO2 E Hz Ver
H
Kuraeva and Protopopov (1974) 0.8×105–4.5×105 19–88 8,10 CO2 E/T Hz H
Petukhov and Polyakov (1974) — — — — T Ver Hz
Shitsman (1974) — — — CO2, T —
H2O, He
Adebiyi and Hall (1976) 0.2×105–2×105 — — CO2 E Hz H
Baskov et al. (1977) 0.95×105–6.4×105 17–212 8,10,12 CO2 E Ver C
{9.2×107–18.8×107}
Protopopov (1977) — — — CO2, T Hz H
water
Petrov et al. (1978) — 40–80 7–10 CO2 E Hz C
Petukhov et al. (1980) — — — CO2 E Hz H
Petrov and Popov (1985a) — — — CO2 T Hz Ver
C
Petukhov et al. (1986) — — — CO2 E Ver H
Renz and Bellinghausen (1986) — — — CO2 T Hz H
Popov and Valueva (1990) — — — SC fluids T H
Griem (1995) — — — Water T H
Kochizuka et al. (1995) — — — Water T Ver C
Zhou and Krishnan (1995) — — — CO2 T —
Lee and Howell (1998) — — — CO2 T Ver H
Tube orientation: Hz = horizontal Ver = vertical H = heating C = cooling
 VOLUME 4, NUMBER 3, JULY 1998 284
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Figure 2a. Conductivity versus Figure 2b. Viscosity versus

temperature temperature

Figure 3a. Specific Heat versus Figure 3b. Density versus

temperature temperature

The sudden reductions in density, viscosity, and thermal conductivity with increasing temper-
ature are due to the change from a “liquid-like” to a “gas-like” supercritical state. The observed
spike in specific heat is associated with the fact that cp → ∞ at the critical point, by definition.
Figures 2a and 2b were generated using the data provided by Vesovic et al. (1990) while Figures
3a and 3b were generated in accordance with Span and Wagner (1996).
A very detailed description of the variations of the viscosity and the thermal conductivity of
carbon dioxide in the pseudocritical region has been given by Vesovic et al. (1990). These
authors used a combination of theoretical predictions and the best available experimental data
to develop equations for the calculation of the viscosity and the thermal conductivity. They
expressed both transport properties as sums of three independent contributions: the transport
property in the limit of zero density, the contribution due to “critical enhancement” and the
“excess property” term. The contribution at the limit of zero-density is associated with the
two-body molecular interactions considered in the kinetic theory of gases. The critical
 285 HVAC&R RESEARCH

enhancement contribution is due to the effect of long-range fluctuations in the properties at and
near the critical point. Rice (1955) may be consulted for an in-depth discussion of this effect.
The last contribution represents effects other than the ones mentioned, including multi-body
interactions. While the first two contributions can be determined by semi-theoretical equations,
the third one can only be determined experimentally. In addition to providing equations, the
authors have also presented graphs and tabulations that show the variation of the transport
properties with temperature and pressure. The authors provided a comparison of data generated
by their equations with several sets of experimental data, indicating that their equations pro-
duced results typically within ± 5% of the data.
Stephan and Lucas (1979) tabulated the viscosity in the temperature range of 300 K to 900 K
and in the pressure range of 3 to 100 MPa. The tabulation is based on a compilation of data from
various sources including experimental data from Herreman et al. (1970) and Kiyama and
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Mikita (1951), and theoretical works such as Kennedy (1954). Vargaftik (1975) and VDI
(1988), have provided other tabulations of the thermal conductivity, and VDI (1988) has also
tabulated the viscosity. VDI (1988) data is based on the experimental work of Ulybin and Baku-
lin (1977) for the thermal conductivity and primarily on Ulybin and Makarushkin (1976) for the
viscosity. In the seventh edition of VDI (1994) the thermal conductivity and the viscosity values
cited are those given by Vesovic et al. (1990) and Span and Wagner (1996). Texas A&M Uni-
versity have also tabulated the values of the viscosity and the thermal conductivity. They used
several references, the most important one being that of Vesovic et al. (1990).
In Figure 4a is shown a comparison of the predicted variations of the viscosity with tempera-
ture at a pressure of 10 MPa, based on data provided by Vesovic et al. (1990), Stephan and
Lucas (1979), and VDI (1988). It is shown in the graph that the viscosity values are in good
agreement with one other. Figure 4b shows a comparison of the predicted variations of thermal
conductivity with temperature based on data from Vesovic et al. (1990), Vargaftik (1975) and
VDI (1994). While the thermal conductivity values of Vesovic et al. (1990) and VDI (1994)
show good agreement, there are some differences from the Vargaftik data.
Vukalovich et al. (1963) analyzed several sets of experimental data on viscosity, thermal con-
ductivity and the Prandtl number of carbon dioxide at atmospheric pressure. They constructed
semi-theoretical equations for the dynamic viscosity using the Lennard-Jones potential and the

Figure 4a. Viscosity versus temperature Figure 4b. Thermal conductivity versus
at 10 MPa temperature at 10 MPa
 VOLUME 4, NUMBER 3, JULY 1998 286

Chapman-Enskog theory. With regard to the thermal conductivity, the authors concluded that
since the specific heat and the viscosity can be calculated accurately, one can use the Prandtl
number to determine the thermal conductivity accurately. Procedures for the direct measurement
of the Prandtl number are available. The maximum deviation in the viscosity predicted by their
formulas and the values that were experimentally observed was about ±5.2%, although a major-
ity of the points were within ± 0.5%. The deviation in the values for the thermal conductivity
was about ± 3%.
The thermodynamic properties, such as the density, enthalpy, entropy and specific heat are
given by Span and Wagner (1996). These authors compared the available correlations and
pointed out the inaccuracy and restrictions of those correlations. They then proposed a new
equation of state for carbon dioxide explicit in Helmholtz free energy A. The dimensionless
Helmholtz energy φ is defined as φ = A/(RT), where R is the gas constant and T is the thermody-
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

namic temperature. The dimensionless Helmholtz energy φ is commonly split into two parts; the
first part, φ°, takes into account ideal gas behavior, and the second part, φr, takes into account
the residual fluid behavior, as shown below:

o r
A ( ρ, T ) ⁄ ( RT ) = φ ( δ, τ ) = φ ( δ, τ ) + φ ( δ, τ ) (1)

where δ = ρ/ρc and τ = Tc /T.

The Helmholtz energy is a function of density and temperature and is in the form of a funda-
mental equation. Therefore, all the thermodynamic properties of a pure substance can be
obtained by combining derivatives of Equation (1). For example, the equations required to cal-
culate the pressure, entropy, and the enthalpy are given in Table 2.
The authors estimated the uncertainty of the equation in the range of ±0.03% to ±0.05% in
determining the density, ±0.03% to ±1% in determining the speed of sound, and ±0.15% to
±1.5% in determining the specific heat at constant pressure.

