30120140505001 2

International Journal of Mechanical Engineering and Technology (IJMET), ISSN 0976 – 6340(Print),

ISSN 0976 – 6359(Online), Volume 5, Issue 5, May (2014), pp. 01-12 © IAEME

1

EFFECT OF DIFFUSER LENGTH ON PERFORMANCE

CHARACTERISTICS OF ELBOW DRAFT TUBE WITH DIVIDING PIER

Rahul Bajaj1, Dr. Ruchi Khare

2, Dr. Vishnu prasad

3

M.Tech. Student1, Asst. Professor

2, Professor

3

Department of Civil Engineering

M.A. National Institute of Technology, Bhopal (M.P.)

ABSTRACT

The hydraulic turbines extract the energy of flowing water and converts into mechanical

energy. The reaction turbine has components namely casing, stay ring, guide vane, runner and draft

tube. Each component plays some role in performance of turbine. Out of above component casing,

stay ring and distributor guide the flow while in runner and draft tube energy transfer and conversion

takes place. In reaction turbine, significant part of input energy goes out of runner unutilized in form

of kinetic energy. Draft tube are provided at exit of runner to connect turbine and tail race providing

closed conduit flow of varying cross sectional area. The development in the design of turbines leads

to different shape of draft tube to recover as much energy as possible. The elbow draft tube is mostly

used with large reaction turbine. It is found that geometry of draft tube, flow space affects its

performance to large extent but due to limitation of experimental investigation, the mesh were

limited to few shapes. The growth on computational power has made it possible to investigate the

many alternative design of draft tube. In the present paper, numerical simulation has been done for a

large elbow draft tube with pier in diffuser. The length of diffuser has been changed to see its effect

on the performance i.e. head and efficiency of draft tube.

Keywords: Draft Tube, CFD, Diffuser, Pier, Kinetic Energy, Efficiency.

INTRODUCTION

Draft tube is an important component of hydraulic reaction turbine and improves the

performance of turbine by converting the kinetic energy entering in draft tube from runner, into

pressure energy. [1] In axial flow reaction turbine, the kinetic energy of water leaving the runner is

INTERNATIONAL JOURNAL OF MECHANICAL ENGINEERING

AND TECHNOLOGY (IJMET)

ISSN 0976 – 6340 (Print)

ISSN 0976 – 6359 (Online)

Volume 5, Issue 5, May (2014), pp. 01-12

© IAEME: www.iaeme.com/ijmet.asp Journal Impact Factor (2014): 7.5377 (Calculated by GISI) www.jifactor.com

IJMET

© I A E M E

International Journal of Mechanical Engineering and Technology (IJMET), ISSN 0976

ISSN 0976 – 6359(Online), Volume 5, Issue

up to 50 % of total input energy where

total energy.

In beginning of 19th

century, the hydrodynamic investigation was carried by F. Frankl and

A.V. Milovich on straight draft tube and

the shape of draft tube on the basis

investigations were carried out on straight diffuser between 1909 and

to design various shapes of draft tube

and small diameters owing to the diameter of runner D

large it is irrational to construct. To overcome these problems,

recover kinetic energy of axial and rotational

large capacity mixed flow hydraulic turbine, bell mouth

runner diameter leads increase in dimensions and weight

were overcome by elbow draft tube which is suit

when diffuser length exceeds 10 to 12 m, the dividing pier is provided

distribution and consequently improves the performance of draft tube and power characteristics of

the turbine.[2]

In hydraulic turbine, determination of optimum shape and dimension as well as prediction of

flow behavior is very difficult task. To overcome these problems the numerical simulation

become more informative tool.[3] At the best operating

minimum whirl and radial velocity

significant whirl velocity.

Since from more than a decade’s Computational Fluid Dynamics (CFD)

tool and widely used by researcher for predicting performance and flow behavior in a domain and

optimization of design of hydraulic turbine components

time consuming and to overcome this

minimized laboratory testing of models.[4]

GEOMETRY AND MESH GENERATION

The elbow draft tube has three parts namely cone, elbow and

of thickness b= 0.25 D1 are used to improve the performance of turbine.

of elbow draft tube are shown in fig. 1.

The modeling of elbow draft tube has been done in ICEM CFD for 3 length L/ D

Model of elbow draft tube for h/D1= 2.3 and L/

mm diameter at inlet and the analysis is done for L = 5 * D

length.

