190530

Research ArticleResearch on Blade Thickness Influencing Pump as Turbine

Sun-Sheng Yang,1 Chao Wang,2 Kai Chen,1 and Xin Yuan1

1 Research Center of Fluid Machinery Engineering and Technology, Jiangsu University, Zhenjiang, Jiangsu 212013, China2 Patent Examination Cooperation Jiangsu Center of the Patent Office, SIPO, Suzhou, Jiangsu 215163, China

Correspondence should be addressed to Sun-Sheng Yang; [email protected]

Received 6 March 2014; Accepted 13 May 2014; Published 12 June 2014

Academic Editor: Dongliang Sun

Copyright 2014 Sun-Sheng Yang et al.This is an open access article distributed under theCreativeCommonsAttribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Research on the efficiency improvement of pump as turbine (PAT) is inadequate. Blade thickness is an important geometryparameter in blade design. To explore effects of blade thickness on the influence of PAT, numerical research on three differentspecific speeds of PATs with different blade thickness was carried out. Their performance changes with blade thickness werepresented. Besides, the variations of hydraulic loss distributionwith increasing blade thickness were performed.Theoretical analysisgives a reasonable explanation for the performance change. Results show that PATs flow versus efficiency curve (-) is lowered;flow versus head (-) curve and flow versus power (-) curve are increased with increasing blade thickness. The increase of- is mainly attributed to the increase of theoretical head caused by increasing blockage of impeller inlet area. Hydraulic lossdistribution analysis indicates that the total hydraulic loss within PAT is increased with increasing blade thickness. The increaseof - curve is a combined effect of the increase in theoretical head and the total hydraulic loss. The decrease of efficiency withincreasing blade thickness indicates that the blade thickness of PAT should be as thin as possible if its strength could be met.

1. Introduction

Pumps are reversible machines. A pump can run as a turbinewith an acceptable efficiency. Due to its apparent advantagesof being cheap and readily available worldwide, pump asturbine is one of the best options for small hydropowerrecovery [14].

The efficiency of PAT is reported to be almost the same asin pumpmode [5].The efficiency improvement of PATmeansgenerating more energy from current limited hydropowerresources. Therefore, researchers have been focused on itsefficiency improvement. Singh and Nestmann [6, 7] carriedvarious possibilities of modifying current pump geometryto improve its efficiency. Yang et al. analyzed the flow field[8] within PAT and investigated effects of splitter blades [9]and blade wrap angle [10] on the performance influence ofPAT. Yang et al. [11] sought to find out a high efficiency PATsvolute designmethod through numerical research into volutemain geometric parameters influencing PAT. Derakhshanet al. [12, 13] used a gradient-based optimization algorithmto optimize PATs blade profiles. As could be seen researchon the efficiency improvement of PAT is inadequate. Blade

thickness is a main geometry parameter to be determined inhydraulic design.Therefore, this paper seeks to find the effectsof blade thickness on the influence of PAT.

Experimental and numerical researches are two waysfrequently adopted in PATs research. Experimental researchcould give people a convincing result. However, it has thedrawbacks of being time consuming and expensive. Besides,limited by the accuracy of test bed, small changes of per-formance may not be easily observed. CFD as a promisingtechnique provides a powerful substitution and has been usedin the research of rotatingmachinery [14, 15] and PAT [16, 17].In this paper, effects of blade thickness influencing differentspecific speeds of PATs performances were investigated usinga validated CFD approach. Hydraulic loss distribution andtheoretical analysis were performed; the reasons for thevariation of PATs performance were explored.

2. Numerical Investigation

2.1.MainGeometric Parameters. PATs covering low,medium,and high specific speeds were designed. Table 1 lists their

Hindawi Publishing CorporationAdvances in Mechanical EngineeringVolume 2014, Article ID 190530, 8 pageshttp://dx.doi.org/10.1155/2014/190530

2 Advances in Mechanical Engineering

XZ

Y

(a) = 57

X

ZY ZYYYYYY

XX

(b) = 119

XZ

Y

XZ XXXXXXXX

(c) = 168

Figure 1: Mesh of PATs.

