See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/340297687

Analysis of tip leakage in external gear pump

Conference Paper · May 2018

CITATION READS

1 601

2 authors:

Deniz İmamoğlu Özgür Ertunç

Ozyegin University Ozyegin University
2 PUBLICATIONS 2 CITATIONS 74 PUBLICATIONS 306 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Ultrasound cavitation and applications View project

Investigating Rubber(EPDM) Viscous Flow View project

All content following this page was uploaded by Özgür Ertunç on 31 March 2020.

The user has requested enhancement of the downloaded file.

 9th International Automotive Technologies Congress, OTEKON 2018

ANALYSIS OF TIP LEAKAGE IN EXTERNAL GEAR PUMP

Deniz İMAMOĞLU, Özgür ERTUNÇ

Ozyegin University, Mechanical Engineering Department, İstanbul, Turkey

Corresponding Author: Deniz İmamoğlu, [email protected]

ABSTRACT
External gear pumps are one of the mostly used pump types in the fluid power applications due to simplicity, low
cost and long operation period. However, there are some disadvantages such as internal leakages, noise and
vibration with gear pumps. This study presents comparison of theoretical calculations and numerical simulations
with considering temperature effects on tip leakage. Numerical simulations have been done for pressure difference
from 2 bars up to 10 bars for seven different rotational speeds. Additionally, the effect of wall temperature varied
from 30 ºC up to 60 ºC has been investigated on the performance of gear pump. Because of extra assumptions
made in theoretical calculations, there are some deviations between theoretical and numerical results. These
deviations are due to pressure drop which occurs because of vorticities at volume packages between teeth in
numerical results and fully developed flow assumption made in theoretical analysis. Results of numerical study
show that the fully developed assumption at the clearance of gear tooth and wall is not appropriate and it leads to
large deviations between numerical and theoretical analysis. Furthermore, it has been shown that leakages increase
linearly with temperature rise at the wall.
Keywords: Gear pump, theoretical calculations, numerical simulations, tip leakages, temperature effect

1. INTRODUCTION
Gear pumps are used in many applications because of specific manufacturing simplicity, smooth and reliable
operation for a long time and relatively low production cost. However, gear pumps are prone to problems such as
leakages, vibration and noise (Ghionea, Ioan, & Tiriplic, 2012). Internal leakages affect volumetric efficiency and
pump operation negatively. Three type of leakages such as face leakage, tip leakage and leakage between two
gears are emphasized theoretically. Tip leakages are in safety levels when the clearance of gear tooth and the wall
is between 5-15 μm (Koç & Canbulut, 1985). In another study, slip flow parameter is introduced to calculate the
leakages based on pressure differential. The effect of slip flow coefficient on the amount of leakages for rotational
speeds ranging from 250 RPM to 3500 RPM are investigated and it has been shown that slip flow coefficient
reduces volumetric efficiency for constant rotational speed (Yusof, Wasbari, Zakaria, & Ibrahim, 2013). Schiffer
et al. (2013) have studied the effect of floating sealing element on tip and face leakages for a mini high-pressure
gear pump. In their research, three different designs of floating sealing elements have been tested and the best one
with highest volumetric efficiency and lowest leakage have been selected based on experimental results (Schiffer,
Benigni, & Jaberg, 2013). External gear pumps also can be used for low pressure applications such as fuel
injection systems and SCR systems. Optimal design of gear pumps for exhaust gas after-treatment applications
(SCR) are studied by Devendran & Vacca, (2013). Flow pulsations, pressure overshoots, cavitation and volumetric
efficiency were optimization parameters, which have been investigated by changing the teeth number, pressure
angle and addendum-dedendum coefficient (Devendran & Vacca, 2013).

