NOVATEUR PUBLICATIONS

INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]

ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

FLOW DISTRIBUTION NETWORK ANALYSIS FOR DISCHARGE SIDE OF

CENTRIFUGAL PUMP
Satish M. Rajmane
Research Scholar,
WIT Research Center, Solapur University, Solapur, India

Dr. S. P. Kallurkar
Principal,
Atharva College of Engineering, Mumbai, India

ABSTRACT
A computational fluid dynamics (CFD) analysis has been conducted to find the pressure losses for dividing
and combining fluid flow through a junction of discharge system. Simulations are performed for a range of
flow ratios and equations are developed for pressure loss coefficients at junctions. A mathematical model
based on successive approximations then would be employed to estimate the pressure losses. The proposed
CFD based strategy can be used for the analysis of all the three pipe branches of some diameter are selected
along with equal length so that only the effect of bend angle can be studied. The effect of bend angle, pipe
diameter, pipe length, Reynolds number on the resistance coefficient is studied. The software used is
CATIA for modeling and ANSYS fluent for analysis purpose.

INTRODUCTION
Centrifugal pumps are a sub-class of dynamic axis symmetric work absorbing turbo machinery. Centrifugal
pumps are used to transport fluids by the conversion of rotational kinetic energy to the hydrodynamic
energy of the fluid flow. The rotational energy typically comes from an engine or electric motor. The fluid
enters the pump impeller along or near to the rotating axis and is accelerated by the impeller, flowing
radially outward into a diffuser or volute chamber (casing), from where it exits. Common uses include
water, sewage, petroleum and petrochemical pumping. The reverse function of the centrifugal pump is
a water turbine converting potential energy of water pressure into mechanical rotational energy.

A computational fluid dynamics (CFD) analysis has been conducted to find the pressure losses for dividing
and combining fluid flow through a junction of discharge system. Simulations are performed for a range of
flow ratios and equations are developed for pressure loss coefficients at junctions. A mathematical model
based on successive approximations then would be employed to estimate the pressure losses. The proposed
CFD based strategy can be used as a substitute to setting up and performing costly experiments for
estimating junction losses.

In this section a review of research work in the area of different inserts was carried out and based on this
review certain observation were made;

Abdul Waheed Badar, Reiner Buchholz, Yongsheng Lou and Felix Ziegler [1] have studied the A CFD
analysis has been conducted to find the pressure losses for dividing and combining fluid flow through a tee
junction of a solar collector manifold. Simulations are performed for a range of flow ratios and Reynolds
numbers, and equations are developed for pressure loss coefficients at junctions. A theoretical model based
on successive approximations then is employed to estimate the isothermal and non-isothermal flow

1|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

distribution in laminar range through a collector consisting of 60 vacuum tubes connected in parallel in a
reverse and parallel flow arrangement.

Chris Smith [2] have investigated using CFD in flow assurance Analysis of low material temperatures
caused by rapid gas expansion (JT cooling) is possible using CFD. This includes cold spots and thermal
gradients, which can go beyond the design limits of the material, and traditional pipeline tools dont give the
distinction required. Situations can be evaluated including choke valves and downstream pipe work at start-
up, and process equipment and pipe work during blow down.

R. Vijayakumar, S.N. Singh, V. Seshadri[3] has studied Unbalanced coal/air flow in the pipe system
feeding the boilers will lead to non uniform combustion in the furnace thereby lowering the efficiency of the
power plant.. The current practice of balancing the flow by introducing orifices in the pipe system is
generally achieved using semi-empirical methods. This paper presents the CFD analysis for deciding the
optimum geometry of the orifices to balance the flow in the existing power plant at NTPC.

Balvinder Singh, Harpreet Singh, Satbir Singh Sehgal[4] the objective of this study was to Pipe fittings
are used in the plumbing systems to connect straight pipe or tubing sections for regulating or measuring
fluid flow. The wye shape fitting will convert into T shape fitting. In the present work, effect of angle of
turn/bend for a Y-shape pipe will be studied computationally. Water and air as a fluid is selected which
flows through the plumbing system. The effect of bend angle, pipe diameter, pipe length, Reynolds number
on the resistance coefficient is studied. It was observed that resistance coefficient vary with the change in
flow parameters.

