Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 69 (2015) 259 – 268

International Conference on Concentrating Solar Power and Chemical Energy Systems,

SolarPACES 2014

Study on Flow Instability inside Multiple Parallel Pipes of Direct

Steam Generation
J. Zhoua, S. Wangb, L. Gaoa, L. Zhaoa, L. Pang*a
a
North China Electric Power University ,Changping District, Beijing 102206, China.
b
Guoneng(Shendong Power) Power Generation Lt d, Shenhua Group

Abstract

Flow instability in parallel pipes is closely related with the application of direct steam generation (DSG) for parabolic trough
solar power technology. Maldistribution of the flow rate may occur in heated parallel pipes with common inlet and outlet
manifolds. The theoretical analysis on four parallel pipes that undergoes a process of heating and evaporation during different
heating loads is carried out to indicate all the possible steady-state solutions. Stable solutions and unstable solutions are
differentiated through linear stability analysis. It is shown that the more number of heated pipes, the more stable steady state
solutions of the splitting ratios, and the greater range of uneven flow rate distribution. Transient analysis is applied to simulate
the response of the system to finite disturbances and the performance due to variations in operating conditions by Simulink of
Matlab. The results confirm the correctness of the linear stability analysis, and a hysteresis phenomenon is observed when a
solution is turning to another solution, even for unstable ones. In addition, A method is proposed to improve flow distribution
uniformity in parallel pipes by installing a resistant orifice at the inlet of each pipe where is not heated.

©
© 2015
2015TheTheAuthors.
Authors.Published
Publishedby by
Elsevier Ltd.Ltd.
Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer review by the scientific conference committee of SolarPACES 2014 under responsibility of PSE AG.
Peer review by the scientific conference committee of SolarPACES 2014 under responsibility of PSE AG
Keywords: direct steam generation; parallel pipes; flow intability; splitting ratio; two-phase flow

1. Introduction

The instability of evaporating two-phase flow in parallel pipes exists widely in boilers, steam generators, heat
exchangers, boiling water reactors and other devices which are related to evaporating fluid. Direct steam generation

* Corresponding author. Tel.: +8610-61772839.

E-mail address: Liping_pang@163.com

1876-6102 © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer review by the scientific conference committee of SolarPACES 2014 under responsibility of PSE AG
doi:10.1016/j.egypro.2015.03.030
260 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268

(DSG) of parabolic trough solar power technology consists of a row of horizontal parallel pipes, using parabolic
trough concentrator focusing solar power on each pipe to generate steam to drive turbine power generation system.
Compared with the traditional double-loop system, the parabolic trough solar power technology for direct steam
generation has advantages of low cost and high efficiency because of eliminating extra heat exchangers and extra
losses incurred while heat is transferred from oil to steam [1-4]. Howeverˈdue to the possible uneven flow rate
distribution and other instabilities relate to liquid-vapor flow in parallel pipes become an obstacle to direct steam
generation technology for commercial application.
Scholars have done a lot of detailed research on the instability of two-phase flow in vertical pipes [5-9]. In the
nuclear industry, two-phase flow in parallel pipes is of major interest related to the instability in boiling water
reactors (BWR) and loss of cooling accident (LOCA). These studies deeply discussed the instability problems such
as pressure drop oscillation (PDO), density wave oscillation (DWO) and flow excursion, get a lot of useful
conclusions and practical application in engineering[10-14]. Evaporating two-phase flow in horizontal pipes differ
from evaporating two-phase flow in vertical pipes by the flow patterns and by the absence of gravitational pressure
drop[15, 16].
The instability of two-phase flow in horizontal pipes mainly depends on the relationship between the fractional
pressure drop and flow rate[17]. Natan et al.[18] analyzed theoretically a system of two parallel pipes with common
inlet and outlet manifolds that undergoes a process of heating and evaporation. They developed a numerical
simulation to calculate the splitting of the flow between the two pipes, and the properties of the fluids along the
pipes. The pressure drop in this model was based on the local flow pattern.
Minzer et al.[19] measured the flow rate distribution and the pressure drop in a set of two parallel horizontal
evaporating pipes. The experimental results compared well with the theoretical steady-state analysis of Natan et
al.[18] . The stability criteria offered by Akagawa et al.[20] was used to distinguish between stable and unstable
steady states. Minzer et al.[21] then suggested a simplified model that is used both for steady-state solutions analysis
and for transient analysis. They presented a stability analysis to demarcate between stable and unstable steady-state
solutions. In addition, they carried out the corresponding experiments to confirm the correctness of the theoretical
analysis. And the experimental results compared well with the model.
Mordechai Baikin et al.[22] extended the theoretical model developed by Minzer et al[21] to a larger number of
pipes and different heating conditions. They mainly discussed a system of three pipes with common inlet and outlet
manifolds by carrying out steady-state solutions analysis and transient analysis to find the flow rate distribution for
different heating combinations. And their experimental results compared well with the theoretical results.
The present work is aimed to extend the work of Minzer et al.[21] and Mordechai Baikin et al.[22] to a system of
more parallel pipes. In this paper, the maldistribution of the flow rate in four parallel pipes with common inlet and
outlet manifolds is mainly discussed.