Several numerical studies have been done to investigate the fluid flow structure inside gear pumps. Ghazanfarian
and Ghanbari (2014) have used k-ε turbulence model with second-order discretization scheme for convective
terms over a dynamically varying unstructured mesh for rotary double external gear pump. Based on their results,
reduction of clearance by 9 µm, increases net flow rate by 30%. Furthermore, region at the contacting teeth near
the suction side is more prone to cavitation (Ghazanfarian & Ghanbari, 2014). Houzeaux and Codina (2007) have
simulated whole gear pump and half gear pump, which includes suction chamber and the teeth at inlet, as 2D and
3D cases. Comparison of half pump and whole pump shows that there is not much difference for pressure contours
and velocity profiles. Additionally, 3D simulations give more realistic results than 2D simulations (Houzeaux &
Codina, 2007). Castilla et al. (2010) have investigated the flow in the suction chamber of an external gear pump.
They have compared laminar flow assumption with different turbulence models such as standard k-ε, realizable k-

621
 9th International Automotive Technologies Congress, OTEKON 2018

ε, RNG k-ε and Reynolds Stress Models (RSM). Based on their results, RSM turbulence model is the most
suitable one according to the experimental data (Castilla et al., 2010). In another work by Castilla et al. (2015) half
of the gear pump using 2D and 3D simulations have been analyzied with different contact ratios. Their results
show that flow rate obtained from 3D simulations are higher than that from 2D simulations with lower fluctuations
in velocity magnitude both at inlet and outlet sections (Castilla et al., 2015). In other research, CFD analysis of
gear pump using with deforming mesh were studied. The simulation was modeled as unsteady, laminar and
Eulerian multiphase flow. Especially, the importance of mesh deforming in the gearing zone is indicated by
authors. The mesh size at deforming zone were retained similar size with boundary cells and growth factor of
mesh was minimized to improve accuracy of results. Furthermore, time step was adjusted as 0.0003 revolution per
time step. External gear pump simulations have moving boundaries such as driving and driven gears. Moving
boundaries makes the problem unsteady. Based on their results, coefficient of spatial variation decrease from 45%
to 5.5% by increasing additive injection stream numbers at the pump intake (Strasser, 2007).

Volumetric efficiency of gear pumps is directly influenced by the amount of leakages. Hence, leakages should be
avoided while designing external gear pumps. To decrease the leakages, lateral and tip clearances of gear pump
should be minimized. Additionally, leakages can be reduced by increasing rotational speed (Koç & Canbulut,
1985).

Theoretical leakage and volume displacement analysis has to be verified numerical simulations. Numerical
simulations help to examine theoretical calculation and limits of the theory. Theoretical leakage and volume
displacement calculations are mostly performed with assumptions such as fully developed flow in literature.
Validity range of the theoretical calculations can be examined in 2D numerical simulations by studying the flow
structures and flow characteristics. Furthermore, temperature effect which can be occurred due to friction on the
pump body decrease dynamic viscosity of fluid and increase tip leakages.

In this study, the gear pump geometry includes only the region which is between gear teeth and the pump body for
one gear and two reservoirs. Figure 1 and 2 shows the geometry in detail. In addition, transient mass flow rate at
inlet and outlet, leakages on tips, which are between case and gear teeth, are investigated on 2D numerical
simulations in this study. Theoretical calculations are compared with numerical simulations to observe which one
is more accurate.

2. NUMERICAL SIMULATION
In this study, numerical simulations were performed on Star-CCM+. STAR-CCM+ can simulate internal and
external fluid flow across a wide range of flow regimes, and for a variety of fluid types. It solves the conservation
equations for mass, momentum, and energy for general incompressible and compressible fluid flows.

Numerical simulation of external gear pump is too hard and time-consuming task due to close gaps. These gaps
cause low mesh size, high number of mesh cells and too long computational time. To decrease number of mesh
cells and computational time, the geometry of gear pump is simplified for numerical simulations in this study. The
geometry conversion is shown on Figure 1.

622
 9th International Automotive Technologies Congress, OTEKON 2018

Figure 1. Geometry conversion and geometry scene of fluid domain

Figure 2. Geometric dimensions of fluid domain

The geometry scene of flow domain including two reservoirs and 10 teeth in contact with the wall is shown in
Figure 1. This simplified geometry has been created based on radius of gear. The dimensions of flow domain are
given on Table 1. In fact, each gear has 14 teeth; but 10 teeth are investigated in this study. Because tip leakages
are observed only in the region which is in contact with wall and includes 10 teeth.