Fig. 1: Geometric parameter of

International Journal of Mechanical Engineering and Technology (IJMET), ISSN 0976

e 5, Issue 5, May (2014), pp. 01-12 © IAEME

2

0 % of total input energy whereas in case of mixed flow reaction turbine it


A.V. Milovich on straight draft tube and they determine the solution to the problem of determining

the shape of draft tube on the basis of theoretical investigation. Therefore large

traight diffuser between 1909 and 1929 and emphasized

draft tubes. The use of straight tubes was restricted to turbine of medium

the diameter of runner D1 increases, the length of the tube becomes so

large it is irrational to construct. To overcome these problems, bell mouth tubes were invented and

kinetic energy of axial and rotational component of flow. With the rapid deve

large capacity mixed flow hydraulic turbine, bell mouth draft tube was also absolute because large

dimensions and weight of draft tube. Later on all these problems

overcome by elbow draft tube which is suitable for large diameter hydraulic turbine.

length exceeds 10 to 12 m, the dividing pier is provided which give

consequently improves the performance of draft tube and power characteristics of

, determination of optimum shape and dimension as well as prediction of

flow behavior is very difficult task. To overcome these problems the numerical simulation

At the best operating condition the flow enters

in a draft tube but at off design condition, the flow enters with

Since from more than a decade’s Computational Fluid Dynamics (CFD)

widely used by researcher for predicting performance and flow behavior in a domain and

ion of design of hydraulic turbine components. Conventional model testing is costly and

to overcome this, CFD is an alternative tool which is a cost effective and

minimized laboratory testing of models.[4]

GEOMETRY AND MESH GENERATION

three parts namely cone, elbow and diffuser. In large draft tube piers

are used to improve the performance of turbine. The geometric dimensions

draft tube are shown in fig. 1.

The modeling of elbow draft tube has been done in ICEM CFD for 3 length L/ D

= 2.3 and L/ D1=15 is shown in fig.2. The draft tube is having 250

mm diameter at inlet and the analysis is done for L = 5 * D1, L = 10 * D1 and L = 15 * D

parameter of elbow draft tube with dividing pier [2]


is up to 15 % of


they determine the solution to the problem of determining

large numbers of

and emphasized the need

tubes was restricted to turbine of medium

, the length of the tube becomes so

bell mouth tubes were invented and

flow. With the rapid development of

draft tube was also absolute because large

Later on all these problems

able for large diameter hydraulic turbine. In case

gives the better flow

consequently improves the performance of draft tube and power characteristics of

, determination of optimum shape and dimension as well as prediction of

flow behavior is very difficult task. To overcome these problems the numerical simulation has

enters mostly axially

the flow enters with

Since from more than a decade’s Computational Fluid Dynamics (CFD) is most effective

widely used by researcher for predicting performance and flow behavior in a domain and

. Conventional model testing is costly and

n alternative tool which is a cost effective and

. In large draft tube piers

The geometric dimensions

The modeling of elbow draft tube has been done in ICEM CFD for 3 length L/ D1 ratio [2].

is shown in fig.2. The draft tube is having 250

and L = 15 * D1 diffuser

[2]



3

Fig. 2: Isometric view of elbow draft tube

In three dimensional CFD simulations, the fluid domain can be discretized by tetrahedral or

hexahedral mesh. Unstructured mesh which mainly consists of tetrahedral elements while structured

mesh consists of hexahedral elements.[5] Ansys CFX uses finite volume method (FVM) for the

discretized domain and N-s equations is solved at every node of the cell.

In the present work, meshing of draft tube is done in Ansys ICEM CFD as shown in fig. 3.

For this flow domain, unstructured three dimensional tetrahedral meshing has been adopted.

Fig.3: Mesh of elbow draft tube

BOUNDARY CONDITIONS

The numerical flow simulation is carried out for 13 whirl component varying from 3.0 m/s

to -3.0 m/s at an interval of 0.5 m/s at 3 different lengths of diffuser i.e. L = 5 * D1, L = 10 * D1

and L = 15 * D1. Cylindrical velocity component at inlet of draft tube is specified as axial

component = -15.41 m/s, radial component = 0 m/s and 13 whirl component at an interval of 0.5 m/s

as inlet boundary condition. Average static pressure is specified at outlet of draft tube as outlet

boundary condition. All the walls in the geometry is assumed as smooth and no slip. Shear Stress

Transport (SST) K-w turbulence model is used with wall function as automatic in Ansys CFX code.

Convergence criteria are set to be 10-6

as RMS value and 500 as maximum iteration.



4

COMPUTATION OF PARAMETARS

The following formulae are used to assess performance of draft tube using the velocity and

pressure distribution from numerical simulation.