Table 2. Relation of Some Thermodynamic Properties to Helmholtz Function

Property and Common Relation to Reduced Helmholtz
Thermodynamic Definition Energy φ and Its Derivatives
Pressure p ( δ, τ ) r
---------------- = 1 + δΦ δ
p ( T, ρ ) = – ( ∂A ⁄ ∂ν ) T ρRT
Entropy ( δ, τ -)
s--------------- o r o r
= τ ( Φ τ + Φτ ) – Φ – Φ
s ( T, ρ ) = – ( ∂A ⁄ ∂T ) V R
Enthalpy h ( δ, τ ) o r r
---------------- = 1 + τ ( Φτ + Φ τ ) + δΦ δ
h ( T, p ) = A – T ( ∂A ⁄ ∂T ) V – ν ( ∂A ⁄ ∂ν )T RT

PHYSICAL FACTORS INFLUENCING FORCED CONVECTION

HEAT TRANSFER AT SUPERCRITICAL PRESSURES
Deterioration/Improvement of Heat Transfer in Supercritical Region
In the supercritical region, the great variations in the thermophysical properties of the fluid
cause the heat transfer coefficient to be greatly dependent on both the local bulk temperature and
the heat flux. As noted by Shitsman (1963), it is possible to distinguish a liquid-like region from
a gas-like region on the state diagram, even under supercritical conditions. Shitsman (1963)
pointed out that a deterioration of heat transfer coefficient occurs during in-tube heating of
supercritical carbon dioxide at high heat fluxes when the bulk temperature is below the critical
temperature and the wall temperature is above it. He noticed that the deterioration of heat trans-
 287 HVAC&R RESEARCH

fer begins in the fluid when its heat content is in the order of about 95 to 103 kJ/kg, which corre-
sponds to a liquid-like phase. During heating, there is a strong variation in temperature in the
radial direction over the cross section of the tube. As the tube is being heated, the fluid near the
wall changes from a liquid-like state to a gas-like state. The author observed that the deteriora-
tion during heating occurred abruptly when this “phase change” occurred, and likened the phe-
nomenon to the well-known boiling crisis observed during transition to film boiling. During
cooling it was observed by Shitsman (1963), Tanaka et al. (1971), and Krasnoshchekov et al.
(1970) that there was an improvement in heat transfer when the wall temperature was less than
the critical temperature and the fluid bulk temperature was greater than the critical temperature.
Qualitative explanations for the observed deterioration of heat transfer in heating and enhance-
ment during cooling have been provided by various investigators.
Krasnoshchekov et al. (1970) attributed the improvement of heat transfer during cooling to
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

the formation of a lower-temperature, liquid-like layer near the wall of the tube. This layer has
higher thermal conductivity than the bulk fluid. A converse explanation has been provided by
Petukhov et al. (1972) for the observed deterioration of heat transfer during heating. They
attribute the diminution of heat transfer to the formation of a thin film with lower thermal con-
ductivity than the bulk fluid in the wall-adjacent region. For a constant wall heat flux, the deteri-
oration in heat transfer coefficient results in a peak in the wall temperature.
Shiralkar and Griffith (1969) determined that the following factors decrease the heat transfer
coefficient of supercritical carbon dioxide:
1. Heat flux and flow rate: Deterioration gets worse as the ratio of heat flux to flow rate
increases.
2. Inlet enthalpy: The greatest deterioration is observed when the inlet enthalpy is
substantially below the critical enthalpy.
3. Upstream conditions: Flow instabilities reduce deterioration.
4. Pressure: Deterioration is largest when the pressure is close to critical pressure.
Shiralkar and Griffith (1969) did both experimental and theoretical work to confirm their
results. They solved the differential energy, momentum, and continuity equations along with the
equations for eddy diffusivity, suggested by Deissler (1950), Goldmann (1954), and van Driest
(1956), using a finite difference method. The authors used all three forms of the eddy diffusivity
and found that they all yielded similar results. They compared their calculated results with their
own experimental data for supercritical carbon dioxide and also with earlier work by Shitsman
(1963) using supercritical water as the fluid. They found that, at moderate heat fluxes, their math-
ematical model underpredicted the wall temperature at the inception of the experimental peak
wall temperature somewhat overpredicted at high heat fluxes. The authors also observed that their
prediction is not accurate in the post peak region. They attribute this to the additional mixing in
the core due to large density gradients that was not taken into account in their calculations.
Shiralkar and Griffith (1970) also studied the effect of swirl, inlet conditions, flow directions,
and tube diameter on the deterioration of heat transfer to supercritical fluids. They observed that
the swirl in the flow increases the heat transfer. They also pointed out that the location and the
extent of the deterioration is sensitive to the system geometry and the inlet subcooling.
Tanaka et al. (1971) performed experiments on supercritical carbon dioxide at high heat
fluxes. They analyzed the mechanism of forced convection heat transfer near the pseudocritical
region, where the properties of the fluid vary greatly. They found that the surface roughness of
the tube increases the heat transfer and reduces the deterioration.
Bourke and Pulling (1971) found that in the upstream part of a vertical heated tube carrying
an upward flow of carbon dioxide at supercritical pressure and subcritical inlet temperature,
there is a sharp deterioration in the heat transfer when heat fluxes are high. These authors sug-
 VOLUME 4, NUMBER 3, JULY 1998 288

gested that the deterioration was due to a reduction in the turbulence in the near-wall layer. The
low turbulence level near the wall is attributed to a reduction in shear stress due to buoyancy
forces arising from the density difference between the high-density core flow and the low-den-
sity near-wall flow, a phenomenon that was described by Hall and Jackson (1969). They pointed
out that further downstream, when the fluid temperature is well above the critical point, the heat
transfer improves substantially and finally approaches that of a normal superheated gas. They
suggested placing small rings inside the tube to increase turbulence upstream and hence improve
the heat transfer. Experimental studies by Petukhov et al. (1972) revealed deterioration of heat
transfer in both vertical and horizontal heated tubes. These investigators also attribute the deteri-
oration to reduced turbulence levels engendered by large density gradients. The reduction in
heat transfer was believed to occur when the layer of large density gradient encompasses the vis-
cous sublayer and the buffer layer. They observed that the heat transfer could be improved con-
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

siderably by the imposition of acoustic pressure pulsations on the flow.

Effect of Buoyancy Driven Secondary Flows on Heat Transfer in Tube Flow

At sufficiently low Reynolds numbers, buoyancy effects give rise to secondary flows that
cause the heat transfer in a horizontal tube to be enhanced at the bottom and to be reduced at the
top of the tube. The circumferentially averaged heat transfer coefficient is usually enhanced by
secondary flow. However, Petukhov and Polyakov (1974) pointed out that a local “worsening”
of heat transfer may occur, depending on the hydrodynamic and thermal characteristics of the
flow regime. According to them, for upward flow in heated vertical tubes, free convection
effects produce an enhancement in heat transfer for Gr/Re 2 > 0.6. They pointed out that deterio-
ration of heat transfer does occur for Gr/Re2 < 0.6 when the inlet temperature is subcritical and
the wall temperature is supercritical. Evidently, in this regime the reduction in near-wall turbu-
lence is sufficient to overcome any enhancement associated with secondary flows.
Adebiyi and Hall (1976) performed experiments to determine the effects of buoyancy. In a
vertical tube, localized temperature peaks were observed in the axial temperature during upflow.
These temperature peaks are associated with worsened heat transfer. During downflow, an
increase in the heat transfer was observed. The authors’ findings are consistent with the observa-
tions of other investigators that buoyancy effects occur both in horizontal tubes and in vertical
tubes. They also verified that in a horizontal tube, the heat transfer at the bottom of the tube is
enhanced due to the effect of buoyancy.
According to Hall and Jackson (1969), in a vertical tube buoyancy effects are negligible when