Table 1. Flow Domain dimensions

Symbol Description Unit Dimension

h Film Thickness µm 36.3
b Face width mm 10
L Top Land mm 0.371
rd Radius of Gear mm 8.69

The flow is modeled as incompressible and laminar. Water is used as fluid. Properties of water are shown on Table
2. Second order implicit unsteady solver has been used for numerical simulations. Time step has been adjusted 0.1
μs. In addition, inlet and outlet boundaries have been modeled pressure inlet and pressure outlet. The remaining
boundaries have been modeled wall. Reference pressure and inlet pressure are given as 1 bar and 0 bar
respectively. Outlet pressure is given 2, 5, 8, 10 bar as gauge pressure. In addition, the translation velocity has

623
 9th International Automotive Technologies Congress, OTEKON 2018

been calculated through rotation speed and radius of gear. Rotation speeds are 400, 600, 800, 1000, 1200, 1400,
1600 rpm.

Table 2. Properties of water

Symbol Description Unit Value

ρ Density kg/m3 997.561
μ Dynamic Pa.s 8.8871 x
Viscosity 10-4

There are two regions as moving body and fixed body. Moving body includes gear teeth and volume packages.
Moving frame to x+ direction is defined to moving body by using translation reference frame. Besides, fixed body
contains two reservoirs and the surface between gear teeth and wall.

Part based meshing has been used for mesh generation. An unstructured 2D polyhedral mesh grid has been
constructed for moving region and fixed region. Also, prism layer mesh has been used to observe more accurate
results at zone close to wall, decrease the number of cells and computational time. Total number of cells is 106163
from which 38217 belongs to moving body and 67946 to fixed body. The mesh scenes of all fluid domain and
between teeth tip and wall are shown in Figure 3 and Figure 4 respectively.

Figure 3. Mesh scene of all fluid domain

Figure 4. Mesh scene between teeth and wall

Theoretical calculations have been done based on Koç an Canbulut (1985) ‘s equations for tip leakages. The
equation of tip leakage volume flow rate is as follows:

(1)

to calculate tip leakages per tooth where b is face width, h is the film thickness, ∆Pt is the pressure difference, rd is
radius of gear, L is top land, ƞ is dynamic viscosity, ω is rotational speed and n is number of teeth which is in
contact with wall. L, b, h, ƞ, rd, and n are constant. Pressure difference and rotational speed have been changed and
observed how to effect on tip leakages. The first term represents the amount of fluid which leaks to inlet due to

624
 9th International Automotive Technologies Congress, OTEKON 2018

pressure difference; the second term represents the amount of fluid which is moved to outlet due to viscous
adhesion to the tooth tip.

In this study, temperature effect on tip leakages have been investigated. In numerical simulations, dynamic
viscosity property of water was used as a polynomial function in T between 5-95 ºC as below:

(2)

3. RESULTS

In this study, tip leakages have been analyzed for various rotational speeds at 2, 5, 8, 10 bars outlet pressure.
Rotational speeds given in numerical simulations are 400 rpm, 600 rpm, 800 rpm, 1000 rpm, 1200 rpm, 1400 rpm
and 1600 rpm. In addition to numerical simulations, theoretical tip leakage calculations have been done and
compared with simulations. Furthermore, numerical simulations have been performed different wall temperatures
to investigate temperature effect on tip leakages in external gear pumps. Temperatures have been given 30ºC, 40
ºC, 50ºC, 60 ºC at the wall which is in contact with gear tips respectively.

Firstly, tip leakages try to return previous tooth, but the back volume package pushes back leakage fluid. Then, it
starts to rotate and vorticity forms between two teeth. In parallel with this, velocity vector scenes of gear pump are
shown in Figure 5 for 2, 5, 8, 10 bars in case 800 rpm. These figures shows that vorticities occur and intensify
volume packages between gear teeth due to pressure changes. In Figure 5 (a), vorticity occurs near the wall at 2
bar. Besides, another vorticity starts to form between teeth. In Figure 5 (b), vorticity near the wall intensifies and
the other vorticity is about to occur at 5 bar. In Figure 5 (c) and Figure 5 (d), vorticity near the wall intensifies
more and more and other one occurs and intensifies too. These vorticities cause pressure drop at clearances and the
velocities decrease between gear tips and wall. In theoretical calculations, pressures at clearances increase linearly
and equally. Therefore, a deviation has been occurred between numerical and theoretical results. In Figure 6,
pressure distribution is shown for 10 bar, 1000 rpm case. Pressure increases linearly but not equally in numerical
simulations.