Table 1: Design parameters of the PATs.

Q/m3h1 /m /rmin1

100 40 1500 57120 43 3000 119125 28 3000 168

Table 2: Main geometric parameters of the PATs.

Impeller

57 119 168Impeller outlet diameter

1/mm 102 90 90

Impeller diameter2/mm 255 158 132

Length of wear ring /mm 15 17 15Impeller hub diameter

/mm 30 30 20

Blade inlet angle 2() 20 30 30

Blade outlet angle 1() 19.5 25.05 31.05

Blade wrap angle () 100 80 60Impeller inlet width

2/mm 14.38 19 27

Blade number 11 10 10Volute

Volute base circle diameter3/mm 266 168 150

Volute inlet width 3/mm 26 34 40

Volute inlet pipe diameter4/mm 65 80 80

Volute cross section shape/mm Round Round Round

design parameters. Table 2 lists their main geometric param-eters [18].

2.2. Mesh Generation. ICEM-CFD was used to generatestructured hexahedral grid for each component part [19].A grid independent test of the low specific speed PAT wasperformed; it was found that when mesh elements werearound 1 million, the variation of efficiency was within 0.5%.The final mesh number of the PAT was over 1 million.The +

near the boundary wall was around 40 [20]. For comparison,mesh number of different designs was almost the same.Figure 1 gives a general view of the generated meshes.

2.3. Solution Parameters. ANSYS-CFX was selected in thesolution of 3DNavier-stokes due to its characteristics of beingrobust and having fast convergence. Steady state simulationwas carried out. The turbulence selected was standard -model. The advection scheme was set to high resolution.The convergence criterion was 106. The fluid selected wasideal water at 25C. All the wall surface roughness withinthe control volume was set to 50m. Standard wall functionwas used for the boundary layer treatment. The boundaryconditions were set to inlet, static pressure, and outlet,mass flow outlet [21, 22]. By changing the mass flow rate,performance curves of the PAT were acquired.The interfacesbetween two stationary components and rotary and station-ary components were set to general grid and rotor statorinterface, respectively.

3. Validation of Numerical Simulation

3.1. Experimental Set-Up. A laboratory model of an openPAT test rig, as shown in Figure 2, was set up at JiangsuUniversity. High pressure fluid required for PATs energyrecovery was supplied by a feed pump. An electric eddycurrent dynamometer (EECD) was installed to measureand consume energy generated by PAT and to regulate itsrotational speed. The discharge was measured by a turbineflow meter. PATs inlet and outlet pressures were measuredby pressure sensors. The uncertainties of measured requiredpressure head , flow rate , hydraulic power

, generated

shaft power , and efficiency are 0.14%, 0.50%, 0.52%,1.08%, and 1.20%, respectively.

3.2. Comparison between Numerical and Experimental Re-sults. PAT with specific speed of 57 was selected for thevalidation of numerical accuracy. The blade thickness oftest PAT is 4mm. Other geometric parameters are listed in

Advances in Mechanical Engineering 3

Flow meter

PATMotor

Pressure transmitterValve

Computer

Pump EECD

Figure 2: An open PAT test rig.

Figure 3: Test PAT.

Table 2. Figure 3 shows the test PAT. Comparison betweenexperimental and numerical results is presented in Figure 4.

As is shown in Figure 4, the tendency of PATs numericalpredicted performance curves is in agreement with thoseof experimental. Numerical predicted efficiency, pressurehead, and shaft power are higher than those of experimental.The overprediction of efficiency, pressure head, and shaftpower may attribute to the neglect of leakage loss throughbalancing holes, surface roughness value set, and mechanicalloss caused bymechanical seal and bearings.The comparisonbetween experimental and steady state numerical resultsindicates that the grid and turbulence model selected arereasonable for PATs performance prediction. As a result,ANSYS-CFX can be used to predict the performance of PAT.