Grb Reb−2.7 < 5 × 10−6 (2)

Petukhov et al. (1974) proposed the following criterion under which the effect of buoyancy in
a horizontal tube may be ignored:

Grb Reb−2.75 Prb−0.5 [1 + 2.4Reb−1/8 (Prb2/3 – 1)]−1 < 3 × 10−5 (3)

HEAT TRANSFER CORRELATIONS FOR IN-TUBE FLOW

OF SUPERCRITICAL FLUID
Several experimental and theoretical investigations have been carried out to study the heat
transfer associated with the flow of supercritical carbon dioxide in tubes. Various correlations
based on the results of each investigator have been produced, differences with each other being
mainly with respect to the index of the variables. Most of these correlations are based on the
Petukhov-Popov-Kirilov equation for the circumferentially averaged, local Nusselt number:
 289 HVAC&R RESEARCH

ξ ⁄ 8 Re Pr
Nu o = --------------------------------------------------------------------- (4)
2⁄3
12.7 ξ ⁄ 8 ( Pr – 1 ) + 1.07

where the friction coefficient ξ is calculated by the correlation proposed by Filonenko:

–2
ξ = [ 0.79ln ( Re ) – 1.64 ] (5)

Equation (4) applies to the flow of a single-phase fluid where the thermophysical properties
are either constant or weakly varying.

Correlations for Heating

Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Based on their experimental results for the heating of supercritical carbon dioxide in tube
flows, Petukhov et al. (1961) developed a correlation for determining the Nusselt number. To
account for the variations of the viscosity, thermal conductivity, and specific heat, they proposed
the relationship:

0.35
µ b 0.11 k b – 0.33  c p 
Nu b = Nu ob  ------   ------   -------- (6)
 µw   kw   cpb 

Nuo is calculated using the Petukhov-Kirillov correlation from Equation (4), and

ib – iw
c p = ------------------ (7)
Tb – Tw

Krasnoshchekov and Protopopov (1966) conducted an experimental study to investigate the

heat transfer coefficient of supercritical carbon dioxide in a horizontal tube during heating, as
did Petukhov et al. (1961). These authors observed that the changes in the thermophysical prop-
erties over the cross section of the tube are great and hence Equation (4), which does not con-
sider any variations in the thermophysical properties in the cross section of the tube, cannot be
used. They assumed that, in the supercritical region, the Nusselt number is a function of the Rey-
nolds number and the Prandtl number, just as it is in the single-phase region. The variations of
the thermophysical properties in the cross-section of the tube were taken into account by intro-
ducing correction factors for density and specific heat ratio. They proposed the correlation:

n
ρ w 0.3  c p 
Nu b = Nu o  ------  -------- (8)
 ρb   c pb

where Nuo is calculated using the bulk temperature of the fluid, and n, a function of Tw/Tc, var-
ies, as shown in Figure 5.
Yamagata et al. (1971) experimentally investigated the heat transfer characteristics of super-
critical water and proposed a correlation based on their study. They also compared the predic-
tions of the correlations proposed by Petukhov et al. (1961), Krasnoshchekov and Protopopov
(1966), and Styrikowitsch et al. (1959), against their own correlation as well as the experimental
data and found that their correlation was the closest to the experimental data. The correlation
that they proposed was:
 VOLUME 4, NUMBER 3, JULY 1998 290

0.85 0.8
Nu b = 0.0135Re b Pr b F c (9)

where
for E > 1 Fc = 1
– 0.05 n1
for 0 ≤ E ≤ 1 F c = 0.67Pr m ( c p ⁄ c pb )
n2
for E < 0 F c = ( c p ⁄ c pb )
and n1 = −0.77(1 + 1/Prm) + 1.49
n2 = 1.44(1 + 1/Prm) − 0.53
Here E > 1 if the fluid is considered to be a liquid, E < 0 if it is considered a vapor, and 0 ≤ E ≤
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

1 if it is considered to be a vapor near the wall and a liquid in the core.

Krasnoshchekov and Protopopov (1972) proposed a generalized relationship for calculation
of heat transfer at supercritical pressures, that was based on an analysis of experimental data for
carbon dioxide and water at pressures of P/Pc = 1.02 – 1.45. The relationship is:
n
 cp  ρw m
Nu b = Nu ob  --------  ------ (10)
 c pb  ρ b 

where m = −0.35 − 0.05π, π = P/Pc, and n = f (Tw/Tb) is given by Figure 5.

Figure 5. Variation of n with wall, bulk, and supercritical temperatures

(Krasnoshchekov and Protopopov 1966)

Furthermore, Krasnoshchekov et al. (1972) measured the local heat transfer coefficients dur-
ing turbulent flow of carbon dioxide in a heated test section. They investigated the variation of
the heat transfer coefficient in the entrance length region where the flow develops. For this
region, they proposed a slightly modified version of the correlation given in Equation (10).
n
 c p ρ w 0.3 x
Nu w = Nu ow  -----  ------ f  --- (11)
 c b  ρ b   d

where f(x/d) = 1 for x/d > 15 and f(x/d) = 0.95 + 0.95 (d/x)0.8 for 2 < f(x/d) < 15
 291 HVAC&R RESEARCH

Ghajar and Asadi (1986) studied some of the correlations that were developed for forced con-
vective heat transfer in the supercritical region. Based on their studies, they concluded that the
correlation given by Jackson and Fewster (1975) was the most accurate one. The correlation is:

b c ρw  cp 
d n
Nub = aReb Pr b  ------  -------- (12)
 ρ b   c pb

Here a, b, c, and d are curve-fitted constants. The exponent n is prescribed as follows:

for Tb < Tw ≤ Tc and Tw > Tb ≥ 1.2Tc, n = 0.4
for Tb ≤ Tc < Tw, n = 0.4 + 0.2 (Tw/Tc – 1)
for Tc ≤ Tb ≤ Tc and Tb < Tw , n = 0.4 + 0.2(Tw/Tc – 1)[1 – 5(Tb/Tc – 1)]
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Ghajar and Asadi (1986) came to the conclusion that the convective heat transfer in the super-
critical region can be predicted by a Dittus-Boelter type heat transfer correlation that has prop-
erty ratios to account for the large variations in the physical properties in this region.
In a recent publication, Griem (1995) proposed a new correlation to determine the heat trans-
fer coefficient of supercritical carbon dioxide. This correlation is based on a selective method
for determination of the representative specific heat. His correlation for the heat transfer in a
vertical smooth tube at and near the pseudocritical pressure is:

Nu = 0.0169Re0.8356 Pr0.432 (13)

Protopopov (1977) characterized the effect of free convection during heating in various
regions. The effect of free convection heat transfer was determined by finding the ratio of the
Grashof and the square of the Reynolds number. For the upward flow in a vertical tube he pro-
posed the following variation to the correlation given by Equation (14):
Nu = Nu c ϕ ( K )
(14)
2
K = Gr ⁄ Re
The value of Nuc can be calculated using Equation (8). Values of the function ϕ(K) are tabu-
lated for values of K between 0.01 to 0.4. In this range, the value of the function is everywhere
slightly below unity. For K > 0.4 in the fully developed region, the function ϕ(K) is substantially
below unity.
Additional information about the effect of free convection can be found under the section that
describes the effect of buoyancy.

