i An update to this article is included at the end

International Journal of Heat and Mass Transfer 63 (2013) 134–140

Contents lists available at SciVerse ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

On heat transfer in tubes

V. Gnielinski ⇑
Institut für Thermische Verfahrenstechnik, Karlsruher Institut für Technologie (KIT), Campus Süd, Kaiserstr. 12, 76131 Karlsruhe, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Heat transfer coefficients for thermal design of heat exchangers are taken from equations in laminar or
Received 18 October 2012 turbulent flow. The equations used to calculate the Nusselt numbers in tubes do not agree in the transi-
Received in revised form 14 March 2013 tion region between laminar and turbulent flow. There is a gap at Re = 2300.
Accepted 8 April 2013
A linear interpolation between the Nusselt numbers at Re = 2300 and Re = 4000, each resulting from
Available online 25 April 2013
well-known equations is presented. The linear interpolation is based on the results of a paper published
recently.
Keywords:
New experimental findings about the Nusselt numbers in tubes at turbulent flow again confirm the
Heat transfer
Tubes
respective equation.
Transition regime Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction That equation takes into account the well-known results ob-
tained by scientists many years earlier.
Equations are needed to calculate heat transfer coefficients for His Pr0.42 referred to the paper by Friend and Metzner [2] who
fluids in a variety of heat exchanger designs. The heat transfer coef- had demonstrated experimentally in fluids with Prandtl numbers
ficients depend on the Reynolds and Prandtl numbers and the between 0.46 and 346 that the 0.42 exponent is a better descrip-
physical properties of the fluids. tion of the dependence of heat transfer for turbulent flow on the
Analytical and numerical solutions for heat transfer for laminar Prandtl number than the exponent of 1/3 appearing in a number
flow have been available for many years whereas heat transfer of earlier correlations on the same subject such as that of the equa-
coefficients for turbulent flow must be calculated from empirical tion named after Dittus–Boelter.
equations based on experimental data. With [1+(d/L)2/3], he took into account the dependence of the
On the basis equations or numerical data for the flow field, ad- heat transfer coefficient on the length of the tube, which is impor-
vanced equations for heat transfer have been developed by analogy tant especially in relatively short tubes. In this way, he includes the
with the flow field and the temperature field. developing hydraulic and thermal boundary layers from the begin-
No numerical descriptions of the flow field in the transition re- ning of the tube.
gion between laminar and turbulent flows in tubes were found in His coefficient of ðlb =lw Þ0:14 referred to the findings by Sieder
the past. In practice however many heat exchangers operate in that and Tate [3] who had matched the experimental data by different
region. A recent paper offered solutions to this problem. authors from the literature for heating and cooling of liquids. He
On the results of the findings in that, equations are developed showed that the inclusion of this factor in the equation makes
which are to provide heat transfer coefficients for all tube flow any distinction between heating and cooling by means of different
regimes. exponents of the Prandtl number unnecessary. Well known
authors used to do that before.
2. A review Colburn [4] was one of the first authors to find the heat transfer
data for turbulent flow below Re = 104 to drop to Re = 2300 from
In 1959, Hausen [1] published his revised heat transfer correla- the straight line drawn through the data in log–log plot. As no
tion for turbulent flow in tubes: equation was available to describe the data in this region, he rec-
h i ommended a ‘‘résumé chart’’ to yield heat transfer coefficients in
Nu ¼ 0:037ðRe0:75  180ÞPr0:42 1 þ ðd=LÞ2=3 ðlb lw Þ0:14 ð1Þ this region for practical use.
The same drop can be seen in the diagrams by Sieder and Tate
[3] demonstrating this influence of fluid properties on heating and
cooling, and from the data by Friend and Metzner [2] showing the
⇑ Tel.: +49 7243 98518.
dependence of the Nusselt number on the Prandtl number. Many
E-mail address: volker.gnielinski@kit.edu
other authors found that drop as well.