4. Results and Analysis

4.1. Performance Analysis. Numerical simulations of threePATs with different blade thickness were performed. Themain geometric parameters of the investigated PATs are listedin Table 2. Figure 5 plots their performance curves. Table 3lists their performance at their BEPs.

Figure 5 andTable 3 show that PATs- and- curvesare increased; - curve is lowered with the increase ofblade thickness. This illustrates that for the same flow rateits required pressure head and generated shaft power areincreased and its efficiency is decreased when there is anincrease of blade thickness.Thus, it can be concluded that for

70 80 90 100 110 12020

30

40

50

60

70

2

4

6

8

10

12

25

30

35

40

45

50

55

(%

)

Q (m3/h)

H(m

)

CFD

CFD HEXP

EXP HCFD PEXP P

P(k

W)

Figure 4: Comparison between experimental andnumerical results.

Table 3: PATs BEPs of impellers with different blade thickness.

/mm /m3h1 /m /kW /%

572 100.00 40.20 7.79 71.174 100.00 40.65 7.85 70.936 100.00 41.01 7.89 70.70

1192 120.00 42.44 11.31 81.614 120.00 42.70 11.37 81.476 120.00 44.39 11.71 80.73

1682 125.00 27.59 7.69 81.914 125.00 28.17 7.81 81.426 125.00 29.20 8.02 80.72

Table 4: Lists of hydraulic loss differences at their BEPs.

/mm volute/m impeller/m pipe/m total/m /m

57 4 0.15 0.02 0.06 0.22 0.456 0.21 0.10 0.11 0.42 0.81

119 4 0.05 0.14 0.01 0.11 0.276 0.18 0.92 0.00 0.75 1.95

168 4 0.08 0.31 0.02 0.24 0.586 0.43 1.03 0.04 0.64 1.61

the efficiency improvement blade thickness should be as thinas possible, supposing that its strength is satisfied.

4.2. Hydraulic Loss Distribution Analysis. The variations ofhydraulic loss distribution with blade thickness within PATsthree zones (volute zone to outlet pipe zone as definedin Figure 6) are discussed in this section. The differenceof hydraulic loss distribution between impellers with bladethickness of 2mm and other blade thicknesses is presentedin Figure 7. Table 4 lists their hydraulic loss difference at theirBEPs.


70 80 90 100 110 120 130

60

62

64

66

68

70

7274

30

35

40

45

50

55

4

6

8

10

12

Q (m3/h)2462H4H

6H2P4P6P

(%

)

H(m

)

P(k

W)

(a) = 57

77

78

79

80

81

82

83

35

40

45

50

55

60

65

6

8

10

12

14

16

18

20

Q (m3/h)100 110 120 130 140 150 160

H(m

)

P(k

W)

(%

)

2 4 6 2 H4 H

6 H2 P4 P6 P

(b) = 119

75

76

77

78

79

80

81

82

83

2224262830323436384042

4

5

6

7

8

9

10

11

12

13

Q (m3/h)100 110 120 130 140 150 160

H(m

)

P(k

W)

(%

)

2 4 6 2 H4 H

6 H2 P4 P6 P

(c) = 168

Figure 5: Performance curves of PATs with different blade thickness.

As indicated in Figure 7 and Table 4, the hydraulic losswithin impeller and the total hydraulic loss are increasedwithincreasing blade thickness. The hydraulic loss within voluteof the low specific speed PAT is grown, while that withinthe volutes of the medium and high specific speed PATs isdropped. And the variation of hydraulic loss within outletpipe is negligible. It can be concluded that the growth ofhydraulic loss within impeller is mainly responsible for theincrease of total hydraulic loss within PAT with increasingblade thickness. The hydraulic loss within impeller can be

calculated by equation impeller = 2. Table 4 shows that, as

blade thickness increases from 2mm to 6mm, the hydraulicloss factor is increased by 1.89%, 19.22%, and 30.50% for thelow, medium, and high specific speeds of the PATs at the BEP.