S. Mokhtari, V.V. Kudriavtsev, M. Danna [5] has studied uniformly of a dielectric film deposited on the
silicon wafers greatly depends on the flow delivery systems &injector geometry. In this paper, we analyze
flow distribution patterns through the simulated closed-end multiple outlet pipe/channel which delivers
chemical precursors to the parallel slot linear injector of a chemical vapor deposition reactor. Instead of
using semiemprical formulas for the flow resistance & flow network methodology, a finite difference multi-
block CFD solution method is used to calculate local pressure patterns in the pipe & establish flow
distribution trends & uniformly as a function of Re number & system aspect ratio. It was found that flow
distribution changes qualitatively as we move from the low Re flow conditions to intermediate & high Re
flow regimes.

2|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

DRAWING AND ANALYTICAL CALCULATION

A) DRAWING:-
Variant No. 1:-

Fig. 1 Variant No. 1

Variant No. 2:-

Design for the variant with angel of 20 degree:-

Fig. 2 Variant No. 2

3|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Variant No. 3:-

Design for the variant with angle 45 degree:-

Fig. 3 Variant No. 3

B) ANALYTICAL CALCULATIONS:-

Single Pipe:-

Given Data:-Inner diameter of pipe (di) = 50 mm

Thickness of pipe (t) = 1.7 mm

Length of pipe (l) = 1m

Mass of flowing water per second = 40 kg/sec

Density of flowing water (w) = 1000 kg/m3

Weight Density of carbon steel (cs) = 7850 kg/m3

4|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Coefficient of friction for pipe () = 0.005

Efficiency of motor () = 80%

a) Cross-sectional area of pipe:

b) Velocity of flowing liquid in pipe per second:

c) Major losses in pipe :

d) Pressure developed in a pipe

8.46

e) Weight of single pipe:

5|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

f) Power required to pump water:

FINITE ELEMENT ANALYSIS

We briefly examine the function of each of these elements within the context of a CFD code. Before
going for those processes we have to identify the problem; in that first includes define area means exactly
for which result we looking for (i.e. pressure drop, mass flow rate, temperature drop etc.). Second things
come that is identification of domain in which exact part of complete system have to identify by which
expected result (i.e. pressure drop, mass flow rate, temperature drop etc.) will affect and try to examine that
part only instead of complete system because it takes long time for simulation and required large memory
space in computer system unnecessary.

For solving we are using ANSYS Fluent Solver. In this interface following parameters used:

System type: Pressure Based system

Analysis type: Steady state condition
Model: k-epsilon (2 equations) with realizable model
Fluid Used: Water
Mass flow rate at inlet: 2 kg/s
Temperature: 27 deg0C.

6|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Pressure Plot:-

Fig. 4 ANSYS Variant 1 Fig. 5 ANSYS Variant 2

Fig shows that pressure plot in Pascal. At inlet is more and decreases as gradually along the length. Color
strip shows indicate the different pressure level. Blue color indicates the minimum pressure level and red
color indicate the max pressure level.

Fig. 6 ANSYS Variant 3

7|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Velocity Plot:-

Velocity plot shown in fig. color strip shows the different velocity levels in geometry. At wall o pipe
velocity is minimum and at the centre of pipe velocity is maximum Velocity vector shows the flow pattern
of fluid. Vortices can be seen using this plot.

Fig. 7 ANSYS Variant 1 Fig. 8 ANSYS Variant 1

Fig. 9 ANSYS Variant 2 Fig. 10 ANSYS Variant 2

8|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Fig. 11 ANSYS Variant 2 Fig. 12 ANSYS Variant 3

Fig. 13 ANSYS Variant 3

RESULT AND DISCUSSION

Discussion: - In this chapter we will discuss how to design the delivery side. The efficiency of the
centrifugal pump can be increased by number of ways such as modifying the geometry of the sump,
increasing the diameter of the discharge pump, etc. The following diagram shows a delivery pipe variant 1.