0017-9310/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijheatmasstransfer.2013.04.015
 V. Gnielinski / International Journal of Heat and Mass Transfer 63 (2013) 134–140 135

Nomenclature

d tube diameter c Intermittency factor

K correction factor
L tube length Subscripts
Nu Nusselt number b bulk
Pr Prandtl number w wall
Re Reynolds number based on tube diameter lam laminar
T Kelvin temperature turb turbulent
FD fully developed
Greek symbols SE square edged
l dynamic viscosity
n friction factor

In the light of these findings, Hausen [1] included the factor of from theoretical derivations. The theoretical solutions for the two
(Re0.75  180) in his equation. boundary conditions are given in [13].
When reviewing the validity of the so-called Hausen equation, Under the assumption that there is no influence of the free-con-
Eq. (1), in 1975 the author [5] became aware of the publication vection case (I), UWT yields
by Hufschmidt et al. [6] who had shown in their experimental data n o1=3
3
with water and Reynolds numbers between 2⁄104 and 5⁄105 that Num;T ¼ Nu3m;T;1 þ 0:73 þ ½Num;T;2  0:7 þ Nu3m;T;3 with
the Reynolds number exponent is 0.87, i.e. higher than the 0.75
ð7Þ
in Eq. (1). All these findings, the recommendations in [6], and the
fact that, in turbulent flow, the exponent of Re is dependent on
Num;T;1 ¼ 3:66 and ð8Þ
the Prandtl number made Gnielinski [5] set up the equation below
Eq. (2), which is a modification of an equation by Petukhov [7]: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3
"  2=3 # Num;T;2 ¼ 1:615 Re Pr d=L and ð9Þ
ðn=8ÞðRe  1000ÞPr d
Nu ¼ pffiffiffiffiffiffiffiffiffiffi 1 þ K ð2Þ
1 þ 12:7 ðn=8ÞðPr2=3  1Þ L  1=6
2
Num;T;3 ¼ ðRe Prd=LÞ1=2 ð10Þ
The K factor is 1 þ 22 Pr

K ¼ ðPr=Prw Þ0:11 ð3Þ For case (II) (constant heat flux), UHF, the result is
 o1=3
3
for liquids instead of (lb/lw)0.14 in Eq. (1). Eq. (3) follows Hufsch- Num;q ¼ Num;q;1 þ 0:63 þ ðNum;q;2  0:6Þ3 þ Nu3m;q;3 with
midt and Burck [8] as well as Yakovlev [9]. For gases,
ð11Þ
K ¼ ðT b =T w Þn ð4Þ
Num;q;1 ¼ 4:354 and ð12Þ
The exponent, n, depends on the gas, e.g., for air it is n = 0.45.
In Eq. (2), n is the friction factor for turbulent flow in tubes. It qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3
was defined first as Num;q;2 ¼ 1:953 Re Prd=L and ð13Þ

n ¼ ð1:84log10 Re  1:64Þ ð5Þ p

3
ffiffiffiffiffiqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Num;q;3 ¼ 0:924 Pr Reðd=LÞ ð14Þ
according to Filonenko [10]. Later, Eq. (5) was replaced by Eq. (6):
Fig. 1 shows the Nusselt numbers from Eqs. (2) and (7) as a function
n ¼ ð1:8log10 Re  1:5Þ ð6Þ of the Reynolds number for (d/L) = 0.001 and the Prandtl numbers
by Konakov [11]. Eq. (6) is a better representation than Eq. (5) of the 0.7 (air), 7 (water), and 70 (oil).
equation for the friction factor in tubes by Blasius and that by Pra-
ndtl [12] for Re > 105, which can be only solved iteratively. 3. Problem
As [5], was translated into English, Eq. (2) with Eqs. (3)–(5) is
often cited in the international literature. Fig. 1 shows a gap between the Nusselt numbers at Re = 2300
In [5] Eq. (2) was checked against a large number of experimen- for laminar and for turbulent flows. The respective equations do
tal data. not overlap. This fact is mentioned in a number of publications
Turbulent flow starts at Re = 2300 in tubes fixed edgewise in the and books e.g. [14].
tube sheet. Zhipeng Duan [15] addresses the problem by saying: ‘‘Since the
This is the way tubes are fixed in the tube sheets in most heat transition mechanism and a definitive picture of the transition pro-
exchangers. The range of validity of Eq. (2) therefore begins at cess are still unclear, the laminar to turbulent transition region
Re = 2300. (There are some exceptions to this rule for high values should be considered a metastable and complicated region. The
of (d/L), see [5]). transition region is a varying mixture of different transport mech-
The flow is laminar at lower Reynolds numbers. It is not neces- anisms and a mixed degree relies on the value of the Reynolds
sary to distinguish between the boundary conditions of ‘‘(I) con- number and other conditions’’.
stant wall temperature (UWT)’’ and ‘‘(II) constant heat flux Designing a heat exchanger with a computer program will re-
(UHF)’’ along the tube for heat transfer in a turbulent flow. For lam- veal jumps at Re = 2300 for the Nusselt number in case of iterations
inar flow, the two boundary conditions produce different equations in the area around Re = 2300.
136 V. Gnielinski / International Journal of Heat and Mass Transfer 63 (2013) 134–140