4.3. Flow Field Analysis. The velocity streamline distributionat the middle span of impeller blade to blade surface ispresented in Figure 8. For the purpose of comparison, theflow rate is the same for both cases. It can be seen fromFigure 8 that the velocity within impeller is increased with


Volute zone

Impeller zone

Outlet pipezone

Figure 6: Flow zones in a PAT control volume.

increasing blade thickness. The local and frictional hydraulicis in proportion to the square of velocity; therefore itshydraulic loss within impeller is increased with increasingblade thickness.

5. Theoretical Analysis

5.1. Output Shaft Power. From the fundamentals of energytransfer in turbines (steady flow equation and Euler turbineequation), the output mechanical shaft power can be repre-sented by (1) [2325]:

= mech leak, (1)

= Euler, (2)

Euler =(22 11)

=(22cot2 1(1 1cot1))

=(2(/

2) cot

2 1(1 (/

1) cot

1))

.

(3)

As could be seen from (1), (2), and (3), PATs - curveis in inverse proportion to inlet area

2. Figure 9 shows two

bladeswith different thickness. An increase of blade thicknesswould cause a decrease of impeller inlet area

2. Therefore,

there would be an increase of the theoretical head

and the generated shaft power . Thus, we could see thatPATs - curve would increase with the increase of bladethickness, which is in agreement with the results presented inSection 4.1.

5.2. Required Pressure Head. PATs required pressure headcan be represented as the sum of theoretical head and thelosses within all the three zones of the PATs control volumeillustrated in

= + total. (4)

Hydraulic loss distribution (Section 4.2) and theoreticalhead (Section 5.1) analysis shows that the total hydraulic losstotal and theoretical head are increased with the increaseof blade thickness. Therefore, PATs - curve will alsoincrease which is in agreement with the results in Figure 5.

5.3. Efficiency. PATs hydraulic efficiency can be representedby

=

+ total. (5)

An increase of blade thickness would cause an increaseof both theoretical head and total hydraulic loss. Therefore,PATs efficiency is decreased when there is an increase ofblade thickness.

6. Conclusions

Research on blade thickness to the influence of PAT wasperformed numerically. PATs performance of different bladethickness was presented. Results show that its - curveis lowered and - and - curves are increased withincreasing blade thickness. It could be concluded that fromthe efficiency point of view the blade thickness is expected tobe as thin as possible.

Hydraulic loss analysis within PAT indicates that, with theincrease of blade thickness, the hydraulic loss within impellerand the total hydraulic loss are increased. The hydraulic losswithin volute of the low specific speed PAT is grown, whilethat within the volutes of the medium and high specificspeed PATs is dropped.The variation of hydraulic loss withinoutlet pipe is negligible. Flow field analysis indicates thatit is the increase of velocity within impeller that causedthe increase of hydraulic loss within impeller. Theoreticalanalysis shows that the theoretical head is increased with theincrease of blade thickness. Therefore, PATs output poweris increased. PATs actual required pressure head is the sumof theoretical head and the total hydraulic loss within PATscontrol volume. Thus, the head curve is increased as bladethickness is increased.

Nomenclature

: Area, m22: Impeller inlet width, mm1: Impeller outlet diameter, mm2: Impeller inlet diameter, mm3: Volute base circle diameter, mm4: Volute inlet diameter, mm: Impeller hub diameter, mm: Acceleration of the gravity, m/s2: Hydraulic loss: Head, m: Length of impeller wear ring, mm: Rotational speed, rpm: Specific speed,

= 3.65/

3/4

: Power, W, kW: Discharge, m3/h


70 80 90 100 110 120

0.0

0.2

0.4

0.6

0.8

1.0

4 volute4 impeller

4 pipe4 total

Q (m3/h)

h

(m)

(a) = 57, = 4mm

70 80 90 100 110 120

0.0

0.2

0.4

0.6

0.8

1.0

6 volute6 impeller

6 pipe6 total

Q (m3/h)

h

(m)

(b) = 57, = 6mm

100 110 120 130 140 150

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Q (m3/h)

h

(m)

