Sie sind auf Seite 1von 10

International Journal of Heat and Mass Transfer 95 (2016) 525534

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer


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

New friction factor equations developed for turbulent flows in rough


helical tubes
Houjian Zhao, Xiaowei Li , Xinxin Wu
Key Laboratory of Advanced Reactor Engineering and Safety of Ministry of Education, Collaborative Innovation Center of Advanced Nuclear Energy Technology, Institute of Nuclear
and New Energy Technology, Tsinghua University, Beijing 100084, China

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

Article history: Helical tubes are widely used and there lacks friction factor equations for turbulent flows in rough helical
Received 17 July 2015 tubes. Turbulent flows in rough helical tubes were investigated theoretically in this paper. Friction factor
Received in revised form 7 December 2015 equations for transitionally and fully rough regime turbulent flows in rough helical tubes were derived
Accepted 16 December 2015
based on the logarithmic velocity distribution law. The parameters of the equations were obtained by
regression analysis of experimental data of Clancy (1949) and verified by experimental data of
McElligott (1948). The characteristics of the friction factors in rough helical tubes were discussed based
Keywords:
on the equations. Friction factors of helical tubes are influenced by Reynolds numbers, relative rough-
Helical tube
Curvature ratio
nesses and curvature ratios. The friction factors of rough helical tubes can also be divided into a transi-
Friction factor tionally rough regime and a fully rough regime according to the roughness height and the Reynolds
Relative roughness number like that for straight tubes. Roughnesses have greater effects on the transition from the transi-
Turbulent flow tionally rough regime to the fully rough regime than that of curvature ratios. Roughness effects on friction
factors increase with the increasing of the curvature ratio and the Reynolds number.
2015 Elsevier Ltd. All rights reserved.

1. Introduction characteristics in small bending radius helical tubes. Friction factor


correlations were obtained by regression analysis of the experi-
Helical tubes are widely used in chemical, petroleum and mental data and the results showed that the critical Reynolds num-
nuclear industries due to the advantages of compact structures, ber for laminar flow to turbulent flow transition in helical tubes
high heat transfer coefficients and good thermal expansion perfor- was much greater than that in straight tubes. Yamamoto et al.
mances. In nuclear industries, helical tubes are usually used in [11] investigated flows in helical tubes with large pitches numeri-
steam generators [35], especially for gas cooled reactors. Flows cally and experimentally. Simulations agreed well with experi-
in helical tubes are more complex than those in straight tubes. Sec- mental data and showed that two vortices transformed into one
ondary flows occur in helical tubes in the cross section due to the single vortex as the pitches increase. Hart et al. [12] presented a
centrifugal force, which causes more friction losses. friction factor chart for flows in curved tubes which covered the
Since the classical works by Dean [6,7], flows in smooth curved laminar flow region and turbulent flow region and derived an
tubes have been extensively studied. Dean obtained an approxi- equation to calculate friction factors. Grundman [13] developed a
mate solution through perturbation over the Poiseuille flow in practical friction factor diagram for helical tubes which accounted
straight tubes and introduced the Dean number. Gnielinski [8] pre- for curvature ratio effects. The diagram also offered a graphic view
sented equations for laminar flows and turbulent flows to calculate of the flow conditions and other parameters. Mishra and Gupta
the friction factors and heat transfer coefficients in helical tubes. [14] investigated the pitch effects on pressure drops in helical
Results were compared with experimental data from literature tubes. Pressure drops were measured and equations for friction
and the deviations were less than 15%. Srinivasan [9] measured factors were derived using the modified Dean number. They con-
the pressure drops of water and oil in helical tubes and developed cluded that the pitch effects could be eliminated by using the mod-
equations to predict friction factors for laminar, transition and ified curvature diameter. Ito [15] measured the frictional pressure
turbulent flow regions. Ju et al. [10] evaluated the hydraulic drop of turbulent flows in smooth curved tubes with zero pitch and
derived an equation using the 1/7th power velocity distribution
law and an equation using the logarithmic velocity distribution
Corresponding author. Tel.: +86 10 62784825; fax: +86 10 62797136. law, which have been used widely for engineers [16].
E-mail address: lixiaowei@tsinghua.edu.cn (X. Li).

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2015.12.035
0017-9310/ 2015 Elsevier Ltd. All rights reserved.
526 H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534