1000

Nu
d / L = 0.001
Pr=0.7 Eq. (7)
100
Pr=7
Pr=70
Pr=0,7 Eq. (2)
Pr=7
10
Pr=70

1
1,000 10,000 100,000 1,000,000
Re

Fig. 1. Nusselt numbers calculated with Eqs. (2) and (7) showing a discontinuity at Re = 2300.

To overcome this problem, Taborek [16] proposed a linear inter- ‘‘To add credence to the results . . ., numerical simulations have
polation between the Nusselt numbers from the different equa- been performed’’ [20]. The numerical results are presented as data
tions for laminar and turbulent flows between Re = 2000 and points in Fig. 2 together with the Nusselt numbers resulting from
Re = 8000. However he did not specify the equations to be used. the application of the above mentioned friction factor in Eq. (2)
After a discussion with Gnielinski [17] in 1995 also proposed a in the range 2300 < Re < 4500. The Nusselt numbers for higher val-
linear interpolation, though between Re = 2300 and Re = 104. ues of Re are calculated from Eq. (2) with the friction factor from
For a laminar flow at Re = 2300 and constant wall temperature, Eq. (6). The data points were calculated for the two boundary con-
the Nusselt number resulting from Eq. (7) must be substituted and ditions, ‘‘Uniform heat flux UHF’’ and ‘‘Uniform wall temperature
for a turbulent flow at Re = 104, the Nusselt number resulting from UWT’’, along the surface of the tube at Pr = 0.7.
the extended equation by Petukhov [7] (Eq. (15)) must be
substituted
" 5. The purpose of this paper
 2=3 #
ðn=8ÞRePr d
Nu ¼ pffiffiffiffiffiffiffiffiffiffi 1þ K ð15Þ As a consequence of the results of [20], the linear interpolation
1 þ 12:7 ðn=8ÞðPr2=3  1Þ L
between the Nusselt numbers resulting from Eq. (7) for laminar
The proposal in [16] was checked against the same large number of flow and boundary condition UWT and from Eq. (11) for boundary
experimental data as before to test Eq. (2) in [6]. There was no condition UHF, respectively, at Re = 2300 and the Nusselt number
change in the scatter data in the range of 2300 < Re < 104. according to the equation by Petukhov [7] at Re = 104 as proposed
Both proposals were empirical only and had no theoretical by the author in [17] must be changed.
function. It should be checked whether linear interpolation between the
This method of calculation was introduced by the author also in results of the laminar equations at Re = 2300 and the results of Eq.
[18,19]. (2) with Eq. (6) at Re = 4000 (4000 presenting a better approach to
the algebraic relation than 4500) would usefully represent the re-
4. New developments sults by Abraham et al. [20] in this region.
The results of the linear interpolation calculated from Eq. (17)
Summarizing the results of some earlier work, Abraham, Spar- together with the results by Abraham et al. are shown in Fig. 2
row and Minkowycz in 2011 published a paper dealing with the
‘‘Internal-flow Nusselt numbers for the low-Reynolds-number Nu ¼ ð1  cÞNulam;2300 þ c Nuturb;4000 ð16Þ
end of the laminar-to-turbulent transition regime’’ [20]. Based on
with
‘‘a model for the fluid flow in all possible flow regimes and imple-
mented for flow in round pipes of constant diameter, parallel-plate Re  2300
channels, and round pipes in which there is an upstream conical c¼ and 06c61 ð17Þ
4000  2300
enlargement’’ they obtained ‘‘definitive information on the rela-
tionship between the fully developed friction factor and the Rey- Nulam,2300 is calculated from Eq. (7) at Re = 2300 for the ‘‘uniform
nolds number for the entire range of Reynolds numbers wall temperature’’, UWT, boundary condition and for ‘‘constant heat
encompassing laminar, intermittent, and turbulent flows. The flux’’, UHF, from Eq. (11).
resulting relationships for the low-Reynolds-number end of the Nuturb,4000 at Re = 4000 is calculated from Eq. (2) with Eq. (6)
transition regime were fitted by algebraic equations’’ [20] and (each for Pr = 0.7, d/L = 0.0001, and K = 1).
were presented in a table as a friction factor. The range of validity For lower values of Re, Nu is calculated from Eq. (7) or Eq. (11),
of this friction factor for a straight round pipe is 2300 < Re < 4500. depending on the boundary condition. For Re > 4000, Nu is calcu-
Application of this friction factor in Eq. (2) enables the calculation lated from Eq. (2) with Eq. (6) for the friction factor.
of the Nusselt number’’ in the specified range. ‘‘For higher Reynolds Reproduction of the data by Abraham et al. [20] from the
numbers, Eq. (2) (in this paper) is used without modification’’[20]. numerical calculation by Eq. (17) for the UWT boundary condition
 V. Gnielinski / International Journal of Heat and Mass Transfer 63 (2013) 134–140 137