2. Analysis

Subcooled water that enters the system of four pipes with common inlet manifold and outlet manifold undergoes
a process of heating and evaporation, and the flow splits into the four heated parallel pipes. The first step of the
theoretical analysis is to obtain the relationship between the frictional pressure drop and flow rate. And it allows
obtaining the inlet pressure versus the flow rate for a specified outlet pressure and heat flux.
Since each pipe has the common inlet and outlet manifolds, the same pressure drop yields in the parallel pipes. It
is possible to obtain different combinations of flow rates in parallel pipes that satisfy the condition of equal pressure
drop. And linear stability analysis is performed to differentiate between stable and unstable solutions.

2.1. Single pipe analysis

According to the water state of the pipe exit, the flow within the pipe is subdivided into three regions, see Fig.1.
When the water exits as hot liquid we just have one section, the subcooled section. When the water exits as water-
vapor mixture we have two sections, the subcooled section and the evaporating two-phase section. When the water
exists as steam we get three sections, the subcooled section at the entrance, the evaporating two-phase section at the
middle and the superheated steam at the exit.
 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268 261

Fig. 1. Three possible flow sections in a single pipe.

The momentum equation is established for two-phase flow mixture inside a single parallel pipe:

μ μ ߲ܲ
(ߩ ‫ ݑ ߙܣ‬+ ߩீ ‫ ) ீݑ ீߙܣ‬+ (ߩ௅ ‫ߙܣ‬௅ ‫ݑ‬௅ଶ + ߩீ ‫ீݑ ீߙܣ‬ଶ ) = െ߬‫ ݏ‬െ ߩ݃‫ ߚ݊݅ݏܣ‬െ ‫ܣ‬ (1)
μt ௅ ௅ ௅ μx ߲‫ݔ‬

Where A is the cross-sectional area of the pipe, ĮL and ĮG are the liquid and gas void fractions, ȡL and ȡG are the
liquid and gas densities, uL and uG are the liquid and gas velocities, IJ is the wall shear stress, s is the pipe periphery,
ȕ is the pipe inclination angle, μt is the time rate of change, ߲‫ ݔ‬is the change rate along pipe length.
By integrating Eq.1 and substituting the velocities and the void fractions in terms of the mass flow rate (W), the
vapor quality (X) and the slip ratio (S):

‫߲ ܮ‬
߂ܲ = (߂ܲ)ƒ + (߂ܲ)௚ + (߂ܲ)௔ + (ܹ) (2)
‫ݐ߲ ܣ‬

Where ǻ3f is the frictional pressure drop, ǻ3g is the gravitational pressure drop, (ȟP)௔ is the accelerated pressure
drop. And where (ȟP)௔ can be expressed as:

ܹ
(ȟP)௔ = ( )ଶ [‫ܤ‬୭୳୲ െ ‫ܤ‬୧୬ ] (3)
‫ܣ‬

ܺ ܵ(1 െ ܺ) 1െܺ
‫ܤ‬୭୳୲ = ( + )(ܺ + ) (4)
ߩீ ߩ௅ ܵ

When the water exits as hot liquid (X=0), ‫ܤ‬୭୳୲ = 1Τߩ௅ . When the water exits as water-vapor mixture (0<X<1), ‫ܤ‬୭୳୲
is the Eq.4. When the water exists as superheated steam (X=1), ‫ܤ‬୭୳୲ = 1Τߩீ . And where ‫ܤ‬୧୬ = 1Τߩ௅ .
The frictional pressure drop and the gravitational pressure drop of three sections are calculated respectively. The
pressure drop of single-phase is calculated using the average of the local pressure drops at the beginning and the end
of corresponding section. In this work, we special introduce the frictional pressure drop for the two-phase mixture.
For the two-phase mixture, a valid correlation is needed to estimate closely the frictional pressure drop in the
middle of the two-phase section. Thus, the pressure drop in the middle of the mixture section is calculated using the
method proposed by Chisholm as following.