Nomenclature

Notations
Cur curvature ratio (dimensionless) Greek letters
fc friction factors for curved tubes (dimensionless) a; b; c constants in the friction factor equation
Ks average wall surface roughness (m) d thickness of boundary layer (m)
M mass flow rates (kg s1) e relative roughness, e = Ks/r0 (dimensionless)
P static pressure (Pa) h angels in the cross section
r0 tube inner radius (m) l dynamic viscosity (kg s1 m1)
R helical radius (m) q density (kg m3)
Rc modified helical radius (m) u angles in the axial direction
V velocity magnitude in curved tubes (m s1)
Vr velocity magnitude in the r direction (m s1)
Subscripts
Vh velocity magnitude in the h direction (m s1) 1 at the edge of the boundary layer
Vu velocity magnitude in the u direction (m s1) c curved rough tubes
V sw =q1=2 , friction velocity (m s1) c, r helical rough tubes
Vm mean axial velocity (m s1)
c, s helical smooth tubes
y distance normal to the wall (m) rough fully rough regime
y0 distance normal to the wall where the velocity magni- str, r straight rough tubes
tude is zero (m).
trans transitionally rough regime
sw wall shear stress (N m2) w near the wall
sh wall shear stress in the h direction (N m2)
su wall shear stress in the u direction (N m2)

Flows in helical tubes also have been investigated numerically. well with others experimental data for commercial tubes. How-
Austin and Seader [17] solved the NavierStokes equation in the ever, the results deviated from the performances of commercial
stream-function form using the finite difference method to simu- tubes at moderate Reynolds number. Colebrook and White [26]
late steady, incompressible and fully developed flows in helical investigated turbulent flows in roughened straight tubes experi-
tubes. They presented a correlation for pressure drops in terms mentally and found Nikuradses [25] deviations from the commer-
of the Dean number. Soeberg [18] simulated laminar flow in helical cial tubes were due to the uniform roughness in tubes. Then
tubes based on the symmetry of the secondary flow and investi- Colebrook [27] presented an implicit equation to calculated friction
gated the velocity profile changes which were influenced by the factors in rough straight tubes. The equation was widely used to
Dean number. Liu and Masliyah [19] and Yamamoto et al. [20] sim- calculate friction factors for turbulent flow in rough straight tubes.
ulated the flows in helical tubes with finite pitches and found the Many researchers [2832] obtained explicit equations to calculate
friction factors would increase as the curvature ratios increase. the friction factors in rough straight tubes with specific limits.
However, the factors would decrease as the pitches increase. Moody [33] presented a composite friction factor chart of all flow
Yanase et al. [21] used the ForierChebyshev spectral method to regions for rough straight tubes which was widely used in engi-
analyze the stability of the two-vortex and the four-vortex solu- neering. For rough helical tubes, Clancy [1] measured the pressure
tions for flows in slightly curved tubes and found that the two- drops in rough helical tubes and straight tubes and found friction
vortex solution was stable in response to any small disturbances, factors were influenced by Reynolds numbers, helical diameters
while the four-vortex solution was unstable to asymmetric distur- and roughness heights. McElligott [2] compared friction factors
bances. Lai et al. [22] simulated turbulent flows in curved tubes in rough helical tubes and friction factors in rough straight tubes
and found that there were three vortex pairs in the cross section. with comparable roughness. The results showed that friction fac-
One was the Dean-type vortex pair. Another existed in the tube tors in helical tubes depended more on roughnesses than on curva-
core and was caused by the pressure imbalance. The third was near ture ratios.
the outer wall and was the turbulence driven secondary flow. Man- This paper describes a theoretical investigation on turbulent
lapaz and Churchill [23] simulated laminar flows in coiled tubes flows in rough helical tubes and derived friction factor equations
with a finite pitch and developed a friction factor equation for considering the roughness effects, curvature ratios and Reynolds
the laminar flow region. Ivan and Michele [24] simulated turbulent numbers. The equations can predict the friction factors accurately
flows using k  e, SST kx and RSMx models. Pressure drops compared with literature data. The comprehensive effects of
were in excellent agreement with experimental data when using roughnesses and curvature ratios on the friction factor are also dis-
SST kx and RSMx models, but results were unsatisfactory cussed based on the equations.
when using k  e model with wall functions.
To the authors knowledge, most previous studies did not con-
sider the roughness effects on friction factors in helical tubes. 2. Equation derivation
There are no reported equations for calculating the friction factors
in rough helical tubes. However, most helical tubes used for indus- 2.1. Physical model and mathematical description
tries are commercial tubes with different height of roughness.
Many investigations have been done on flows in rough straight Fig. 1 shows a helical tube schematic with the main parameters
tubes which can be taken as the bases for investigation on rough and a secondary flow illustration. The helical radius is defined as R,
helical tubes. Nikuradse [25] experimentally investigated turbu- the tube inner radius is defined as r 0 and the helical pitch is defined
lent flows in rough straight tubes by roughening the tubes with as b. When the helical pitch is small compared with the helical
uniform sand grains. At high Reynolds numbers, the results agreed diameter, torsion effects can be neglected at moderate Reynolds
H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534 527

Fig. 1. Schematic of a helical tube and secondary flow in the cross section.