20

Nu

Pr = 0.7

Abraham et al. algebr.

10
Gnielinski Eq.(2)
Abraham et al. num. UWT
Abraham et al. num. UHF
Gnielinski Eq. (17),T
Gnielinski Eq.(17),q

0
1000 2000 3000 4000 5000 6000
Re

Fig. 2. The results of Abraham et al. [20] and the proposed method of calculation.

is reasonable good. Differences of up to 30% resulted for the UHF the Nusselt number from Eq. (2) with (6), are recommended also
boundary condition. No explanation has been found so far. for higher values of the Prandtl number and higher values of (d/L).
However using Eq. (17) is more reliable for practical application Fig. 3 represents the findings of this paper. Eq. (2) has been pro-
in heat exchanger design. ven to work and is recommended for turbulent flow.
The line in Fig. 2 marked ‘‘Abraham et al. algebra’’ results from Eqs. (7) and (11) were developed theoretically and numerically
Eq. (1) in [20] being identical with Eq. (2) for (d/L) = 0 and K = 1 in and can be recommended for laminar flow under various boundary
this paper, and a friction factor given by Abraham et al. in [20]. A conditions.
correlation for this friction factor was developed in [20] only for Connecting Eq. (7) (or Eq. (11)) at Re = 2300 with Eq. (2) at
the region 2300 < Re < 4500. Hence there are no differences be- Re = 4000 by Eq. (17) avoids the gap between the equations for
tween the two boundary conditions. laminar and turbulent flows and leads to a method of calculation
supplying heat transfer coefficients for all flow regions in tubes.

6. Results
7. Comparison with experimental data
Even though the results by Abraham et al. [20] were obtained
only for Pr = 0.7 and a fully developed flow, the new linear interpo- No experimental data were found which have allowed checking
lation procedure according to Eqs. (16) and (17), and calculation of the Nusselt numbers obtained by the calculation procedure out-

1000

d / L = 0.01
100

Nu

10 Eqs.(2,6,17,18) Pr=0.7

Eqs.(2,6,17,18) Pr=7

Eqs.(2,6,17,18) Pr=70

1
1000 10000 100000
Re

Fig. 3. Proposed method of calculation for different Prandtl numbers and d/L = 0.01, UWT boundary condition.
138 V. Gnielinski / International Journal of Heat and Mass Transfer 63 (2013) 134–140

100

Pr = 0.7
Nu

S. Lau [22]

A. Black III [23]

R. Kemink [24]

D. Wesley [25]

Nu, Eqs.(2),(6), d/L=0.001

10
1000 10000 100000
Re

Fig. 4. Experimental findings by several authors and the method of calculation proposed.