‫ܥ‬ 1
(߂ܲƒ )ெ = ߂ܲ௅ை (1 + + ଶ )‫ܮ‬ெ (5)
ܺெ ܺெ

ȟܲƒ ȟܲƒ
ଶ
ܺெ = ( )௅ை ൘( )ீை (6)
ȟ‫ܮ‬ ȟ‫ܮ‬

Where ܺெ is Lockhart-Martinelli coefficient. ߂ܲ௅ை and ߂ܲீை are the pressure drop of liquid-phase and the gas-phase
through the pipe alone. The coefficent of Eq. 5 is listed in reference[23].
Rearranging Eq.2, the equation can be written as:
262 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268

‫ ܮ‬dܹ
= െ‫ )ܹ(ܨ‬+ ܲ୧୬ െ ܲ୭୳୲ (7)
‫ ܣ‬d‫ݐ‬

Where ‫ )ܹ(ܨ‬is the total pressure drop along the pipe:

‫( = )ܹ(ܨ‬ȟP)ƒ + (ȟP)௚ + (ȟP)௔ (8)

For multiple parallel pipes, a momentum equation is obtained for each pipe:

‫ ܮ‬dܹ௜
= െ‫ܨ‬௜ (ܹ௜ ) + ܲ୧୬ െ ܲ୭୳୲ (9)
‫ ܣ‬d‫ݐ‬

Where the subscript i is the number of the pipes, and ‫ܨ‬௜ (ܹ௜ ) is the pressure drop of each pipe. By summing Eq.9, we
can obtain:

‫ ܮ‬dܹ௜௡
= െ ෍ ‫ܨ‬௜ (ܹ௜ ) + ܰܲ୧୬ െ ܰܲ୭୳୲ (10)
‫ ܣ‬d‫ݐ‬
௜

2.2. Steady-state solutions

For the case of steady-state, dܹ/d‫ = ݐ‬0, Eq.7 can be written as:

ܲ୧୬ = ‫ )ܹ(ܨ‬+ ܲ୭୳୲ (11)

The relationship between the pressure drop along a single pipe and the flow rate is obtained in fig.2. For a special
inlet pressure one may have 1, 2 or 3 steady-state solutions. For multiple pipes, since each pipe has the common
inlet and outlet manifold, the same pressure drop yields in the parallel pipes. It is possible to obtain different
combinations of flow rates in parallel pipes that satisfy the condition of equal pressure drop. The total inlet flow rate
is:

ܹ௜௡ = ෍ ܹ௜௝ , (݅ = 1,2, … , ܰ; ݆ = 1 ‫ ݎ݋‬2 ‫ ݎ݋‬3) (12)

௜,௝

Where the subscript i is the number of the parallel pipes, the subscript j is possible steady-state solutions in a single
pipe. Stable and unstable state solutions are described through linear stability analysis by Minzer et al. [21].

2.3. Transient analysis

Eq.9 and Eq.10 are time-dependent equations for Wi and Win. They are used to simulate the transient response of
the system to finite disturbances by applying Simulink of Matlab.
In this work, transient simulations are carried out to predict the system response to finite disturbances and the
system performance due to variations in operating conditions such as total inlet flow rate and heating power.

3. Steady-state analysis

It is assumed that the cross-sectional area of the heated pipe and the heating power of per unit length are constant.
Results of the calculation model are showed for the following case: pipe length 3 m, pipe diameter 6 mm, inlet water
temperature 60ć, heating power 1.2 KW/m. Fig.2 shows the inlet pressure versus flow rate for a single pipe.
 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268 263

Fig. 2. The inlet pressure versus flow rate for a single pipe

For four parallel pipes system, we main discussed three different heating conditions:

x Case 1: 1 pipe heated 1.2KW/m, 3 pipes unheated.

x Case 2: 2 pipes heated 1.2KW/m, 2 pipes unheated.
x Case 3: 3 pipes heated 1.2KW/m, 1 pipe unheated.