Fig. 2. The curved tube and velocity vectors in the orthogonal coordinates.

number [19]. Moreover, the effects of pitch on friction factors can section, h denotes the angle between r and the symmetry of the
be eliminated if the helical radius is modified by Eq. (1) [14]. cross section, and u denotes the angular distance from a fixed cross
"  2 # section. V r , V h and V u are taken as the velocity magnitudes in
b directions of r, h and u. The continuity and momentum equations,
Rc R 1 1
2pR Eqs. (2)(5), are obtained based on the following assumptions:

For this analysis, the helical tube shown in Fig. 1 is simplified to


(1) The helical tubes are simplified to be curved tubes with zero
be a helical tube with zero pitch as shown in Fig. 2(a). The toroidal
pitch.
orthogonal coordinate system shown in Fig. 2(b) is adopted, where
(2) The flows are incompressible and steady.
r denotes the distance from the center of the tube in the cross
528 H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534

Z Z Z
(3) The flows are fully developed, so, Vr, Vh and Vu are indepen- 1 r0
@ r0 V 2u sin h r0
1 @P
dent of u. V 2h dr V h V r rr00 d dr  dr
r0 r0 d @h r 0 d R r0 d qr0 @h
(4) The helical radius is much larger than the tube radius, i.e.  r
@V h 0
R  r0. m 8
@r r0 d
@V r V r 1 @V h
0 2 According to Ito [36],
@r r r @h
@P r0
q V 2u1 sin h
2
@V r V h @V r V 2h V u cos h @h R
Vr  
@r r @h r R
  V r V h 0 r r0
1 @P m @ @V h V h 1 @V r
   3  
q @r r @h @r r r @h @V h
V h1 0 r r0  d
2
@r 1
@V h V h @V h V h V r V u sin h
Vr where V u1 and V h1 are the velocity magnitudes at the edge of the
@r r @h r R
  boundary layer in u direction and h direction, respectively. Then
1 @P @ @V h V h 1 @V r
 m  4 Eq. (8) is reduced to,
qr @h @r @r r r @h
Z Z
! @ d
dr 0 2 r0 d

2 2 q V 2h dy sh r 0 q V sin h  q sin h V 2u dy 9


@V u V h @V u 1 @P @ V u 1 @V u 1 @ V u @h 0 R u1 R 0
Vr  m 2 5
@r r @h qR @ u @r 2 r @r r @ u2 where y is the distance normal to the wall near the wall region.
As the velocities in r direction in the boundary layer are
According to Adler [34], the boundary layer will become thinner
neglected. The velocities in the curved tubes are composed by
when the Dean number increases. The friction factor equation is
velocities in axial direction, V u , and velocities in circumferential
derived based on the boundary layer approximation method. This
method divides the cross section of curved tubes into two regions, direction, V h . Then the velocities are assumed to be:
the central core in the central region and the thin boundary layer V KV
near the wall region. The detailed information of the boundary
Vu 1=2
Vh 1=2
10
1 K 2 1 K 2
layer is shown in Fig. 3.
In the boundary layer, the viscous effect is significant. The thick- where K is the ratio of V h and V u . V is the velocity magnitude in
ness of the boundary layer is denoted as d. The axial velocities, V u , curved tubes.
will gradually fall to zero when approaching the wall because of According to the relationship between velocities and wall
the no slip wall conditions. Then V h and V u are comparable in stresses,
the boundary layer. Order-of-magnitude analysis was done to  
@V u  @V h 
reduce Eqs. (2)(5). Similar to the boundary layer theory for plates su l sh l 11
@r w @r w
[35], we know that in the boundary layer, Vh  O(1), Vu  O(1),
Vr  O(d), @=@r  Od1 , and d  r0. In the boundary layer, the Then wall stresses in u direction and h direction can be written
value of r is approximate to r0. Then Eqs. (2) and (4) are reduced to, as,
@V r 1 @V h sw K sw
0 6 su sh 12
@r r 0 @h 1 K 2 1=2
1 K 2
1=2

2 Y is defined to have the same order-of-magnitude as V m =V  .


@V h V h @V h V u sin h 1 @P @2V h
Vr  m 2 7
@r r 0 @h R qr0 @h @r Vm
Y 13
V
The pressure gradient in h direction in Eq. (7) is set as indepen-
dent of r and equal to the pressure gradient at the edge of the where V m is the mean axial velocity magnitude and V  is the friction
boundary layer. Eq. (7) is integrated by using Eq. (6). velocity magnitude, which is defined by,

Fig. 3. Illustration of boundary layer in the cross section.