lined above in the region of linear interpolation, i.e., between In 2011, Li Xiao-Wei et al. [26] published data on turbulent con-
1000 < Re < 4000. vective heat transfer for water in tubes with two-dimensional
Eq. (2) with Eq. (5) for Re > 4000 was checked against a large roughness’s. First, Nusselt numbers in smooth tubes were measured
number of experimental data in [5]. This process will not be re- in a double-pipe heat exchanger. The author thanks Li Xiao Wei for
peated here. making available the data of these experiments. As the data were
However there are some new data in this case. measured at different water temperatures, the Prandtl numbers
First, Abraham et al. [21] in 2009 published a number of data varied between Pr = 5.5 and Pr = 3.2. They were normalized there-
from four Ph.D. Theses [22–25] written at the University of Minne- fore by the factor of (Pr/Prw)0.11(4.5/Pr)0.42 (the mean dependence
sota. These data were used to check the applicability of Eq. (2). of the Nusselt numbers in the region around Pr = 5 to 10 is Pr0.42).
Abraham was kind enough to send this author the tabulated data Fig. 5 shows the normalized experimental Nusselt numbers.
of these theses. The respective experiments were performed with Agreement with the plot of Nusselt numbers calculated for
air, long tubes, and the UHF boundary condition. Fig. 4 shows the Pr = 4.5 is very good. The differences for Re < 104 may have been
data in comparison to Eq. (2). Agreement between the data and caused by the design of the inlet of the tube or by free convection
the equation is excellent. effects.

1.000

Nucorr = Nuexp(4 / Pr)0.42 (Pr / PrW)0.11

Nu
Nu corr
d / L = 0.0086

100
Li Xiao-Wei [26], Pr= 5.6 - 5.8

Li Xiao-Wei [26], Pr= 4.0 - 4.2

Li Xiao-Wei [26], Pr= 3.1- 3.3

Nu, Eqs. (2,6,17,18), Pr=4

10
1.000 10.000 100.000 1.000.000
Re

Fig. 5. Experimental data by Li Xiao-Wei et al. [26] and the method of calculation proposed.
 V. Gnielinski / International Journal of Heat and Mass Transfer 63 (2013) 134–140 139

100

Pr = 0.7
Nu

Smith Eiamsa-Ard [27] d/L=0.032

Buyukalaca [28], d/L= 0.019

Nu, Eqs.(2,6,17,18), d/L=0.032

Nu Eqs. (2,6,17,18), d/L= 0.019

10
1000 10000 100000
Re

Fig. 6. Experimental data by Smith Eimsa-Ard et al. [27] and by Buyucalaca et al. [28] and the method of calculation proposed.

Smith Eiamsa-Ard et al. [27] reported on heat transfer in an Unlike the plot of Nusselt numbers as a function of Reynolds
electrically heated tube (UHF) with wire coil turbulators and air. numbers shown above the dependence on Reynolds numbers is
This author is indebted to Prof. Smith Eiamsa-Ard for supplying different for the Nusselt numbers measured by Olivier and Meyer
data of the plain tubes tested first for control. They are shown in [29] for the counterflow of water in the inner pipe of a double pipe
Fig. 6. Even though the Nusselt numbers are systematically lower heat exchanger. Heat transfer was investigated for various inlet
than the respective calculation from Eq. (2) the dependence of configurations. The tube under study was installed downstream
the Nusselt numbers on the Reynolds numbers is evident. of a calmimng section. Fig. 7 shows the results of the investigation
Buyucalaca et al. [28] investigated the effect of ring turbulators for a fully developed flow (FD) and for a square entrance (SE).
on heat transfer in electrically heated tubes (UHF) The experimen- Again, Nusselt numbers experimentally determined were normal-
tal data they gathered from tests in their device with a smooth ized by the factor of (4.5/Pr)0.42 and corrected by (Pr/Prw)0.11 for
tube first seemed to be influenced by natural convection at the temperature dependence of the physical properties. The gradi-
decreasing Reynolds numbers. As they were unable to find an ent of the straight line through the Nusselt numbers is different
explanation of the increasing differences (max. 21% for from the respective line of all the other data shown in the preced-
Re = 3115) at decreasing Reynolds numbers and the data from ing diagrams. The author has no explanation of this fact. The data
Eq. (2) (for (d/L) = 0 and K = 1), they developed a correlation of their were reproduced here only to present as complete as possible the
own for 3115 6 Re 6 25:000. data on heat transfer in tubes published in the past few years.