Correlations for Cooling

During cooling of the supercritical carbon dioxide, which is the main point of interest for refrig-
eration, it was shown that there is an increase in the heat transfer in comparison to the heat transfer
for the flow of a constant property fluid. This section of the paper discusses the correlations that
were proposed to capture the effect of varying heat transfer coefficient in the supercritical region.
Krasnoshchekov et al. (1970) carried out an experimental study of the heat transfer character-
istics during turbulent flow in a round tube with carbon dioxide at supercritical pressures under
cooling conditions. They proposed the following correlation:
m
ρw n  cp 
Nu w = Nu ow  ------  -------- (15)
 ρ b   c pw

where m and n are constants that vary as shown in Figure 6.

 VOLUME 4, NUMBER 3, JULY 1998 292
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Figure 6. Variation of indexes in correlation proposed by Krasnoshchekov et al. (1970)

Curves for pressure: 12 MPa – 1: 10 MPa – 2: 9 MPa – 3: 8.5 MPa – 4: 8 MPa – 5: 7.845 MPa

The authors compared results using the correlation given in Equation (15) with the experi-
mental results and found that 91% of the experimental values of Nuw were within ±20% of the
calculated values. They also compared the correlation to experimental results by Tanaka et al.
(1971) and found satisfactory agreement.
Baskov et al. (1977) experimentally investigated the variation of local heat transfer coeffi-
cients at high Reynolds numbers in cooling conditions and found no appreciable influence of
free convection heat transfer. They compared the value of Nuw using the correlation shown in
Equation (15) and found that their experimental value of Nuw was lower. They concluded that
the correlation proposed by Krasnoshchekov et al. (1966) was for horizontal tubes and that it
was not valid for their experiment that used a vertical tube. Therefore, they proposed the follow-
ing new correlation for vertical tubes:

m
 cp  ρb n
Nu w = Nu ow  --------  ------ (16)
 c pw  ρ w

The values of the exponents m and n are to be determined from tabular data provided in the
paper. The authors found that 90% of their experimental results were within ±15% of the values
predicted by their correlation.
Petrov and Popov (1985) solved the governing conservation equations numerically and devel-
oped the following formula:

n
Nu w mq w  c p 
- =  1 – ----------
------------ -  -------- (17)
Nu ow  ρω   c pw

where

ρω )
 0.66 – k ( q w ⁄ if c p ⁄ c pw ≤ 1
n = 
 0.9 – k ( q w ⁄ ρω ) if c p ⁄ c pw > 1

For more details see the later section on Numerical Methods for Determining Heat Transfer
Coefficients.
 293 HVAC&R RESEARCH

COMPARISON OF HEAT TRANSFER CORRELATION

DEVELOPED FOR COOLING
During the literature search only three correlations were found that were developed strictly to
calculate the heat transfer coefficient of supercritical carbon dioxide during gas cooling. A com-
parative study of these correlations was carried out, together with an additional textbook correla-
tion in order to quantify the deviations in their predictions. The physical parameters used for this
study are shown in Table 3.

Table 3. Parameters for Which Heat Transfer Coefficient Has Been Plotted
Temperature Temperature
Tube I.D. Tube O.D. Mass flow Pressure Range of CO2 Range of Water
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

4.572 mm 6.35 mm 0.03 kg/s 10 MPa 120 – 32°C 17 – 32°C

Figure 7 shows the predictions of the supercritical heat transfer coefficient as a function of
temperature. Of the four correlations compared here, only the Krasnoshchekov-Kuraeva-Proto-
popov [Equation (15)] was developed for cooling in horizontal tubes. Baskov et al.(1977)
[Equation (16)] developed their correlation for vertical tube cooling, and compared their results
to the earlier correlation of Krasnoshchekov et al. (1970). They found that the heat transfer
coefficients calculated with their own correlation were about 20% lower than the values
obtained with the Krasnoshchekov et al. (1970) correlation. This can also be observed from Fig-
ure 7. Hence it can be concluded that there is a difference in the available heat transfer correla-
tions for vertical and horizontal tubes. Petrov and Popov (1985a) developed their correlation
[Equation (17)] for cooling in a vertical tube. The heat transfer coefficient calculated with their
correlation is in good agreement with the Baskov et al. (1977) correlation. The correlation for

Figure 7. Comparison of three cooling correlations and

the single-phase Petukhov-Popov-Kirilov correlation
 VOLUME 4, NUMBER 3, JULY 1998 294

the heat transfer coefficient of single-phase fluids (Petukhov-Popov-Kirilov), as typically seen

in heat transfer textbooks [Equation (4)] has also been plotted. The heat transfer coefficient cal-
culated with the Petukhov-Popov-Kirilov equation varies quite a bit from the values calculated
with the other correlations and is therefore not suitable for predicting the heat transfer coeffi-
cient at supercritical parameters.

COEFFICIENT OF FRICTION FOR IN-TUBE FLOW

OF A SUPERCRITICAL FLUID
The coefficient of friction enables one to calculate the pressure drop in the tube. The most
popular correlation for determining the friction coefficient for the internal flow in a tube is the
one proposed by Filonenko [Equation (5)]. This correlation was developed for the flow in a tube
for single-phase fluids. Some investigators have used the Filonenko correlation [Equation (5)]
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

for fluids in the supercritical region as well.

Kuraeva and Protopopov (1974) proposed a new relationship for determining the friction
coefficient for supercritical fluids and verified their correlation against their own experimental
data. In the absence of free convection effects, they found that for Re < 10 5, the friction coeffi-
cient ξ0 agrees with the Filonenko equation. For Re > 105, ξ0 = 0.02. The authors considered the
effect of free convection on friction by taking into account the ratio of Gr and Re2. Based on
their experimental data, they obtained the following formulas:
For Gr/Re2 < 5 ×10−4

ξ = ξ0 (µw /µl )0.22 (18)

For 5 × 10-4 < Gr/Re2 ≤ 0.3

ξ = ξ0 (µw /µl )0.22 × 2.15(Gr/Re2 )0.1

It was observed that 80% of the experimental data was within ±20%.
The energy and momentum equations are coupled and, hence, later researchers investigated
the pressure drop and heat transfer simultaneously. Petukhov et al. (1979) experimentally inves-
tigated the drag (local and average) in turbulent flow of supercritical fluids. A one-dimensional
flow model is not accurate to determine the local and the average drag. However, a one-dimen-
sional momentum balance can be used to separate the measured changes in static pressure into
the components of wall friction, flow acceleration and the hydrostatic pressure.
Medvetskaya (1981) noted that there is a relationship between the heat transfer coefficient and
the ratio of the frictional and inertial drag. This ratio takes into account the losses due to the defor-
mation of the flow density and velocity profiles. Razumovskiy et al. (1985) proposed the follow-
ing correlation, which is valid for Gr/Re2 < 0.003 to determine the local frictional drag force