4 volute4 impeller

4 pipe4 total

0.2

(c) = 119, = 4mm

100 110 120 130 140 150

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Q (m3/h)

6 volute6 impeller

6 pipe6 total

h

(m)

0.2

(d) = 119, = 6mm

100 110 120 130 140 150

0.0

0.5

1.0

1.5

2.0

Q (m3/h)

h

(m)

4 volute4 impeller

4 pipe4 total

0.5

(e) = 168, = 4mm

6 volute6 impeller

6 pipe6 total

Q (m3/h)

h

(m)

100 110 120 130 140 1500.5

0.0

0.5

1.0

1.5

2.0

(f) = 168, = 6mm

Figure 7: Difference of hydraulic loss distribution with different blade thickness.


18.63

13.97

9.31

4.66

0.00

Velo

city

(ms

1 )

(a) = 168, = 2mm (b)

= 168, = 6mm

Figure 8: Streamline distribution within PAT.

= 2mm

= 6mm

Figure 9: Variation of blade thickness.

: Peripheral velocity, m/s: Absolute velocity, m/s+: plus: Blade number.

Greek Symbols

: Blade wrap angle, (): Efficiency: Absolute flow angle, (): Relative flow angle, (): Blade thickness.

Subscripts

1: Low pressure side fluid2: High pressure side fluid: Hydraulicleak: Leakage: Theoretical: Meridionalmech: MechanicalV: Volumetric.

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper.

Acknowledgments

This paper is financially supported by the Natural Sci-ence Foundation of Jiangsu Province entitled Researchon impeller internal flow and design theory of pump asturbine (BK20130517) and Jiangsu University Foundation(13JDG081), A Project Funded by the Priority Academic Pro-gram Development of Jiangsu Higher Education Institutions(PAPD).

References

[1] J. M. Chapallaz, P. Eichenberger, and G. Fischer, Manual onPumps Used as Turbines, Vieweg, Braunschweig, Germany,1992.

[2] S. V. Jain and R. N. Patel, Investigations on pump running inturbine mode: a review of the state-of-the-art, Renewable andSustainable Energy Reviews, vol. 30, pp. 841868, 2014.

[3] H. Nautiyal and A. Kumar, Reverse running pumps analytical,experimental and computational study: a review, Renewableand Sustainable Energy Reviews, vol. 14, no. 7, pp. 20592067,2010.

[4] T. Wang, F. Kong, Y. He, and S. Yang, Researching status ofcentrifugal pump as turbine, Journal of Drainage and IrrigationMachinery Engineering, vol. 31, no. 8, pp. 674680, 2013.

[5] S. Derakhshan and A. Nourbakhsh, Theoretical, numericaland experimental investigation of centrifugal pumps in reverseoperation, ExperimentalThermal and Fluid Science, vol. 32, no.8, pp. 16201627, 2008.

[6] P. Singh, Optimization of the internal hydraulic and of systemdesign in pumps as turbines with field implementation andevaluation [Ph.D. thesis], University of Karlsruhe, Karlsruhe,Germany, 2005.


[7] P. Singh and F. Nestmann, Internal hydraulic analysis ofimpeller rounding in centrifugal pumps as turbines, Experi-mentalThermal and Fluid Science, vol. 35, no. 1, pp. 121134, 2011.

[8] S. S. Yang, F. Y. Kong, F. Shao, and L. Xue, Numerical calcu-lation and experiment of hydraulic turbine, Journal of JiangsuUniversity (Natural Science Edition), vol. 33, no. 2, pp. 165169,2012.

[9] S. S. Yang, F. Y. Kong, and J. H. Fu, Numerical research oneffects of splitter blades to the influence of pump as turbine,International Journal of Rotating Machinery, vol. 2012, ArticleID 123093, 10 pages, 2012.

[10] S. S. Yang, F. Y. Kong, H. Chen, and X. Su, Effects of bladewrap angle influencing a pump as turbine, Journal of FluidsEngineering, vol. 134, no. 6, Article ID 061102, 8 pages, 2012.