Head loss in draft tube

03 05( )

LD

P PH

γ

−=

Head recovery in draft tube

2 2

03 05( )

2RD LD

V VH H

g

−= −

Draft tube efficiency

2

03

2100RD

D

gH

Vη = ×

Relative head loss

2

03

2100LD

RL D

gHH

V= ×

RESULTS AND DISCUSSIONS

The 3D flow analysis has been carried out in elbow draft tube with three L / D1 ratios of 5, 10

and 15 for 13 whirl component varying from 3.0 m/s to -3.0 m/s at an interval of 0.5 m/s for constant

n1’=135.6 rpm. Fig. 4 shows the various sections at which pressure and velocity contours are taken.

Fig. 4: Pressure contour for different section (L=5D1)



5

Fig. 5.(a): Pressure contour at elbow section for L= 5 D1 and theta component= -1 m/s

Fig. 5.(b): Pressure contour at elbow section for L= 10 D1 and theta component= -1 m/s

Fig. 5.(c): Pressure contour at elbow section for L= 15 D1 and theta component= -1 m/s



6

Fig. 6.(a): Velocity contour at elbow section for L= 5 D1 and theta component= -1 m/s

Fig. 6.(b): Velocity contour at elbow section for L= 10 D1 and theta component= -1 m/s

Fig. 6.(c): Velocity contour at elbow section for L= 15 D1 and theta component= -1 m/s



7

It is seen in figure 5.(a),(b) and (c) that maximum pressure variation is obtained at outer side

of elbow and in fig. 6. (a),(b) and (c) that at the periphery, velocity is zero and velocity is increases

as moves from outer side to inner side of elbow.

Fig. 7.(a): Pressure contour at inlet of diffuser for L= 5 D1 and theta component= -1 m/s

Fig. 7.(b): Pressure contour at inlet of diffuser for L= 10 D1 and theta component= -1 m/s

Fig. 7.(c): Pressure contour at inlet of diffuser for L= 15 D1 and theta component= -1 m/s

Fig. 8.(a): Velocity contour at inlet of diffuser for L= 5 D1 and theta component= -1 m/s

Fig. 8.(b): Velocity contour at inlet of diffuser for L= 10 D1 and theta component= -1 m/s



8

Fig. 8.(c): Velocity contour at inlet of diffuser for L= 15 D1 and theta component= -1 m/s

From fig 7.(a),(b) and (c), it is seen that the maximum pressure zone occurs at divider and in fig

8.(a),(b) and (c) minimum velocity zone is occuer at divider and periphery of diffuser.

Fig. 9.(a): Pressure contour at diffuser for L= 5 D1 and theta component= -1 m/s

Fig. 9.(b): Pressure contour at diffuser outlet for L= 10 D1 and theta component= -1 m/s

Fig. 9.(c): Pressure contour at diffuser outlet for L= 15 D1 and theta component= -1 m/s

Fig. 10.(a): Velocity contour at diffuser outlet for L= 5 D1 and theta component= -1 m/s



9

Fig. 10.(b): Velocity contour at diffuser outlet for L= 10 D1 and theta component= -1 m/s

Fig. 10.(c): Velocity contour at diffuser outlet for L= 15 D1 and theta component= -1 m/s

It is observed from fig 9.(a),(b) and (c) that pressure variation is nearly constant and in fig

10.(a),(b) and (c), velocity is minimum at upper side of diffuser due to sharp curvature at divider.

Fig. 11.(a): Pressure Contour at L= 5 D1 and theta component= -1 m/s

Fig. 11.(b): Pressure Contour at L= 10 D1 and theta component= -1 m/s



10

Fig. 11.(c): Pressure Contour at L= 15 D1 and theta component= -1 m/s

The pressure contour for a draft tube is shown in Fig.11. (a), (b) and (c) at L/ D1 = 5, 10 and

15 respectively. From these contours it can be seen that after achieving a optimum pressure at the

inlet of dividing pier, pressure is nearly constant throuhout the length of draft tube, there is sudden

drop in pressure due to curvature at elbow section, where as .

Fig. 12.(a): Velocity stream line at L= 5 D1 and theta component= -1 m/s

Fig. 12.(b): Velocity stream line at L= 10 D1 and theta component= -1 m/s



11

Fig. 12.(c): Velocity stream line at L=1 5 D1 and theta component= -1 m/s

The streamline patterns for a draft tube is shown in Fig.12. (a), (b) and (c) at L/ D1 = 5, 10

and 15 respectively. From the above figure it can be seen that the amount of whirl incresase with

increase in diffuser length. The whirl is higher in the left of the dividing pier as compared to whirl in

right of dividing pier , it may be because of eccentricity of the dividing pier with the centre line of

draft tube inlet. The velocity reduces from inlet to outlet which confirms the cherecterstic of draft

tube.