µ w ρw 0.18
ξ = ξ o  ------ ------  (19)
 µb ρb 

where ξo is given by Filonenko’s Equation (5). The experimental data is within 20% of that cal-
culated using this formula, and the rms deviation is 9.1%.
Kurganov et al. (1986) used a pitot microtube and thermocouples along several locations
along the tube to collect thermal and hydraulic data. They then used this data in the integral
equations of momentum and energy to determine the local friction coefficient and heat transfer
for an ascending flow of supercritical carbon dioxide under conditions of constant heat flux.
 295 HVAC&R RESEARCH

Kurganov and Kaptilnyi (1993) compared the cases in which buoyancy either assisted or
opposed the flow. They observed deterioration in heat transfer in upward flow in a heated tube
and an improvement in heat transfer in downward flow. The deterioration in heat transfer during
upward was attributed to the formation of a near-wall fluid layer with reduced turbulence.
Popov and Petrov (1985b) developed Equation (21) numerically (discussed in the next sec-
tion) to determine the coefficient of friction.

NUMERICAL METHODS FOR DETERMINING

HEAT TRANSFER COEFFICIENTS
The heat transfer coefficient can also be determined by solving the governing conservation
equations of continuity, momentum and energy. Since these equations are non-linear, coupled,
partial differential equations, analytical solutions are precluded and the solutions have to be
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

obtained numerically. In addition, since the flow in the tube is generally strongly turbulent, the
use of a turbulence model, such as a K-ε model, is necessary. Information presented in the liter-
ature is inconclusive with regard to the effect of thermophysical property variations on the tur-
bulence. Therefore, careful validation of turbulence models is needed before they can be applied
with confidence to the prediction of supercritical heat transfer. The standard K-ε model uses a
coarse grid near the wall and uses wall functions. Consequently, the property changes near the
wall, which are very substantial, may not be fully captured by the model. Various investigators
have used a one-equation turbulence model developed by Petukhov and Medvetskaya (1978)
that utilizes a different conservation equation for K, and a mixing length model developed by
Bellmore and Reid (1983) that considers the effect of density fluctuations.
Bellmore and Reid (1983) developed a mixing length model to predict turbulent flow of
supercritical fluids. Hendricks et al. (1966) noted that the data could be grouped using the ther-
mal expansion coefficient (ß). The compressible time-averaged governing conservation equa-
tions have terms involving the fluctuating density. Bellmore and Reid (1983) used the thermal
expansion coefficient (ß) to relate the turbulent density fluctuations to the energy fluctuations in
the flow field. Hence, they proposed the following mixing length model, which takes into
account the fluctuations of density:

β l m ∂i β l m ∂i 2
- 1 –  ----- ------ -----  –  ----- ------ ----- 
2 ∂u
µ t = pl m ----- (20)
∂y c p Prt ∂y c p Pr t ∂y

The authors found their model to be in good agreement with experimental results over a wide
range of mass flux, heat flux and inlet enthalpy. They also showed that the radial fluid density
variations have a significant effect on turbulent transport mechanisms in fluids near the critical
region. In their work, Bellmore and Reid did not account for the variations of thermal conductiv-
ity, viscosity and specific heat in the wall-adjacent layer. In addition, their model is not valid at
the velocity extremum where the velocity gradient is zero.
Melik-Pashaev (1966) calculated the convective heat transfer at supercritical pressures. Dur-
ing turbulent flow of a fluid, fluctuations in the temperature give rise to variations in the physi-
cal properties of the fluid. According to Melik-Pashaev (1966), the most important fluctuation of
the physical properties is that of density as the transport of energy depends on the density along
with the velocity and the enthalpy. The fluctuations in the viscosity and the thermal conductivity
are of less importance, according to the author, as their effect is felt near the wall where the
effect of turbulence is less. Melik-Pashaev (1966) used a mixing length model of turbulence that
he developed and is similar to the one proposed by Deissler (1950). The author divided the
cross-sectional area into two regions: the viscous sub-layer (which he calls viscous substrate)
 VOLUME 4, NUMBER 3, JULY 1998 296

and the turbulent core. In the viscous sub-layer the fluctuations in density are not great and are
neglected. In the turbulent core, molecular transport is neglected. The author separated the inte-
gral equation of energy into two parts corresponding to the two-layer flow model and solved it
numerically.
Shlykov et al. (1971) calculated the heat transfer for the turbulent flow of supercritical water.
The authors calculated the variation of the turbulent heat transfer coefficient due to the variation
of the physical properties of water. They formed a non-dimensional integral equation, which
they solved numerically using the successive approximation method. The authors found that if
the wall temperature were less than the critical temperature and the fluid temperature were
greater than the critical temperature, an improvement of heat transfer was observed. A deteriora-
tion of heat transfer was seen during heating if the wall temperature was above the critical tem-
perature and the fluid temperature was below the critical temperature.
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Schnurr et al. (1976) developed a two-dimensional numerical method to predict the heat
transfer near the critical region for the turbulent flow of fluids. They solved the conservation
equations, which were in the von Mises coordinate system, numerically using a variation of the
Patankar-Spalding (1970) method. The authors calculated the effective Prandtl number by
assuming the effective conductivity of the fluid is composed of laminar and turbulent contribu-
tions. They observed that the entrance effects continue further downstream then in the constant
property flows. This effect is enhanced as the heat flux is increased.
Petrov and Popov (1985b) numerically solved the differential momentum, energy, and conti-
nuity equations using the explicit finite difference method for calculating the heat transfer coef-
ficient and the friction coefficient in a tube that was being cooled. They used the Prandtl mixing
length model of turbulence. The authors divided the friction coefficient into two components,
one due to the influence of inertia and the other due to the influence of hydraulic resistance.
These components are given as

τ r p
ξw = ξw + ξw

–2 ρw µw
s
ξ w = ( 1.82 ln Re w – 1.64 ) ------  ------ where  s = 0.023 -------
r qw 
ρb µb   -
ρω (21)
p dp ⁄ dx
ξ w = ---------------------------------
2
ρ ω ⁄ ( 2ρ w d )