[11] S. S. Yang, F. Y. Kong, and B. Chen, Research on volutedesign method of pump as turbine using CFD, InternationalAgricultural Engineering Journal, vol. 20, no. 3, pp. 2532, 2011.

[12] S. Derakhshan, B. Mohammadi, and A. Nourbakhsh, Thecomparison of incomplete sensitivities and Genetic algorithmsapplications in 3D radial turbomachinery blade optimization,Computers and Fluids, vol. 39, no. 10, pp. 20222029, 2010.

[13] S. Derakhshan, B. Mohammadi, and A. Nourbakhsh, Incom-plete sensitivities for 3D radial turbomachinery blade optimiza-tion, Computers and Fluids, vol. 37, no. 10, pp. 13541363, 2008.

[14] B. Ji, X. Luo, Y. Wu, X. Peng, and Y. Duan, Numerical analysisof unsteady cavitating turbulent flow and shedding horse-shoe vortex structure around a twisted hydrofoil, InternationalJournal of Multiphase Flow, vol. 51, pp. 3343, 2013.

[15] B. Ji, X. Luo, Y. Wu, X. Peng, and H. Xu, Partially-AveragedNavier-Stokes method with modified k- model for cavitatingflow around a marine propeller in a non-uniform wake,International Journal of Heat and Mass Transfer, vol. 55, no. 23-24, pp. 65826588, 2012.

[16] O. Fecarotta, A. Carravetta, and H. M. Ramos, CFD andcomparisons for a pump as turbine: mesh reliability andperformance concerns, International Journal of Energy andEnvironment, vol. 2, no. 1, pp. 3948, 2011.

[17] C. Santolaria Morros, J. M. Fernandez Oro, and K.M. ArguellesDaz, Numerical modelling and flow analysis of a centrifugalpump running as a turbine: unsteady flow structures and itseffects on the global performance, International Journal forNumerical Methods in Fluids, vol. 65, no. 5, pp. 542562, 2011.

[18] X. F. Guan,Modern PumpsTheory andDesign, China AerospacePress, 1st edition, 2011.

[19] Help Navigator, ANSYS CFX, Release 12. 0 CFX-Solver model-ing Guide.

[20] F. J.Wang,Computational Fluid Dynamics Analysis-CFDPrinci-ples and Application, Tsinghua University Press, Beijing, China,2004.

[21] R. Spence and J. Amaral-Teixeira, Investigation into pressurepulsations in a centrifugal pump using numerical methodssupported by industrial tests, Computers and Fluids, vol. 37, no.6, pp. 690704, 2008.

[22] S. S. Yang, F. Y. Kong, and B. Chen, Research on pump volutedesign method using CFD, International Journal of RotatingMachinery, vol. 2011, Article ID 137860, 7 pages, 2011.

[23] P. Singh and F. Nestmann, Experimental optimization of afree vortex propeller runner for micro hydro application,Experimental Thermal and Fluid Science, vol. 33, no. 6, pp. 9911002, 2009.

[24] S. L. Dixon, Fluid Mechanics andThermodynamics of Turboma-chinery, chapters 2, 6 and 9, Elsevier, Oxford, UK, 2005.

[25] C. Pfleiderer and H. Petermann, Stromngsmaschinen, chapter 1,Springer, 1964.

Submit your manuscripts athttp://www.hindawi.com

VLSI Design

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of

RotatingMachinery


Hindawi Publishing Corporation http://www.hindawi.com

Journal of

EngineeringVolume 2014


Shock and Vibration


Mechanical Engineering

Advances in


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of


Distributed Sensor Networks


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

SensorsJournal of


Modelling & Simulation in EngineeringHindawi Publishing Corporation http://www.hindawi.com Volume 2014


Active and Passive Electronic Components

Advances inOptoElectronics

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

RoboticsJournal of


Chemical EngineeringInternational Journal of


Control Scienceand Engineering

Journal of


Antennas andPropagation




Navigation and Observation


Documents

190530