The draft tube efficiency and relative loss in each numerical simulation are computed and

graphical representation is shown in fig. 13 and fig. 14. As shown in fig.13 the maximum draft tube

efficiency obtained at 0 m/s whirl component in case of L/D1 = 10. All these numerical simulation

shows that when diffuser length is L > 10 D1, the highest efficiency is achieved. From fig. 14, it is

seen that, minimum relative loss is for L/D1 = 5 and 10.

Fig. 13: Efficiency of Draft Tube at Fig. 14: Relative Loss in Draft Tube at

different L / D1 ratios different L / D1 ratios

10

20

30

40

50

60

70

80

90

-4 -3 -2 -1 0 1 2 3 4

Dra

ft T

ub

e E

ffic

ien

cy (

% )

Theta Component (m/s)

L= 5 D1

L= 10 D1

L= 15 D1

0

10

20

30

40

50

60

70

80

-4 -3 -2 -1 0 1 2 3 4

Re

lati

ve

Lo

ss (

% )

Theta Component (m/s)

L= 5 D1

L= 10 D1

L= 15 D1

Poly. (L= 15

D1)



12

CONCLUSION

From numerical simulation of elbow draft tube with dividing pier it may be observed that that

maximum efficiency is achieved at L= 10 * D1 length of the draft tube. The comparative study of the

pressure variations and velocity contours at the inlet section of draft tube and just after the elbow

section shows that location of dividing pier effects the velocity distribution significantly. Efficiency

of draft tube is affected due to whirl component.

REFERENCES

1. Motycak Lukas, Skotak Ales, Obrovsky Jiri, 2010, “Conditions of Kaplan Turbine CFD

Analysis”, ANSYS Conference.

2. M.F Gubin, 1973, “Draft Tubes of Hydraulic Station”, Amerind Publishing Company Pvt.

Ltd.

3. Khare Ruchi, Prasad Vishnu, Mittal Sushil Kumar, 2012, “Effect of Runner Solidity on

Performance of Elbow Draft Tube”, Energy Procedia 14, pp 2054-2059.

4. Khare Ruchi, Prasad Vishnu, Kumar Sushil, 2010, “Derivation of Global Parametric

Performance of Mixed Flow Hydraulic Turbine Using CFD”, Hydro Nepal.

5. Choi Hyen-Jun, Zullah Mohammed Asid, Roh Hyoung-Woon, Ha Pil-Su, Oh Sueg-Young,

2013, “CFD Validation of Performance of a 500 kW Francis Turbine”, Renewable Energy 54,

pp 111-123.

6. Prasad Vishnu, Khare Ruchi, Abhas Chincholikar,2010, “Hydraulic Performance of Elbow

Draft Tube for Different Geometric Configurations Using CFD”, IGHEN-2010, Oct.21-23,

2010, AHEC, IIT Roorkee, India.

7. U. Andersson, J. Jungstedt and M.J. Cervantes, 2008, "Model Experiments of Dynamic Loads

on a Draft Tube Pier", 24th Symposium on Hydraulic Machinery and Systems.

8. S Tridon, S Barre, G D Ciocan, P Leroy and C Segoufin, 2010, "Experimental investigation of

draft tube flow instability", 25th IAHR Symposium on Hydraulic Machinery and Systems.

9. ANSYS, 2010, "ANSYS CFX 13 Software Manual, ANSYS, Inc., southpointe, Canonsburg,

PA.

10. P. K. Sinha, A.K.das and B. Majumdar,, “Numerical Investigation of Flow Through Annular

Curved Diffuser”, International Journal of Mechanical Engineering & Technology (IJMET),

Volume 2, Issue 2, 2011, pp. 1 - 13, ISSN Print: 0976 – 6340, ISSN Online: 0976 – 6359.

11. Prof. N.V.Hargude and Prof. P.R.Patil, “Design and Development of Plc Operated Hydraulic

System for Machining Process”, International Journal of Mechanical Engineering &

Technology (IJMET), Volume 3, Issue 2, 2012, pp. 366 - 373, ISSN Print: 0976 – 6340,

ISSN Online: 0976 – 6359.

12. Tarun Singh Tanwar , Dharmendra Hariyani and Manish Dadhich, “Flow Simulation (CFD) &

Static Structural Analysis (FEA) of a Radial Turbine”, International Journal of Mechanical

Engineering & Technology (IJMET), Volume 3, Issue 3, 2012, pp. 252 - 269, ISSN Print:

0976 – 6340, ISSN Online: 0976 – 6359.