100

90

80
0.42 0.11
70 Nucorr = Nu exp(4.5/Pr) (PrW /Pr)
60
Nu, Nu corr

50
FD 15.88, d/L= 0.0029
40
SE 15.88, d/L= 0.0029
30
FD 19.02, d/L=0.0037
20 SE 19.02, d/L=0.0037

10 Eqs., Nu, Pr=4.5, d/L=0.0029

Eqs., Nu, Pr=4.5, d/L=0.0037

0
1.000 3.000 5.000 7.000 9.000 11.000 13.000 15.000
Re

Fig. 7. Experimental data by Olivier and Meyer [29] and the method of calculation proposed.
140 V. Gnielinski / International Journal of Heat and Mass Transfer 63 (2013) 134–140

The author is grateful to Olivier and Meyer for providing the [12] L. Prandtl, Neuere Ergebnisse der Turbulenzforschung, Zeitschr. VDI 77 (1933)
105–114.
tabulated data.
[13] V. Gnielinski, Heat transfer in laminar flow, in: VDI Heat Atlas, second ed.,
Springer Verlag, 2010 (Chapter G1, Section 3).
8. Summary [14] R.K. Shah, P. Sekulic, Fundamentals of Heat Exchanger Design, John Wiley &
Sons, Hoboken, NJ, 2003.
[15] Zhipeng Duan, New correlative models for fully developed turbulent heat and
A new linear interpolation between the equations for laminar mass transfer in circular and noncircular ducts, Trans. ASME J. Heat Transfer
and turbulent flow in the transition region was developed on the 134 (1) (2012). 014503-1–014503-6.
[16] J. Taborek, Design method for tube-side laminar and transition flow regime
basis of a recent paper. with effects of natural convection, in: Open Forum Session Forum of the Ninth
In addition to earlier data, new experimental findings by vari- Int. Heat Transfer Conference, Jerusalem, Israel, 1990.
ous authors confirm the equation for heat transfer in turbulent [17] V. Gnielinski, Ein neues Berechnungsverfahren für die Wärmeübertragung im
Übergangsbereich zwischen laminarer und turbulenter Rohrströmung, Forsch.
flow used for linear interpolation.
Ing. Wes. 61 (9) (1995) 240–248.
[18] V. Gnielinski, Heat transfer in turbulent flow in pipes, VDI Heat Atlas, second
References ed., Springer Verlag, 2010 (Chapter G1, Section 4).
[19] V. Gnielinski, Heat Exchanger Design Handbook (HEDH), Single phase
[1] H. Hausen, Neue Gleichungen für die Wärmeübertragung bei freier und convective heat transfer: forced convection in ducts, Heat Exchanger Design
erzwungener Strömung, Allg. Wärmetechnik 9 (4/5) (1959) 75–79. Updates, Begell House, 1999 (Chapter 5).
[2] W.L. Friend, A.B. Metzner, Turbulent heat transfer inside tubes, J. Am. Inst. [20] J.P. Abraham, E.M. Sparrow, W.J. Minkowycz, Internal-flow Nusselt numbers
Chem. Eng. 4 (1958) 393–402. for the low-Reynolds-number end of the laminar-to-turbulent transition
[3] E.N. Sieder, G.E. Tate, Heat transfer and pressure drop of liquids in tubes, Ind. regime, Int. J. Heat Mass Transfer 54 (2011) 584–588.
Eng. Chem. 28 (1936) 1429–1435. [21] J.P. Abraham, E.M. Sparrow, J.C.K. Tong, Heat transfer in all pipe flow regimes:
[4] A.P. Colburn, A method of correlating forced convection heat-transfer data and laminar, transition/intermittent and turbulent, Int. J. Heat Mass Transfer 52
comparison with fluid friction, Int. J. Heat Mass Transfer 7 (1964) 1359–1384 (2009) 557–563.
(Reprinted from Trans. Amer. Inst. Chem. Eng. Vol. 29, 1933. [22] S. Lau, Effect of plenum length and diameter o turbulent heat transfer in a
[5] V. Gnielinski, Neue Gleichungen für den Wärme- und Stoffaustausch in downstream tube and on plenum-related pressure loss, Ph.D. Thesis,
turbulent durchströmten Rohren und Kanälen, Forsch. Ing. Wes. 41 (1) (1975) University of Minnesota, 1981.
8–16. Englisch Translation: New equations for heat and mass transfer in [23] A. Black III, The effect of circumferentially-varying boundary conditions on