H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534 529

r
sw 2.3. Friction factor equation in the fully rough regime
V 14
q
In fully rough regimes, y0 jK s , where j is constant [35]. y0 in
As K is the ratio of V h and V u , we can get,
Eq. (24) is replaced to give,
Vu  Vm V h  kV m 15 V u1
B ln d=K s C 25
Eq. (13) is combined with Eqs. (12) and (14) to get, V
where B and C are constants.Eq. (25) is then combined with Eqs.
qV 2m K qV 2m
su  sh  16 (13) and (15), we can obtain the relationship between Y and d=K s .
Y2 Y2
d  eY K s 26
The pressure gradient in u direction is,
Z p The order-of-magnitude of each term in Eq. (9) is obtained after
@P 2R divided by qV 2m r 0 .
 su dh 17
@u pr0 0 R R
d
q @h@ 0d V 2h dy sh r0 q drR0 V 2u1 sin h q rR0 sin h 0 V 2u dy
Eq. (16) is combined with Eq. (17) to get,  27
qV 2m r0 qV 2m r0 qV 2m r0 qV 2m r0
@P RqV 2m
 18
@u r0 Y 2 K 2 eY K s K eY K s eY K s
r0 2 R R
The friction factor of a helical tube, f c is defined by,
Y
  Each term must have the same order-of-magnitude. Comparing
@P 1 R
 f c qV 2m 19 K 2 eY K s =r0 with eY K s =R gives,
@u 2 2r 0
K  r 0 =R1=2 28
Substitute Eq. (19) into Eq. (18) to get the relationship between Y
and f c . Define e, e K s =r0 as the relative roughness. Comparing K=Y 2
2 Y
with K e K s =r0 , we can get that,
f c  Y2 20
1
Define that, Y 2 eY  29
Ke
A The relationships in Eqs. (28) and (29) can be rewritten as,
fc 21
Y2 r 1=2
0
KD 30
where A is constant. R

E
2.2. Logarithmic law for turbulent flow near walls Y 2 eY 31
Ke
The logarithmic velocity distribution law used in this analysis is where D and E are constants.
Prandtls velocity-distribution law which was derived from the Eq. (31) can be solved to get Y,
mixing length theory [35]. It is widely accepted and used for wall
E
turbulence, usually it is called law of the wall. The law can be writ- Y ln  ln K eY 2 ln E 32
ten as, K eY 2
Eq. (32) can be rewritten by using Eqs. (21) and (30).
V aV  ln y b  
1 Ge
where V is the velocity magnitude, a and b are constants, V  is the 1=2
F ln r 0 =R1=2 33
fc fc
friction velocity and y is the distance from the wall.
Then for the boundary layer in the helical tube, we can get, where F and G are constants. r 0 =R is defined as the curvature ratio.
V 1  V aV  ln d=y 22
2.4. Friction factor in the transitionally rough regime
where V 1 is the velocity magnitude at the edge of the boundary
layer in the cross section. According to Colebrook [27], in the transitionally rough regime,
For rough helical tubes, the velocity magnitude is zero at a
y0 HK s Il=qV  34
point, just like that for the straight rough tubes near the wall.
The distance normal to the wall at this point is set to be y0 [35]. where H and I are constants.
V 1 can be expressed as, Eq. (24) can be rewritten as,

V 1 aV  ln d=y0 23 d
V u1 AV  ln 35
HK s Il=qV 
At the edge of the boundary layer, V h1 0 and V 1 V u1 . Eq. (23)
can be rewritten as, Eq. (35) is combined with Eqs. (13) and (15),

V u1 aV  ln d=y0 24 d  eY HK s Il=qV  36

The relationships between y0 and the mean roughness, K s , in the Each side of Eq. (36) is divided by r 0 ,
fully rough regime and the transitionally rough regime are differ-
d LeY Y
ent. The following analysis will be divided into two sections, fric-  JeY e 37
r0 Re
tion factors in the fully rough regime and friction factors in the
transitionally rough regime. Re 2qV m r0 =l; is the Reynolds number. J and L are constants.
530 H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534

The order-of-magnitude of each term in Eq. (9) is obtained after decreasing when the Reynolds number increased, the flows were
divided by qV 2m r0 , in transitionally regime. The parameters in Eq. (43) can be obtained
R R by the regression analysis of experimental data.
d
q @h@ 0d V 2h dy sh r0 q drR0 V 2u1 sin h q rR0 sin h 0 V 2u dy The differential pressures were measured by water manometer
 38
qV 2m r0 qV 2m r0 qV 2m r0 qV 2m r0 at low flow rates and by mercury manometer at high flow rates.
Flow rates were obtained by weighting a mass of water during a
d 2 K d d specific time interval. The experimental errors were not presented
K in reference [1]. The relative errors for pressure drops are set as 2%
r0 Y2 R R
based on the minimum calibration of 1 mm. The maximum relative
Comparing d=R and dK 2 =r0 , we can get the relationship shown in errors for mass flow rates and length are set as 2% and 0.1%, respec-
Eq. (39), which is the same as Eq. (28). tively, based on the significant digits of the experimental data. The
relative errors are calculated by the following equations.
K  r0 =R1=2 39 s
2 2 Du Xn @ ln u2
Comparing dK =r 0 with K=Y , we can get, Dxi 2 45
u i1 @xi
d 1
 40
r 0 KY 2
DP D 2DP D5 p 2
f q 46
Then substitute Eq. (37) into Eq. (40), we can get, 0:5qV 2m L M2 L 4
LeY Y 1 s
 2  2    2
JeY e  41
Re KY 2 Df 1 2 2 2 5 2 1 2
DDP DM DD DL
f DP M D L
Eq. (41) can be rewritten as,
47
Y LeY Y M
Je e 42 The experimental errors for friction factors are below 5%, which
Re KY 2 is calculated by Eq. (47).
where M is constant.
The friction factor equation is obtained by solving Eq. (42), 3.2. Determination of the equation parameters
using Eqs. (21) and (30),
" # As Eq. (43) is a multiple parameter equation and the friction fac-
1 be r0 1=2 c r0 1=2 tor is in implicit expression. Eq. (43) is transformed to Eq. (48) for
1=2
a ln 3=2 Re 43
fc fc R fc R regression analysis to get the parameters.
," 1=2 #
where a, b and c are constants.
be e1=a=f c
c
Re 3=2 48
2.5. Equation compatibility for different flow regimes fc fc r 0 =R1=2