They also proposed the relationship to calculate the heat transfer coefficient already described in
Equation (17).
Renz and Bellinghausen (1986) used the K-ε turbulence model and numerically solved for the
heat transfer in a vertical tube. They used the axisymmetric two-dimensional forms for the con-
tinuity, the x-momentum, and the energy equations and solved them using the finite difference
method. They found that, for upflow in a vertical pipe, there was a large reduction in heat trans-
fer coefficient in the pseudocritical region. For downflow, a slight enhancement was predicted
for the pseudocritical region. They compared the theoretical values with measured wall temper-
atures in upflow and found that the predictions were too high. Nevertheless, the predictions were
found to be in qualitative agreement with experimental data. They attributed the difference to
additional turbulence that might be caused by the changes in the fluid properties.
Popov and Valueva (1990) numerically solved the energy, momentum and continuity equa-
tion along with the Prandtl mixing length model of turbulence for supercritical carbon dioxide in
a vertical tube during heating. They predicted that the effect of the density variations was to
 297 HVAC&R RESEARCH

increase the turbulent viscosity near the wall, where there is a sizable velocity gradient, and
decrease it near the core, where the velocity gradient is negligible.
Koshizuka et al. (1995) numerically analyzed the deterioration of heat transfer for upward
flow of supercritical water in a vertical tube. Their calculation was based on the parabolic solver
of Patankar-Spalding (1972) in the r-z dimensions, and the K-ε turbulence model by Jones-
Launder was used. The results were compared to the experimental work done be Yamagata et al.
(1971) The authors indicate that, at high flow rates and high heat fluxes, heat transfer deteriora-
tion takes place downstream in the tube as the thickness of the viscous sub layer increases. Due
to the increase in the thickness of the viscous sub-layer, the near-wall turbulence energy dimin-
ishes, leading to a reduction in heat transfer. At low flow rates turbulence in the near-wall layer
is diminished due to buoyancy effects, and the heat transfer is diminished correspondingly.
Zhou and Krishnan (1995) numerically solved the governing conservation equations for lami-
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

nar flow, and with a low Reynolds number K-ε turbulence model for turbulent flow. The
low-Reynolds number K-ε model uses a fine grid near the wall and does not require wall func-
tions. In laminar flow they compared their results with the experimental data of Nishikawa et al.,
and found the two to be in good agreement. They came to the conclusion that in a vertically ori-
ented tube buoyancy forces play an important role in influencing heat transfer even at Reynolds
numbers as high as 100,000.
Lee and Howell (1998) performed a numerical simulation to study the turbulent heat transfer
of fluids near their critical point for a vertically oriented tube during heating. In the critical
region the Prandtl number varies and hence the ratio of the diffusivities of the momentum and
energy vary from one region to another. The effect of property variation on the coefficients and
constants in the K-ε turbulence model is not known. Thus the authors used the Bellmore and
Reid (1983) mixing length model, which takes into account the effect of density fluctuations.
They used the SIMPLE solution algorithm with a staggered grid system. The authors compared
their results with experimental data of Wood and Smith (1964) and found the two to be in good
agreement.

SUMMARY
This paper comprises a comprehensive literature review of the heat transfer and pressure drop
characteristics of supercritical carbon dioxide during in-tube. The results of this review can be
summarized as follows. The heat transfer in the supercritical region varies greatly with tempera-
ture. There has been some research on the heat transfer in this region but none of it has been
related to refrigeration or air-conditioning applications. Compared to the heating process of
supercritical carbon dioxide, much less work has been done in determining the heat transfer and
pressure drop during the cooling process of supercritical carbon dioxide, which is required for
the design of a gas cooler. Most of the research on this topic was done by Russian scientists in
the 1960s and 70s. A few correlations were developed, as shown by Yamagata et al. (1971), but
these correlations display significant differences amongst themselves. Some authors observed
improvement of heat transfer during cooling and a deterioration of heat transfer during heating
for some particular values of heat flux. Baskov et al. (1977) showed that the heat transfer is dif-
ferent for horizontal and vertical tubes. The effect of turbulence on heat transfer has not been
fully understood to date. To extend and refine the knowledge base, accurate temperature and
pressure measurements are needed in experimental studies. These are a few of the difficulties
encountered while studying the heat transfer in the supercritical region. The heat transfer can be
calculated numerically by solving the energy, momentum and continuity equations using a tur-
bulence model. However, information presented in the literature is inconclusive with regard to
the effect of thermophysical property variations on the turbulence. Therefore, careful validation
of turbulence models is needed before they can be applied with confidence to the prediction of
 VOLUME 4, NUMBER 3, JULY 1998 298

supercritical heat transfer. More research should be done in determining the heat transfer and
pressure drop during cooling of supercritical carbon dioxide as experienced in refrigeration and
air conditioning applications.

ACKNOWLEDGMENTS
This work is sponsored by ASHRAE research project RP-913. The support of ASHRAE is
gratefully acknowledged.

NOMENCLATURE
cp = specific heat ξ = coefficient of friction
d = inside tube diameter µ = viscosity
g = acceleration due to gravity ρ = density
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Gr = Grashof number ρω = mass flow rate

i = enthalpy Subscripts
k = thermal conductivity b = bulk of fluid
K = kinetic energy c = critical
Nu = Nusselt number m = supercritical
P = pressure o = single phase
Pr = Prandtl number w = inside wall of tube
Re = Reynolds number Superscripts
T = temperature p = inertia
β = thermal expansion coefficient r = flow
ε = dissipation in kinetic energy τ = total

REFERENCES
Adebiyi, G.A., and W.B. Hall. 1976. Experimental Investigation of Heat Transfer to Supercritical Pressure
Carbon Dioxide in a Horizontal Pipe. International Journal of Heat and Mass Transfer 19: 715-720.
Baskov, V.L., I.V. Kuraeva, and V.S. Protopopov. 1977. Heat Transfer with the Turbulent Flow of a Liq-
uid at Supercritical Pressure in Tubes under Cooling Conditions. Teplofizika Vysokikh Temperatur 15
(1): 96-102.
Bellmore, C.P., and R.L. Reid. 1983. Numerical Prediction of Wall Temperatures for Near-Critical
Para-Hydrogen in Turbulent Upflow inside Vertical Tubes. Transactions of the ASME 105: 536-541.
Bourke, P.J., and D.J. Pulling. 1971. Experimental Explanation of Deterioration in Heat Transfer to Super-
critical Carbon Dioxide. ASME Publication No. 71-HT-24.
Bourke, P.J., D.J. Pulling, L.E. Gill, and W.H. Denton. 1970. Forced Convective Heat Transfer to Turbu-
lent CO2 in the Supercritical Region. International Journal of Heat and Mass Transfer 13: 1339-1348.
Deissler, R.G. 1950. Analytical and Experimental Investigation of Adiabatic Turbulent Flow in Smooth
Tubes. NASA TN-2138.
Gnielinski, V. 1976. New Equation for Heat and Mass Transfer in Turbulent Pipe and Channel Flow. Int.
Chem. Engg. 16: 359-368.
Goldmann, K. 1954. Heat Transfer to Supercritical Water and Other Fluids with Temperature Dependent
Properties. Chemical Engineering Progress, Symposium Series, Nuclear Engineering Part 1 50(11).
Ghajar, A.J., and A. Asadi. 1986. Improved Forced Convective Heat-Transfer Correlations for Liquids in
the Near-Critical Region. AIAA Journal 24(12): 2030 – 2037.
Griem, H. 1995. A New Procedure for the Prediction of Forced Convection Heat Transfer at Near and
Supercritical Pressure. Wärme und Stoffübertragung—Thermo and Fluid Dynamics 31(5): 301-305.
Hall, W.B. 1971. Heat Transfer Near the Critical Point. Advances in Heat Transfer 7: 1-86
Hall, W.B., and J.D. Jackson. 1969. Laminarization of a Turbulent Pipe Flow by Buoyancy Forces. ASME
Paper 69-HT-55.
Hendricks, R.C., V.M. Yeroshenko, L.A. Yaskin, and A.D. Starostin. 1979. Bulk Expansion Factors and
Density Fluctuations in Heat and Mass Transfer. Paper B/119 In Proc. XV International Congress of
Refrigeration, Venice, Italy.
 299 HVAC&R RESEARCH