turbulent pipe and channel flow, Int. Chem. Eng.16 (2) (1976) 359–368. turbulent heat transfer in a tube, Ph.D. Thesis, University of Minnesota, 1966.
[6] W. Hufschmidt, E. Burck, W. Riebold, Die Bestimmung örtlicher und mittlerer [24] R. Kemink, Heat transfer in a downstream tube of a fluid withdrawal branch,
Wärmeübergangszahlen in Rohren bei hohen Wärmestromdichten, Int. J. Heat Ph.D. Thesis, University of Minnesota, 1977.
Mass Transfer 9 (6) (1966) 539–565. [25] D. Wesley, Heat transfer in pipe downstream of a Tee, Ph.D. Thesis, University
[7] B.S. Petukhov, Heat transfer and friction in turbulent pipe flow with variable of Minnesota, 1976.
physical properties, Adv. Heat Transfer 6 (1970) 503–564. New York, Academic [26] Li Xiao-Wei, Meng Ji-an, Li Zhi-xin, Roughness enhanced mechanism for
Press. turbulent convective heat transfer, Int. J. Heat Mass Transfer 524 (2011) 1775–
[8] W. Hufschmidt, E. Burck, Der Einfluss temperaturabhängiger Stoffwerte auf 1781.
den Wärmeübergang bei turbulenter Strömung von Flüssigkeiten in Rohren [27] Smith Eiamsa-Ard, V. Kongkaitpaiboon, P. Promvongue, Thermal performance
bei hohen Wärmestromdichten und Prandtlzahlen, Int. J. Heat Mass Transfer assessment of turbulent tube flow through wire coil turbulators, Heat Transfer
11 (6) (1968). 1041-104. Eng. 32 (11–12) (2011) 957–967.
[9] V.V. Yakovlev, Örtliche und mittlere Wärmeübetragung bei turbulenter [28] O. Buyukalaca, V. Ozceyyhan, S. Günes, Experimental investigation of thermal
Rohrströmung nichtsiedenden Wassers und hohen Wärmebelastungen (Local performance in a tube with detached circular ring turbulators, Heat Transfer
and average heat transfer in the turbulent flow of nonboiling water and high Eng. 33 (8) (2012) 682–692.
heat loadings), Kernenergie 3 (10/11) (1960) 1098–1099. [29] J.A. Olivier, J.P. Meyer, Single-phase heat transfer and pressure drop of the
[10] G.K. Filonenko, Hydraulischer Widerstand von Rohrleitungen (Orig.Russ. ), cooling of water inside smooth tubes for transitional flow with different inlet
Teploenergetika 1 (4) (1954) 40–44. Hydraulic resistance in pipes. geometries, HVAC & Res. 16 (4) (2010) 471–496 (RP-1280).
[11] P.K. Konakov, Eine neue Formel für den Reibungskoeffizienten glatter Rohre
(Orig. Russ), Berichte der Akademie der Wissenschaften der UdSSR L1 (7)
(1946) 503–506. A new equation for the frictional resistance in smooth tubes..
 Update
International Journal of Heat and Mass Transfer
Volume 81, Issue , February 2015, Page 638

DOI: https://doi.org/10.1016/j.ijheatmasstransfer.2014.10.063
 International Journal of Heat and Mass Transfer 81 (2015) 638

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

Corrigendum to ‘‘On heat transfer in tubes’’ [International Journal

of Heat and Mass Transfer 63 (2013) 134–140]
V. Gnielinski ⇑
Institut für Thermische Verfahrenstechnik, Karlsruher Institut für Technologie (KIT), Campus Süd, Kaiserstr. 12, 76131 Karlsruhe, Germany

The author regrets that (a) the exponent 2 in Eqs. (5) and (6) behind the last bracket was forgotten. The correct equations are:
2
n ¼ ð1:84log10 Re  1:64Þ ð5Þ

2
n ¼ ð1:8log10 Re  1:5Þ ð6Þ
(b) In the legends of Figs. 3, 5, and 6 Eqs. (16) and (17) must be cited instead of Eqs. (17) and (18).
The authors would like to apologise for any inconvenience caused.

DOI of original article: http://dx.doi.org/10.1016/j.ijheatmasstransfer.2013.04.015

⇑ Tel.: +49 7243 98518.
E-mail address: volker.gnielinski@kit.edu

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.10.063
0017-9310/Ó 2014 Elsevier Ltd. All rights reserved.