In Clancys experiments [1], pressure drops are measured in 12


The thickness of the viscous sub-layer decreases as the
helical tubes for different curvature ratio and different roughness.
Reynolds number increases. The roughness will dominates the wall
stresses when the roughnesses are all out of the viscous sub-layer. The Reynolds number varied from 1  103 to 1  105 . The experi-
Turbulent flows in the transitionally rough regime will transform mental data for each tube are defined as one group of data and is
to the fully rough regime. With very high Reynolds numbers, marked by the tube number. As the equation is derived based on
Eq. (43) can be written as Eq. (44) by eliminating the term includ- the logarithmic velocity distribution law, the friction factors in
ing the Reynolds number. the laminar flow region should not be used. The critical Reynolds
  number is calculated by Eq. (49) [15].
1 be r 0 1=2
1=2
a ln 44 r 0:32
fc fc R 0
Recri 2  104 49
R
Eq. (44) is the same as Eq. (31), which shows that the derived
friction factor equations for transitionally and fully rough regimes Several groups of data are used for the regression analysis. The
are compatible.
data with Reynolds number below 1  104 are not used. Then
about 5 points in each group are left. The total points used are
3. Equation parameter determination about 30. The curvature ratios and the relative roughnesses are
set as constant in Eq. (48) for each group. The results of the regres-
Friction factor equations for transitionally and fully rough sion analysis are shown in Table 1 and the confidence level is 95%.
regime turbulent flows have been derived. However, the constants The values of a, b and c shown in Table 1 are averaged, then
in Eqs. (33) and (43) need to be determined using experimental gives,
data.
a = 0.923033
3.1. The experiments of Clancy b = 0.104455
c = 1.142495
The experimental data from Clancy [1] are used to approximate
the derived equations. Fig. 4 shows a schematic diagram of the test Then Eq. (43) can be expressed as,
apparatus of Clancy [1]. The pressure drops in helical tubes were " #
measured with the curvature ratio varied from 0.056 to 0.063, 1 0:104e r0 1=2 1:142 r 0 1=2
0:923 ln 3=2 50
mean roughness varied from about 0.64 lm and Reynolds num- 1=2
fc fc R f c Re R
bers varied from 1  103 to 1  105 . As the friction factors kept
H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534 531

Fig. 4. Schematic diagram of the test apparatus of Clancy [1].

Table 1 Table 2
Regression analysis results. Equation verification.

Tube number 2 3 4 5 9 10 Re Relative roughness Experiment Eq. (50) Relative error (%)
a 0.92589 0.92879 0.91636 0.92058 0.92567 0.92091 5 3 0.0257 0.025916 0.8
1  10 0:315  10
b 0.10158 0.09756 0.09101 0.12272 0.09693 0.11693 1  105 0:2849103 0.0247 0.024705 0.01
c 1.19671 1.18591 0.98235 1.19694 1.11030 1.18276
2  104 0:315  103 0.035 0.034533 1.3
Relative roughness 0.00016 0.00018 0.00013 0.00010 0.00013 0.00011
Curvature ratio 0.06214 0.06101 0.06115 0.06186 0.06101 0.06158 2  104 0:302  103 0.033 0.032989 0.03
Adj. R-square 0.99877 0.99915 0.99411 0.99840 0.99858 0.99750 2  10 4
0:284  103 0.0324 0.032957 1.7

According to the discussion in Section 2.5, when Re is high


enough, the turbulent flows in the transitionally rough regime will
convert to the fully rough regime. Influence of Re will diminishes,
left only the relative roughnesss influence. We set c as zero in Eq.
(50). The equation for the fully rough regime is obtained.
 
1 0:104e r 0 1=2
1=2
0:923 ln 51
fc fc R

The coefficients in Eq. (50) are obtained by regression analysis


of experimental data with random roughnesses. But the uniform
roughness effects and random roughness effects on friction factors
are different [26,27]. Then coefficients of the equations to predict
friction factors in uniform roughened helical tubes [3740] in the
transitionally rough regime and fully rough regime will be
different.