Herreman, W., W. Grevendonk, and A. DeBock. 1970. Shear Viscosity Measurements of Liquid Carbon
Dioxide. Journal Chem. Phys. 53: 185-189.
Hsu, Y., and J.M. Smith. 1961. The Effect of Density Variation on Heat Transfer in the Critical Region.
Transactions of the ASME (May) 176-182.
Jackson, J.D., and J. Fewster. 1975. AERE – R8158, Harwell, UK.
Kennedy, G.C. 1954. Pressure-Volume-Temperature Relations in Carbon Dioxide at Elevated Tempera-
tures and Pressures. Am. Journal Sci. 252: 225-235.
Krasnoshchekov, E.A., and V.S. Protopopov. 1966. Experimental Study of Heat Exchange in Carbon
Dioxide in the Supercritical Range at High Temperature Drops. Teplofizika Vysokikh Temperatur 4(3):
389-398.
Krasnoshchekov, E.A., I.V. Kuraeva, and V.S. Protopopov. 1970. Local Heat transfer of Carbon Dioxide at
Supercritical Pressure Under Cooling Conditions. Teplofizika Vysokikh Temperatur 7(5) : 922-930.
Krasnoshchekov, E.A., and V.S. Protopopov. 1972. A Generalized Relationship for Calculation of Heat
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Transfer to Carbon Dioxide at Supercritical Pressure. Teplofizika Vysokikh Temperatur 9(6): 1314.
Krasnoshchekov, E.A., V.S. Protopopov, I.A. Parkhovnik, and V.A. Silin. 1972. Some Results of an
Experimental Investigation of Heat Transfer to Carbon Dioxide at Supercritical Pressure and Tempera-
ture Heads of up to 850 C. Teplofizika Vysokikh Temperatur 9(5): 1081-1084.
Kiyama, R., and T. Mikita. 1951. A New Simple Viscometer for Compressed Gases and Viscosity of Car-
bon dioxide. Rev. Phys. Chem. Japan Vol. 21: 63-68.
Kochizuka, S., N. Takano, and Y. Oka. 1995. Numerical Analysis of Deterioration Phenomenon in Heat
Transfer to Supercritical Water. International Journal of Heat and Mass Transfer 38(16): 3077-3084.
Koppel, L.B., and J.M. Smith. 1961. Turbulent Heat Transfer in the Critical Region. ASME International
Developments in Heat Transfer Part 3: 585-590.
Kuraeva, I.V., and V.S. Protopopov. 1974. Mean Friction Coefficients for Turbulent Flow of a Liquid at a
Supercritical Pressure in Horizontal Circular Tubes. Teplofizika Vysokikh Temperatur 12(1): 218-220.
Kurganov, V.A., V.B. Ankudinov, and A.G. Kaptilnyi. 1985. Experimental Study of Velocity and Temper-
ature Fields in an Ascending Flow of Carbon Dioxide at Supercritical Pressure in a Heated Vertical
Pipe. Teplofizika Vysokikh Temperatur 24(6): 1104-1111.
Kurganov, V.A., and A.G. Kaptilnyi. 1993. Flow Structure and Turbulent Transport if a Supercritical Pres-
sure Fluid in a Vertical Heated Tube Under the Conditions of Mixed Convection. Experimental Data.
International Journal Heat and Mass Transfer 36(13): 3383-3392
Lee, S.H., and J.R. Howell. 1998. Turbulent Developing Convective Heat Transfer in a Tube for Fluids
Near the Critical Point. Int. J. Heat Mass Transfer 41(10): 1205-1218.
Medvetskaya, N.V. 1981. Heat Transfer and Drag in Pipe Flows of Single-Phase Working Fluids at Trans-
critical Pressures. Author’s abstract of candidate thesis, Moscow.
Melik-Pashaev. 1966. Calculation of Convective Heat Transfer at Supercritical Pressure. Teplofizika
Vysokikh Temperatur 4(6): 853-864.
Patankar, S.V., and D.B. Spalding. 1970. Heat and Mass Transfer in Boundary Layers, 2nd ed. London:
Intertext Books.
Patankar, S.V., and D.B. Spalding. 1972. A Calculation Procedure for Heat, Mass and Momentum Transfer
in Three-Dimensional Parabolic Flows. Int. J. Heat Mass Transfer 15: 1787-1806.
Petrov, V.P., S.P. Beschastnov, and V.I. Belozeror. 1978. Natural Convection Heat Transfer from Super-
critical CO2 to a Cooled Horizontal Cylinder. Heat Transfer-Soviet Research 10(5): 150-153.
Petrov, N.E., and V.N. Popov. 1985a. Heat Transfer and Resistance of Carbon Dioxide Being Cooled in the
Supercritical Region. Thermal Engineering 32(3): 131-134.
Petrov, N.E., and V.N. Popov. 1985b. Heat Transfer and Turbulent Flow of Cooled Helium with Supercrit-
ical State Parameters in a Tube. Teplofizika Vysokikh Temperatur 27(3): 506-514.
Petukhov, B.S. 1968. Heat Transfer in a Single-Phase Medium under Supercritical Conditions. Teplofizika
Vysokikh Temperatur 6(4): 732-742.
Petukhov, B.S., E.A. Krasnoshchekov, and V.S. Protopopov. 1961. An Investigation of Heat Transfer to
Fluids Flowing in Pipes under Supercritical Conditions. ASME International Developments in Heat
Transfer Part 3: 569-78.
 VOLUME 4, NUMBER 3, JULY 1998 300

Petukhov, B.S., V.S. Protopopov and V.A. Silin. 1972. Experimental Investigation of Worsened Heat
Transfer Conditions with the Turbulent Flow of Carbon Dioxide at Supercritical Pressure. Teplofizika
Vysokikh Temperatur 10(2): 347-354.
Petukhov, B.S., and A.V. Polyakov. 1974. Boundaries of Regimes with “Worsened” Heat Transfer for
Supercritical Pressure of Coolant. Teplofizika Vysokikh Temperatur 12(1): 221-224.
Petukhov, B.S., V.A. Kurganov, V.B. Ankudinov, and A.G. Kaptilliyi. 1980. Effect of Dissolved Gas on
Heat Transfer during Turbulent Pipe Flow of Carbon Dioxide at Supercritical Pressure. Teplofizika
Vysokikh Temperatur 23(4) : 742-747.
Petukhov, B.S., V.A. Kurganov, V.B. Ankudinov, and V.S. Grigoriev. 1986. Experimental Investigation of
Drag and Heat Transfer in a Turbulent Flow of Fluid at Supercritical Pressure. Teplofizika Vysokikh
Temperatur 18(1): 100-111.
Petukhov, B.S., A.F. Polyakov, V.A. Kulcohnov, and Y.L. Sheckler. 1974. Turbulent Flow and Heat
Transfer in Horizontal Tubes with Substantial Influence of Thermogravitational Forces. Proc. 5th Int.
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Heat Transfer Conf. Tokyo, Paper No. NC 4.8 ASME.