Fig.3-Fig.5 show the steady state solutions of inlet pressure and flow rate splitting ratio (ܴ௜ = ܹ௜ Τܹ୧୬ ) versus
total flow rate for three different cases. The red solid line represent the unstable solutions, the other color solid lines
represent the stable solutions with different splitting ratios.
Fig.3a and b shows the steady state solutions of inlet pressure and splitting ratio versus total flow rate for the case
1. It yields two kinds of splitting ratios from low to high total inlet flow rate: When flow rate is very low, the stable
steady-state solutions shown in grey lines where three pipes which unheated get almost 1/3 of total inlet flow rate,
and the heated pipe just get quite a small part of the total flow rate. The flow rate distribution is not uniform. By
increasing the flow rate it starts to get into the two-phase flow region. Since the impact of the negative slope of
pressure drop versus flow rate one can obtain two solutions which is shown in grey lines and black line. By
increasing the flow rate even more, even distribution is obtained in black line.
The steady state solutions of inlet pressure and splitting ratio versus total flow rate for the case 2 is shown in
fig.4a and b. It yields three different splitting ratios from low to high flow rates: When flow rate is very low, the
stable steady-state solutions shown in green lines where two unheated pipes get almost 50% of the total inlet flow
rate each, and two heated pipes just get the rest of the total flow rate. By increasing the flow rate it starts to get into
the two-phase flow region that one may obtain three different solutions as shown in green lines, grey lines and black
line. By increasing the flow rate even more, an even distribution is obtained in black line.
Fig.5a and b shows the steady state solutions of inlet pressure and splitting ratio versus total flow rate for the case
3. It yields four kinds of splitting ratio from low to high total inlet flow rate: When flow rate is very low, the stable
steady-state solutions shown in blue lines where one pipe which unheated get almost 94% of the total inlet flow rate,
and three heated pipes just get 2% of the total flow rate each. By increasing the flow rate it reaches the two-phase
flow region one may obtain four different solutions which are shown in blue lines, green lines, grey lines and black
line. By increasing the flow rate, the water of each pipe exists as liquid, an even distribution is obtained as shown in
black line.
264 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268

Fig. 3. (a) inlet pressure versus total flow rate for case 1; (b) splitting ratio versus total flow rate for case 1

Fig. 4. (a) inlet pressure versus total flow rate for case 2; (b) splitting ratio versus total flow rate for case 2

Fig. 5. (a) inlet pressure versus total flow rate for case 3; (b) splitting ratio versus total flow rate for case 3
 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268 265

Fig.6 shows the steady state solutions for the splitting ratio versus total flow rate for the case of equally heated
Q=1.2KW/m. Through Fig.6 the stable steady-state solution is not unique even for the case of equal heating of the
four pipes. For unequal heating higher flow rate may take place in the less heated pipes, and vice versa. It causes the
negative influence to the system during actual operation.

Fig. 6. Steady state solutions for the splitting ratio versus total flow rate, each pipe equally heated Q=1.2KW/m

4. Transient analysis

In this work, transient simulations are carried out to predict the system response to finite disturbances and the
system performance due to variations of inlet flow rate. Simulations are under the case of equal heating power of 1.2
KW/m for each pipe.
The steady state solution for splitting ratio versus time is showed in Fig.6. A case that the spontaneous process
from a hypothetical unstable steady state (point A) where each pipe receives the flow rate of 0.01kg/s is simulated.
By applying a small perturbation on the flow rate of the third pipe, the transient response of the flow rate and inlet
pressure of each pipe is shown in Fig.7.
Fig.7a presents the variations of the flow rate and the inlet pressure versus time, and Fig.7b indicates the transient
path of inlet pressure versus flow rate in each of the four pipes. It is clearly seen that there is a hysteresis
phenomenon at the beginning of the simulation, the flow rate of each pipe keeps constant and changes after nearly
1s. Then the system is shifted to a case where three of the four pipes gets almost 30% of the total inlet flow rate
corresponding to the points B which is indicated in Fig.6 and Fig.7b, and the other pipe decreases to the flow rate of
0.0015kg/s, quite a small part of the total flow rate. As the simulation goes on, another hysteresis phenomenon is
observed at point B, a solution which is unstable. Then two of them decrease to the flow rate of 0.0015kg/s each,
and another pipe increases to about 0.036kg/s, which can be observed in Fig.6 or Fig.7b at points C and D. The
system reaches to the stable steady state.
Fig.8 shows the results of transient simulation for the case of linear decreasing the total inlet flow rate from 0.12
kg/s to 0.04 kg/s (from the 1 second to the 3 second), namely from a stable solution to a unstable solution. As we
can see, the simulation starts with even flow rate distribution at point A, keeping constant until the flow rate get into
the two-phase flow region. As the simulation goes on, the splitting ratios of three pipes decreases to 0.02 each,
corresponding to the point B, and the other pipe increase to about 0.94 corresponding to the point C which are
indicated in Fig.8b. The system reaches to the stable steady state then.
266 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268