3.3. Equation verification Fig. 5. Equation verification.

The equations are verified by the experimental data from


become equivalent to that for smooth tubes. We will discuss this
McElligott [2]. In Ref. [2], the Reynolds numbers are 2  104 and
in Section 4.1. Eq. (50) shows that the friction factors of rough heli-
1  105. The relative roughnesses are from 0.284  103 to
cal tubes are influenced by the combined effects of the Reynolds
0.315  103 which are different from those of Ref. [1]. Friction fac-
number, the curvature ratio and the relative roughness. We will
tors measured in the experiments, factors calculated by Eq. (50)
discuss these effects in Sections 4.2, 4.3 and 4.4.
and the relative errors are shown in Table 2 and Fig. 5, which
shows a good prediction by using Eq. (50).
4.1. Compatibility with Itos equations

4. Discussion Equation (43) is derived on the bases of Itos theoretically inves-


tigation on turbulent flows in smooth curved tubes. Ito [15] inves-
A good friction factor equation for rough helical tubes should be tigated turbulent flow pressure drops in smooth curved tubes
compatible with that for smooth helical tubes, which means when experimentally and theoretically. Two equations are derived based
the roughness diminishes, the equation for rough tubes should on the 1/7th power velocity distribution law and the logarithmic
532 H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534

velocity distribution law, respectively, which are shown in Eqs.


(52) and (53). The equations for smooth tubes have been widely
used.
 1=2
R 0:316
fc h
i1=5 52
r0 2
Re rR0

"  1=2 #
1 3=2 R
1=2
2:11log10 Ref c  0:23 53
fc r0

We set the relative roughness as zero in Eq. (50) to calculate the


friction factors in smooth helical tubes to see whether the equation
is consistent with Itos equations. Friction factors are also calcu-
lated by Eqs. (52) and (53) with the same curvature ratio and the
same Reynolds number. The results are shown in Fig. 6. The rela-
tive errors are less than 4%. The curvature ratios are 0.1 and 0.01.
Cur means the curvature ratio. Fig. 7. Comparison of friction factors in helical tubes with different roughnesses
and the same curvature ratios.

4.2. Roughness effects on the friction factors


effects on the transition progress. Fig. 8 shows that the curvature
Friction factors are calculated using Eq. (50) for the same curva-
ratio has little effect on the transition. Roughness dominates the
ture ratio but different relative roughnesses. The Reynolds num-
transition process, like that in straight tubes.
bers varies from 1  104 to 6  105. The relative roughness varies
from 4  104 to 8  103 . The curvature ratio is 0.1, where the 4.4. Combined effects of roughness and curvature ratio on friction
helical radius is 50 mm and the tube inner radius is 5 mm. The factors
results are shown in Fig. 7. Fig. 7 shows that the friction factors
decrease as the Reynolds numbers increase in a range of Re with Eq. (50) shows that the roughness effects on friction factors are
relatively low values. The range is different for tubes with different influenced by Reynolds numbers and curvature ratios. The factors,
roughness, like the transitionally rough regime in Moody chart. In f c;s , in smooth helical tubes are calculated by Eq. (50) with the rel-
the transitionally rough regime, the factors depend both on the
ative roughness set as zero. The friction factors, f c;r , in rough helical
roughness and the Reynolds number. When Re is high enough, fric-
tubes are calculated by Eq. (50) with the relative roughness set as
tion factor keeps unchanged as Reynolds numbers increase, like
0.004 and the Reynolds number varies from 1  105 to 6  105. The
the fully rough regime in Moody chart.
ratio, f c;r =f c;s reflects the roughness effects on the friction factors in
4.3. Roughness and curvature ratio effects on the transition helical tubes. f c;r =f c;s is calculated with different curvature ratio
and the results are shown in Fig. 9.
Friction factors in the transitionally rough regime, ftrans, are cal- The results in Fig. 9 shows that f c;r =f c;s increases as the curva-
culated by Eq. (50) and friction factors in fully rough regime, f rough , ture ratio increases or the Reynolds number increases. That means
are calculated by Eq. (51). The curvature ratios are 0.1 and 0.05. roughness effects on friction factors increase with increasing cur-
The relative roughnesses are 0.001, 0.002 and 0.004 for each curva- vature ratios and Reynolds numbers. When the curvature ratios
increase or the Reynolds numbers increase, the centrifugal force
ture ratio. Re varies from 1  104 to 6  105 . Ratios, f trans =f rough are
increases and the secondary flow intensities increase, which will
calculated and the results are shown in Fig. 8. When the flows in
cause more pressure drops in the helical tube.
the transitionally rough regime transfer to the fully rough regime,
Eq. (50) also shows that the curvature ratio effects on friction
the ratio approaches to unit. The trends of the ratio as the Reynolds
factors are influenced by relative roughness and the Reynolds
number increases reflect the curvature ratio effects and roughness

Fig. 8. Comparison of f trans =f rough with different curvature ratios and different
Fig. 6. Comparison of friction factors calculated by Eqs. (50), (52) and (53). roughnesses.
H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534 533

rough helical tubes are theoretically derived in this paper. The


parameters in the equations are determined by regression analysis
of experimental data of Clancy [1] and verified by experimental
data of McElligott [2]. The influences of curvature ratio, relative
roughness and the Reynolds number on friction factors in rough
helical tubes are analyzed. The main conclusions are,