9|Page
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Fig. 14 Delivery pipe Variant 1 Fig. 15 Delivery pipe Variant 1

In the similar way we can create all the models in CATIA for analysis. The following Figure shows a model
of variants shape of manifold.

Fig. 16 Delivery pipe Variant 2 Fig. 17 Delivery pipe Variant 3

Fig. 18 Variant 1 Fig. 19 Variant 2

10 | P a g e
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

Fig. 20 Variant 3 Chart 1 Gambit operation

Result:-

The comparison between benchmark and modified model of delivery pipe proposed in our thesis can be
very well represented in a tabular format, as shown

Pressure developed
Sr. No Details Remark
(Analytical) in MPa
Variant 1 i.e.
1 Benchmark 3.161 More vortices developed
Geometry
Less vortices developed as compared to
2 Variant 3 3.125
benchmark geometry

Very less vortices developed as

3 Variant 2 1.12
compared to all variants

Table 1 Comparison between benchmark and modified model

Comparing the pressure drop in different geometry, Pressure developed in benchmark geometry is more as
compared to other variant. When we apply the angle for delivery pipe, fewer vortexes generated in 20 deg
bend pipe i.e. variant 2. Smooth fluid flow observed in variant 2. Hence less power required for variant 2.

CONCLUSION
Vortices and cavitations introduce inefficiency on the operation of the centrifugal pump.
Changing the bend angle of delivery pipe reduces the vortices and cavitations in pipe.
The suction head and the delivery head has a bearing on the output of the pump in terms of
discharge achieved per KW of pump power.
The intake pumping stations needs a uniform flow distribution of the sumps in order to ensure
the operation of pump units.
CFD model used to study the effect of various parameters which reduces time as well as cost and
hence could become an important tool for optimization of pump sump geometry.
Redesign of the delivery side of the pump facilitated the flow of water and improves the
discharge and consequently the performance of the centrifugal pump.
11 | P a g e
 NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 7, JULY-2015

REFERENCES
[1] Abdul Waheed Badar, Reiner Buchholz, Yongsheng Lou and Felix Ziegler; CFD based analysis of flow
distribution in a coaxial vacuum tube solar collector with laminar flow conditions; International Journal of
Energy and Environmental Engineering 2012.

[2] Chris Smith (Production Assurance Manager at Xodus Group); A dynamic approach to flow assurance
modelling; Technical Paper, Xodus Group.

[3] R. Vijayakumar, S.N. Singh, V. Seshadri, (Deptt. of Applied Mechanics, IIT Delhi) CFD; Analysis for
the Design of Optimum Geometry of the Orifices in the PC pipes for ensuring uniform feed rate of Coal
through various Coal Burners in a Boiler.

[4] Balvinder Singh, Harpreet Singh, Satbir Singh Sehgal; CFD Analysis of Fluid Flow Parameters withinA
Y-Shaped Branched Pipe; International Journal of Latest Trends in Engineering and Technology (IJLTET).

[5] S. Mokhtari, V.V. Kudriavtsev, M. Danna; Flow informality and pressure variation in Multi-outlet flow
distribution pipes, Advance in analytical, Experimental and computational Technologiesin Fluids,
Structures, Transient sand Natural Hazards; PVP-Vol.355, AME 1997

AUTHORS BIOGRAPHY:

Mr. S. M. Rajmane

He is having ME from Solapur University. He is Research Scholar for WIT Research Center, Solapur University, and
Solapur. He is having more than 10 years teaching experience His area of interest is in Fluid machinery, FEM,
Design engineering.

Dr. S. P. Kallurkar

He is having Phd from NITIE Mumbai. Presently working as Principal in Atharva College of Engineering, Mumbai.
He is Research Guide at WIT Research Center, Solapur University, and Solapur. He is having more than 25 years
teaching experience.

12 | P a g e