625
 9th International Automotive Technologies Congress, OTEKON 2018

Figure 5. Velocity vector scene for 800 rpm, (a) 2 bar, (b) 5 bar, (c) 8 bar, (d) 10 bar

Figure 6. Absolute pressure distribution in gear pump (10 bar, 1000 rpm)

Koç and Canbulut (1985) have done theoretical calculations for tip leakages in external gear pumps. In this study,
these calculations have been used to calculate tip leakages for the geometry which is shown in Figure 1. Tip
leakages have been calculated for all cases which have been done numerical simulations. The comparison of
numerical simulations and theoretical calculations are shown in Figure 7. Tip leakages do not change much with
different rotational speeds, but it increases as pressure raises. Actually, tip leakages decrease when rotational speed

626
 9th International Automotive Technologies Congress, OTEKON 2018

is increased. However, this reduce is approximately 10%. As it is seen in Figure 7, there is a difference between
theoretical calculations and numerical simulations. Theoretical calculations at 2 bar are two times higher than
numerical simulations and this difference gets higher and higher as pressure increases. Mass flow rate of leakage
raises by three times when pressure increases from 2 bar to 10 bar in numerical simulations. Besides that, it rises
by five times for this situation in theoretical calculations.

Figure 7. Comparison of numerical simulations and theoretical calculations for the tip leakage

Theoretical calculations of tip leakages have been done based on fully developed flow assumption. However, this
assumption cannot be seen between gear tips and wall in numerical simulations. This difference arises from fully
developed assumption. To observe whether fully developed flow exists, 19 plane sections have been created in this
area at the middle tooth of gear pump. Velocity profiles have been created for each plane sections, but there is not
any velocity profile which demonstrate fully developed flow. In Figure 8, plane sections’ locations are shown on
the graph and 5 out of 19 sections and theoretical velocity profiles are demonstrated. In numerical results, the
velocity profiles of the plane sections on the right do not match the fully developed flow. But, plane sections on
the left side of the tooth, it approaches the fully developed flow state but cannot be observed clearly. However,
theoretical velocity profile shows that the fully developed assumption has been existed.

627
 9th International Automotive Technologies Congress, OTEKON 2018

Figure 8. Comparison of velocity profile of theoretical and some plane sections between gear tip and wall for 2bar, 600rpm
case (fully developed flow and pressure effects)

Besides effect of fully developed assumption, pressure drop at clearances due to vorticity in volume packages
increases the difference between numerical and theoretical results., For instance, the differential pressure should be
raised 0.2 bar after each tooth for 2 bar cases theoretically. However, pressure difference is 0.13 bar in numerical
simulations at the tooth tip which is shown velocity profiles in Figure 8. To observe only fully development flow
assumption effect on tip leakages, the velocity profiles and mass flow rate of leakage has been calculated again at
0.13 bar as theoretical and they are shown in Figure 9 and Table 3. Theoretical and numerical in theory
calculations of tip leakages has a high difference.

Figure 9. Comparison of velocity profile of numerical value in theory and some plane sections between gear tip and wall for 2
bar, 600 rpm case (only fully developed flow effect)

Table 3. Comparison of theoretical, theoretical with numerical pressure difference and numerical for 2 bar and 600 rpm at the
middle tooth

Mass flow rate of Mass flow rate of Mass flow rate of

leakage (Theoretical) leakage (Numerical leakage
in theory) (Numerical)
0.0046 kg/s 0.0029 kg/s 0.0021 kg/s

628
 9th International Automotive Technologies Congress, OTEKON 2018

The difference arising from pressure drop is approximately 57% and the difference based on fully developed flow
is 40%.

Temperature rise on the wall due to friction increases tip leakage. Because, enhancement of temperature causes
decreasing dynamic viscosity of the fluid. Therefore, numerical simulations have been done for various wall
temperatures at the same pressure and rotational speed to observe temperature effect on tip leakage. Especially
between gear tips and wall, dynamic viscosity reduces dramatically because of high wall temperature as shown in
Figure 10. In Figure 11, temperature effect on tip leakages is shown. The graph shows that each 10ºC temperature
rise causes a 10% increase of tip leakage.

Figure 10. The change of dynamic viscosity in gear pump at T=60ºC

Figure 11. Temperature effect on tip leakages

4. CONCLUSIONS
In this study, comparison of theoretical and numerical calculations on tip leakages were presented. Numerical
simulations have been performed for four pressure difference at 7 different rotational speeds. Theoretical
calculations have been done to validate numerical simulations. The effect of heat transfer from heated casing to the
fluid on tip leakage were also presented.