(1) Two equations are theoretically derived for calculating the


friction factors for turbulent flows in rough helical tubes.
One is for the transitionally rough regime and the other is
for the fully rough regime. The friction factors are influenced
by the Reynolds number, curvature ratio and the relative
roughness.
(2) The parameters of the equations are obtained by regression
analysis of experimental data.
(3) The roughness has similar effects on friction factors for tur-
bulent flow in rough helical tubes as that in rough straight
Fig. 9. Comparison of f c;r =f c;s with different curvature ratios. tubes. When the roughness and the Reynolds number are
high enough, the friction factors become independent of
the Reynolds number.
(4) For turbulent flow in rough helical tubes, the roughness
effects on friction factors increase as the curvature ratio
increases. The roughness effects and curvature ratio effects
increase as the Reynolds number increase.

Acknowledgments

This work was financially supported by the National Natural


Science Foundation of China (51576103) and the National S&T
Major Project (Grant No. ZX06901).

References

[1] A.H. Clancy, Experimental Investigation of Pressure Drop in Helical Coils (AE
thesis), California Institute of Technology, Pasadena, California, 1949.
[2] R.H. McElligott, Investigation of Pressure Drop Through Helical Coils (AE
thesis), California Institute of Technology, Pasadena, California, 1948.
[3] Z.Y. Zhang, Z.X. Wu, D.Z. Wang, Y.H. Xu, Y.L. Sun, F. Li, Y.J. Dong, Current status
Fig. 10. Comparisons of f c;r =f str; r with different Reynolds numbers. and technical description of Chinese 2 times 250 MWth HTR-PM
demonstration plant, Nucl. Eng. Des. 239 (7) (2009) 12121219.
[4] H. Nariai, M. Kobayashi, T. Matsuoka, Friction pressure-drop and heat-transfer
coefficient of 2-phase flow in helically coiled tube once-through steam-
number. Friction factors in helical tubes, f c;r , are calculated by Eq. generator for integrated type marine water-reactor, J. Nucl. Sci. Technol. 19
(11) (1982) 936947.
(50) with curvature ratio of 0.1, and the relative roughness varies
[5] S.H. Yang, S.H. Kim, Y.J. Chung, H.S. Park, K.K. Kim, Experimental validation of
from 0.0008 to 0.0166. Friction factors in rough straight tubes with the helical steam generator model in the TASS/SMR code, Ann. Nucl. Energy 35
the same relative roughness, f str;r , are calculated by Eq. (54) pre- (1) (2008) 4959.
[6] W.R. Dean, Note on the motion of fluid in a curved pipe, London, Edinb. Dubl.
sented by Churchill [30]. Philos. Mag. J. Sci. 4 (20) (1927) 208223.
( "  0:9 #)2
[7] W.R. Dean, The stream-line motion of fluid in a curved pipe (second paper),
London, Edinb. Dubl. Philos. Mag. J. Sci. 5 (30) (1928) 673695.
e 7
f str; r 2log10 54 [8] V. Gnielinski, Heat transfer and pressure drop in helically coiled tubes, in:
7:4 Re Proceedings 8th International Heat Transfer Conference, San Francisco,
Hemisphere, Washington DC, 1986, pp. 28472854.
The ratio, f c; r =f str; r , reflects the curvature ratio effects on frica- [9] P. Srinivasan, S. Nandapurkar, F. Holland, Friction factors for coils, Trans.
Instrum. Chem. Eng 48 (1970) 156161.
tion factors. f c; r =f str; r is calculated with different Reynolds number [10] H.M. Ju, Z.Y. Huang, Y.H. Xu, B. Duan, Y. Yu, Hydraulic performance of small
and the results are shown in Fig. 10. The results in Fig. 10 shows bending radius helical coil-pipe, J. Nucl. Sci. Technol. 38 (10) (2001) 826831.
[11] K. Yamamoto, A. Aribowo, Y. Hayamizu, T. Hirose, K. Kawahara, Visualization
that the ratio, f c;r =f str; r , increases with the increasing roughness
of the flow in a helical pipe, Fluid Dyn. Res. 30 (4) (2002) 251267.
height, which means the curvature ratio effects on friction factors [12] J. Hart, J. Ellenberger, P.J. Hamersma, Single-phase and 2-phase flow through
in rough helical tubes increase with the increasing roughness. helically coiled tubes, Chem. Eng. Sci. 43 (4) (1988) 775783.
[13] R. Grundmann, Friction diagram of the helically coiled tube, Chem. Eng.
Process. Process Intensif. 19 (2) (1985) 113115.
[14] P. Mishra, S.N. Gupta, Momentum-transfer in curved pipes. 1. Newtonian
5. Conclusions fluids, Ind. Eng. Chem. Process Des. Dev. 18 (1) (1979) 130137.
[15] H. Ito, Friction factors for turbulent flow in curved pipes, J. Basic Eng. 81 (2)
(1959) 123134.
Helical tubes are extensively used as heat exchangers in chem-
[16] S. Vashisth, V. Kumar, K.D. Nigam, A review on the potential applications of
ical, petroleum and nuclear industries. Turbulent flows in helical curved geometries in process industry, Ind. Eng. Chem. Res. 47 (10) (2008)
tubes are more complex than those in straight tubes. However, 32913337.
previous investigations on friction factors in helical tubes are all [17] L.R. Austin, J. Seader, Fully developed viscous flow in coiled circular pipes,
AIChE J. 19 (1) (1973) 8594.
for smooth helical tubes. All the commercially used helical tubes [18] H. Soeberg, Viscous flow in curved tubesI. Velocity profiles, Chem. Eng. Sci.
are rough tubes. Two equations for calculating friction factors in 43 (4) (1988) 855862.
534 H. Zhao et al. / International Journal of Heat and Mass Transfer 95 (2016) 525534