Fig. 7. (a) variations of the flow rate and the inlet pressure versus time; (b) transient path of inlet pressure versus flow rate in each pipe

Fig. 8. (a) variations of the splitting ratio and the inlet pressure versus time; (b) transient path of inlet pressure versus flow rate in each pipe

5. Flow rate distribution improvement

By installing a resistive device at the inlet of each pipe where is not heated, the pressure drop equation becomes:

‫߲ ܮ‬
ȟP = (ȟP)ி + (ȟP)௚ + (ȟP)௔ + (ȟP)௧ + (ܹ) (13)
‫ݐ߲ ܣ‬

Where (ȟP)௧ is the pressure drop produced by the inlet resistive device. It is assumed that:

ߩ௅ ‫ݑ‬ଶ
(ȟP)௧ = ߦ (14)
2

Where ߦ is the coefficient of local resistance, ‫ ݑ‬is the liquid velocity of pipe inlet.
Fig.9 shows the inlet pressure versus flow rate for a single pipe with extra resistance for different local resistant
coefficient ߦ. The bold line 1 represents the pressure drop versus flow rate without the inlet resistance. The dotted
line 2 and 3 are the pressure drop of resistive device with different coefficient ߦ alone. And the solid line 2' and 3'
are the pressure drop versus flow rate by installing a resistive device at the inlet of the pipe corresponding to the line
 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268 267

2 and 3. By increasing the resistance of inlet, the negative slope of pressure drop curve becomes smaller. Finally, it
becomes monotonically rise show in line 3'. It fundamentally eliminates the side-effect of multiple solutions (the
negative slope of the evaporating zone), so as to achieve the purpose of uniform distribution. That means every
parallel pipe inlets will install an orifice to overcome the two-phase flow instability.

Fig. 9. (a) The inlet pressure versus flow rate for a single pipe with extra resistance

6. Conclusion

The splitting ratios in four parallel pipes with a common inlet and outlet manifolds are calculated. Three different
heating conditions are simulated.
It is shown that the more number of heated pipes, the more stable steady state solutions of the splitting ratios, and
the greater range of uneven flow rate distribution. In addition, the steady-state analysis shows that the stable steady-
state solution is not unique even for the case of equal heating. For unequal heating higher flow rate may take place
in the less heated pipes, and vice versa. The result causes the negative influence to the system for actual operation.
Transient analysis simulates the transient response of the system to finite disturbances by applying Simulink of
Matlab. The final results confirm the correctness of the linear stability analysis. In addition, a hysteresis
phenomenon is observed when a solution is turning to another solution, even for unstable ones.
It is found that the flow resistance characteristic along the pipes is changed by installing a resistive device at the
inlet of each pipe where is not heated. It fundamentally eliminates the side-effect of the pressure drop curve (the
negative slope of the evaporating zone), so as to achieve the purpose of uniform distribution.
The final conclusion could be used for the actual design of DSG parabolic trough solar power technology. The
designer could avoid the unstable regions when planning the operating states at different sunlight heating load.

7. Acknowledgment

This work is supported by Project on Key Basic Research of Science and Technology Department of Hebei
Province in P. R. China (13964306D).