Petukhov, B.S., and N.V. Medvetskaya. 1978. Turbulent Flow and Heat Exchange in Vertical Pipes Under
Conditions of Strong Influence of Upward Forces. Teplofizika Vysokikh Temperatur 16(4): 778-786.
Petukhov, B.S., and N.V. Medvetskaya. 1979. Turbulent Flow and Heat Transfer in Heated Tubes for
Single-Phase Heat Carriers with Near-Critical Parameters. Teplofizika Vysokikh Temperatur 17(2):
343-350.
Polyakov, A.F. 1991. Heat Transfer under Supercritical Pressures. Advances in Heat Transfer 21: 1-53.
Popov, V.N., and N.E. Petrov. 1989. Calculation of Heat Transfer and Flow Resistance in the Turbulent
Pipe Flow of Cooled Carbon Dioxide in the Supercritical Region. Teplofizika Vysokikh Temperatur
23(2): 309-316.
Popov, V.N., and E.P. Valueva. 1990. Distinctive Features of Heat exchange During Turbulent Forced and
Mixed Convection of a Fluid in the Supercritical Region. Atomnya Energiya 70(5): 326-335.
Protopopov, V.S. 1977. Generalizing Relations for the Local Heat-Transfer Coefficients Turbulent Flows
of Water and Carbon Dioxide at Supercritical Pressure in a Uniform Heated Circular Tube. Teplofizika
Vysokikh Temperatur 15(4): 815–821.
Razumovskiy, V.G., A.P. Ornatskiy, Y.E.M. Mayevskiy, and N.V. Igol’nikova. 1985. Determination of
Drag Coefficients in Turbulent Flow of Water at Supercritical Pressures in Smooth Channels. Fluid
Mechanics – Soviet Research 14(5): 1-6.
Renz, U., and R. Bellinghausen. 1986. Heat Transfer in a Vertical pipe at Supercritical Pressure. Heat
Transfer, Proceedings of the International Heat Transfer Conference 8th 3: 957-962.
Rice, O.K. 1955. Chapter 5E, Critical Phenomena. Thermodynamics and Physics of Matter, F.D. Rossini,
ed. Princeton, NJ, Princeton University Press.
Robinson, D.M., and E.A. Groll. 1998. Efficiencies of Transcritical CO2 Cycles With and Without an
Expansion Turbine. Accepted for publishing in the International Journal Refrigeration.
Sastry, V.S., and N.M. Schnurr. 1975. An Analytical Investigation of Forced Convection Heat Transfer to
Fluids Near the Thermodynamic Critical Point. Transactions of the ASME: 226-230.
Schnurr, N.M., V.S. Sastry, and A.B. Shapiro. 1976. A Numerical Analysis of Heat Transfer to Fluids Near
the Thermodynamic Critical Point Including the Thermal Entrance Region. Transactions of the ASME:
609-615.
Shiralkar, B.S., and P. Griffith. 1969. Deterioration in Heat Transfer to Fluids at Supercritical Pressure and
High Heat Fluxes. ASME Journal of Heat Transfer (Feb): 27-36.
Shiralkar, B., and P. Griffith. 1970. The Effect of Swirl, Inlet Conditions, Flow Direction, and the tube
diameter on the Heat Transfer to Fluids at Supercritical Pressure. ASME Journal of Heat Transfer
(Aug): 465-471.
Shitsman, M.E. 1963. Impairment of the Heat Transmission at Supercritical Pressures. Teplofizika Vyso-
kikh Temperatur 1(2): 267-275.
Shitsman, M.E. 1974. Heat Transfer to Supercritical Helium, Carbon Dioxide, Water: Analysis of Thermo-
dynamic and Transport Properties and Experimental Data. Cryogenics (Feb): 77-83.
 301 HVAC&R RESEARCH

Shykov, Y.P., A.N. Koblakov, and A.D. Leongardt. 1971. Theoretical Calculation of Heat Transfer During
the Turbulent Flow of Water at Supercritical Conditions. Teplofizika Vysokikh Temperatur 9(4):
765-770.
Span, R., and W. Wagner. 1996. A New Equation of State for Carbon Dioxide Covering the Fluid Region
from the Triple-Point Temperature to 1100 K at Pressures up to 800 MPa. Journal of Physical and
Chemical Reference Data 25(6): 1509-1596.
Stephan, K., and K. Lucas. 1979. Viscosity of Dense Fluids: 75-79.
Styrikowitsch, M.A., S.L. Miropolsky, and M.E. Shitsman. 1959. Wärmeübergang im kritischen Druck-
gebiet bei Strömung Arbeitsmediums. Mitteilungen der VGB 61: 288-294.
Tanaka, H., N. Nishiwaki, M. Hirate, and A. Tsage. 1971. Forced Convection Heat Transfer to a Fluid near
Critical Point Flowing in Circular Tube. International Journal of Heat and Mass Transfer 14: 739-750.
TRC Thermodynamic Tables—Non-hydrocarbons. The Texas A&M University System: cb-1004.0– 1004.3
and gb-1004.0–1004.3.
Downloaded by [Fondren Library, Rice University ] at 10:03 03 May 2013

Ulybin, S.A., and S.S. Bakulin. 1977. The Thermal Conductivity of Carbon Dioxide at Temperatures of
220-1300 K and at Pressures up to 300 MPa. Thermal Engineering 24 (1): 68-72.
Ulybin, S.A., and V.I. Makarushkin. 1976. The Viscosity of Carbon Dioxide at Temperatures of
220-1300 K and at Pressures up to 300 MPa. Thermal Engineering 23(6): 65-70.
van Driest, E.R. 1956. On Turbulent Flow near a Wall. Journal of the Aeronautical Sciences 23: 1007.
Vargaftik, N.B. 1975. Tables on the Thermophysical Properties of Liquids and Gases, 2nd Ed. Washington
DC, Hemisphere Publishing Corp.: 208-210.
VDI. 1988. VDI Wärmatlas, 5th Edition.
VDI. 1994. VDI Wärmatlas, 7th Edition.
Vesovic, V., W.A. Wakeham, G.A. Olchowy, J.V. Sengers, J.T.R. Watson, and J. Millat. 1990. The Trans-
port Properties of Carbon Dioxide. Journal Physical Chemical Reference Data 19(3): 763 –808.
Vukalovich, M.P., V.V. Altunin, and V.V. Blinov. 1963. Thermophysical Properties of Carbon Dioxide.
Teplofizika Vysokikh Temperatur 1(3): 356 – 367.
Wood, R.D., and J.M. Smith. 1964. Heat Transfer in the Critical Region—Temperature and Velocity Pro-
files in Turbulent flow. AIChE Journal 10(2): 180 – 186.
Yamagata, K., K. Nishikawa, S. Hasegawa, T. Fuji, and S. Yoshida. 1971. Forced Convective Heat Trans-
fer to Supercritical Water Flowing in Tubes. International Journal of Heat and Mass Transfer 15:
2575-2593.
Zhou, N., and A. Krishnan. 1995. Laminar and Turbulent Heat Transfer in Flow of Supercritical CO2.
HTD-Vol. 307 National Heat Transfer Conference Vol. 5 ASME: 53-63.