[19] S.J. Liu, J.H. Masliyah, Axially invariant laminar-flow in helical pipes with a [31] P.K. Swamee, A.K. Jain, Explicit equations for pipe-flow problems, J. Hydraul.
finite pitch, J. Fluid Mech. 251 (1993) 315353. Div. 102 (5) (1976) 657664.
[20] K. Yamamoto, S. Yanase, T. Yoshida, Torsion effect on the flow in a helical pipe, [32] N.H. Chen, An explicit equation for friction factor in pipe, Ind. Eng. Chem.
Fluid Dyn. Res. 14 (5) (1994) 259273. Fundam. 18 (3) (1979) 296297.
[21] S. Yanase, N. Goto, K. Yamamoto, Dual solutions of the flow through a curved [33] L.F. Moody, Friction factors for pipe flow, Trans. ASME 66 (8) (1944) 671684.
tube, Fluid Dyn. Res. 5 (3) (1989) 191. [34] M. Adler, Currents in curved pipes, ZAMM J. Appl. Math. Mech./Z. fr Angew.
[22] Y. Lai, R. So, H. Zhang, Turbulence-driven secondary flows in a curved pipe, Math. Mech. 14 (1934) 257275.
Theor. Comput. Fluid Dyn. 3 (3) (1991) 163180. [35] H. Schlichting, Boundary Layer Theory, McGraw-Hill, New York, 1979. pp.
[23] R.L. Manlapaz, S.W. Churchill, Fully developed laminar flow in a helically 128131, 587591, 615623.
coiled tube of finite pitch, Chem. Eng. Commun. 7 (13) (1980) 5778. [36] H. Ito, Laminar flow in curved pipes, ZAMM J. Appl. Math. Mech. 49 (11) (1969)
[24] D.P. Ivan, C. Michele, Numerical prediction of turbulent flow and heat transfer 653663.
in helically coiled pipes, Int. J. Therm. Sci. 49 (4) (2010) 653663. [37] Y.X. Li, J.H. Wu, H. Wang, L.P. Kou, X.H. Tian, Fluid flow and heat transfer
[25] J. Nikuradse, Strmungswiderstand in rauhen Rohren, ZAMM J. Appl. Math. characteristics in helical tubes cooperating with spiral corrugation, Energy
Mech./Z. fr Angew. Math. Mech. 11 (6) (1931) 409411. Procedia 17 (2012) 791800.
[26] C.F. Colebrook, C. White, Experiments with fluid friction in roughened pipes, [38] A. Zachr, Analysis of coiled-tube heat exchangers to improve heat transfer
Proc. R. Soc. London Ser. A Math. Phys. Sci. (1937) 367381. rate with spirally corrugated wall, Int. J. Heat Mass Transfer 53 (19) (2010)
[27] C.F. Colebrook, Turbulent flow in pipes, with particular reference to the 39283939.
transition region between the smooth and rough pipe laws, J. ICE 11 (4) (1939) [39] S. Rainieri, F. Bozzoli, L. Cattani, G. Pagliarini, Compound convective heat
133156. transfer enhancement in helically coiled wall corrugated tubes, Int. J. Heat
[28] L.F. Moody, An approximate formula for pipe friction factors, Trans. ASME 69 Mass Transfer 59 (2013) 353362.
(12) (1947) 10051011. [40] S. Rainieri, F. Bozzoli, G. Pagliarini, Experimental investigation on the
[29] D.J. Wood, An explicit friction factor relationship, Civil Eng. 36 (12) (1966) 60 convective heat transfer in straight and coiled corrugated tubes for highly
61. viscous fluids: Preliminary results, Int. J. Heat Mass Transfer 55 (1) (2012)
[30] S.W. Churchill, Empirical expressions for the shear stress in turbulent flow in 498504.
commercial pipe, AIChE J. 19 (2) (1973) 375376.