References

[1] Eck M, Steinmann W-D. Direct steam generation in parabolic troughs: first results of the DISS project[J]. Journal of solar energy
engineering. 2002, 124(2): 134-139.
[2] Price H, Lupfert E, Kearney D et al. Advances in parabolic trough solar power technology[J]. Journal of solar energy engineering. 2002,
124(2): 109-125.
268 J. Zhou et al. / Energy Procedia 69 (2015) 259 – 268

[3] Zarza E, Valenzuela L, Leon J et al. Direct steam generation in parabolic troughs: Final results and conclusions of the DISS project[J].
Energy. 2004, 29(5): 635-644.
[4] Zarza E, Valenzuela L, León J et al. The DISS project: direct steam generation in parabolic trough systems. Operation and maintenance
experience and update on project status[J]. Journal of solar energy engineering. 2002, 124(2): 126-133.
[5] Aritomi M, Aoki S, Inoue A. Instabilities in parallel channel of forced-convection boiling upflow system,(I) Mathematical model[J]. Journal
of Nuclear Science and Technology. 1977, 14(1): 22-30.
[6] Davies A, Potter R. Hydraulic stability: an analysis of the causes of unstable flow in parallel channels: UKAEA Atomic Energy
Establishment; 1966.
[7] .DNDo 6 9H]LUR÷OX 7 g]ER\D ',1 et al. Transient boiling flow instabilities in a multi-channel upflow system[J]. Wärme-und
Stoffübertragung. 1977, 10(3): 175-188.
[8] Veziroglu T, Lee S. Boiling flow instabilities in parallel channels. In: Proceedings of the Institution of Mechanical Engineers, Conference
Proceedings; 1969: SAGE Publications; 1969. p. 7-17.
[9] Xiao M, Chen X, Zhang M et al. A multivariable linear investigation of two-phase flow instabilities in parallel boiling channels under high
pressure[J]. International journal of multiphase flow. 1993, 19(1): 65-77.
[10] Kakac S, Veziroglu T. A review of two-phase flow instabilities. In: Advances in Two-phase Flow and Heat Transfer: Springer; 1983:577-667.
[11] Ozawa M. Flow instability problems in steam-generating tubes[J]. Steam Power Engineering-Thermal and Hydraulic Design Principles.
1999: 323-385.
[12] Pederson RJ, May EK. Flow instability during direct steam generation in a line-focus solar-collector system[J]. NASA STI/Recon Technical
Report N. 1982, 83: 20404.
[13] Yadigaroglu G. Two-phase flow instabilities and propagation phenomena. In: In Von Karman Inst. for Fluid Dyn. Two-Phase Flows in Nucl.
Reactors, Vol. 2 71 p (SEE N79-28497 19-34); 1978; 1978.
[14] Yüncü H, Kakaç S. Fundamentals of two-phase flow instabilities. In: Two-Phase Flow Heat Exchangers: Springer; 1988:375-406.
[15] Collier JG, Thome JR. Convective boiling and condensation: Oxford University Press; 1994.
[16] Ozawa M, Nakanishi S, Ishigai S. Flow instabilities in multi-channel boiling systems[J]. ASME paper. 1979: 1-7.
[17] Ledinegg M. Instability of flow during natural and forced circulation[J]. Die Warme. 1938, 61(8): 891-898.
[18] Natan S, Barnea D, Taitel Y. Direct steam generation in parallel pipes[J]. International Journal of Multiphase Flow. 2003, 29(11): 1669-1683.
[19] Minzer U, Barnea D, Taitel Y. Evaporation in parallel pipes––splitting characteristics[J]. International Journal of Multiphase Flow. 2004,
30(7-8): 763-777.
[20] Akagawa K, KONO M, SAKAGUCHI T et al. Study on distribution of flow rates and flow stabilities in parallel long evaporators[J].
Bulletin of JSME. 1971, 14(74): 837-848.
[21] Minzer U, Barnea D, Taitel Y. Flow rate distribution in evaporating parallel pipes—modeling and experimental[J]. Chemical Engineering
Science. 2006, 61(22): 7249-7259.
[22] Baikin M, Taitel Y, Barnea D. Flow rate distribution in parallel heated pipes[J]. International Journal of Heat and Mass Transfer. 2011,
54(19): 4448-4457.
[23] Chisholm D. A theoretical basis for the Lockhart-Martinelli correlation for two-phase flow[J]. International Journal of Heat and Mass
Transfer. 1967, 10(12): 1767-1778.