There is some deviation between theoretical calculations and numerical simulations. A part of this deviation comes
from pressure drop because of vorticity existence in volume packages. The other part of deviation is based on fully
developed flow assumption in theoretical calculations. To check this assumption in numerical simulations, 19

629
 9th International Automotive Technologies Congress, OTEKON 2018

plane sections have been created and observed each planes’ velocity profiles. However, fully developed flow
assumption cannot be observed on numerical simulations. Additionally, vorticity occurrence in volume packages
causes to pressure drop at clearances. In theory, pressure increases linearly and equally after each tooth in external
gear pumps. However, numerical simulations show that pressure rises linearly but not equally in this study. This
creates a high difference between numerical and theoretical results as much as 57%. Besides, fully developed flow
assumption causes another 40% difference.

Furthermore, temperature effect on tip leakages has been investigated in this study. As the wall temperature
increase due to friction between gear tips and wall, dynamic viscosity decreases. This means that amount of fluid
which is forwarded to outlet by viscosity decreases; so tip leakages increases.

As future work, 3D numerical simulations of all gear pump body will be performed. Beside tip leakages, face
leakages will be investigated and compared with theoretical calculations.

ACKNOWLEDGEMENTS

The authors would like to gratefully acknowledge Scientific and Technological Research Council of Turkey
(TÜBİTAK) for providing financial support to this research with the 115M093 project.

REFERENCES

[1] Castilla, R., Gamez-Montero, P. J., del Campo, D., Raush, G., Garcia-Vilchez, M., & Codina, E. (2015).
Three-Dimensional Numerical Simulation of an External Gear Pump With Decompression Slot and Meshing
Contact Point. Journal of Fluids Engineering, 137(April), 41105. https://doi.org/10.1115/1.4029223
[2] Castilla, R., Gamez-Montero, P. J., Ertrk, N., Vernet, A., Coussirat, M., & Codina, E. (2010). Numerical
simulation of turbulent flow in the suction chamber of a gearpump using deforming mesh and mesh
replacement. International Journal of Mechanical Sciences, 52(10), 1334–1342.
https://doi.org/10.1016/j.ijmecsci.2010.06.009
[3] Devendran, R. S., & Vacca, A. (2013). Optimal design of gear pumps for exhaust gas aftertreatment
applications. Simulation Modelling Practice and Theory, 38, 1–19.
https://doi.org/10.1016/j.simpat.2013.06.006
[4] Ghazanfarian, J., & Ghanbari, D. (2014). Computational Fluid Dynamics Investigation of Turbulent Flow
Inside a Rotary Double External Gear Pump. Journal of Fluids Engineering, 137(2), 21101.
https://doi.org/10.1115/1.4028186
[5] Ghionea, G., Ioan, C., & Tiriplic, P. (2012). Simulation of the Working Conditions for a Gear Pump Using
Finite Element Analysis Method, 2012(Xxvi), 21–28.
[6] Houzeaux, G., & Codina, R. (2007). A finite element method for the solution of rotary pumps. Computers &
Fluids, 36(4), 667–679. https://doi.org/10.1016/j.compfluid.2006.02.005
[7] Koç, E., & Canbulut, F. (1985). Pozitif Deplasmanlı Pompa ve Motorlarda İç Akışkan Kaçağı. Kayseri:
Erciyes Üni, Fen Bilimleri Dergisi.
[8] Schiffer, J., Benigni, H., & Jaberg, H. (2013). Development of a novel miniature high-pressure fuel pump
with a low specific speed. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of
Automobile Engineering, 227(7), 997–1006. https://doi.org/10.1177/0954407013476820
[9] Strasser, W. (2007). CFD Investigation of Gear Pump Mixing Using Deforming/Agglomerating Mesh.
Journal of Fluids Engineering, 129(4), 476. https://doi.org/10.1115/1.2436577
[10] Yusof, A. A., Wasbari, F., Zakaria, M. S., & Ibrahim, M. Q. (2013). Slip flow coefficient analysis in water
hydraulics gear pump for environmental friendly application. IOP Conference Series: Materials Science and
Engineering, 50(1), 12016. https://doi.org/10.1088/1757-899X/50/1/012016

630

View publication stats