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

net/publication/320022019

Numerical Modeling and Validation of Oil Distribution and Churning Losses in

Gearboxes

Conference Paper · September 2017

CITATION READS

1 570

4 authors:

Hua Liu Thomas Jurkschat

Technische Universität München Technische Universität München
8 PUBLICATIONS   107 CITATIONS    10 PUBLICATIONS   84 CITATIONS   

SEE PROFILE SEE PROFILE

Thomas Lohner Karsten Stahl

Technische Universität München Technische Universität München
81 PUBLICATIONS   376 CITATIONS    397 PUBLICATIONS   1,050 CITATIONS   

SEE PROFILE SEE PROFILE

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

Ressource-efficient Electromechanical Drive Technology View project

Late Tooth Root Breakage due to non-metallic inclusions View project

All content following this page was uploaded by Hua Liu on 25 September 2017.

The user has requested enhancement of the downloaded file.


Numerical Modeling and Validation of Oil
Distribution and Churning Losses in Gearboxes
Hua Liu1) *, Thomas Jurkschat1), Thomas Lohner 1) and Karsten Stahl1)
1
Gear Research Centre (FZG), Technical University of Munich, Boltzmannstraße 15,
85748 Garching b. München, Germany
*Corresponding author: liu@fzg.mw.tum.de
1. Introduction
Three main concerns are in the focus of the devel-
opment of geared transmissions nowadays. These are
load carrying capacity, efficiency and
noise-vibration-harshness (NVH) behavior. Increasing
demand for reduced energy consumption and resource
efficiency leads to the requirement of modern transmis-
sions with higher efficiency. Losses always occur when
machine elements are moving relative to each other and
when oil molecules interact with each other. Besides the
load dependent losses, the no-load losses often play a
significant role especially in high-speed and
dip-lubricated transmissions with a high oil fill level.
Nowadays, many geared transmissions are designed
with a high oil fill level to overcome different driving
scenarios. In many cases, this leads to unnecessarily
high churning losses. Therefore, detailed knowledge on
the oil distribution inside a transmission is extremely
important to avoid insufficient oil supply or churning
losses inside gearboxes.
CFD (Computational Fluid Dynamics) methods of-
fer a very flexible way to gain a deeper insight into the
oil flow inside gearboxes with almost no restrictions on
the geometry and operating conditions. In this work, a
finite volume CFD model building on Liu et al. [10] is
applied to investigate the oil distribution and the churn-
ing losses of a single-stage gearbox at various operating
conditions. The influence of different oil viscosities and
circumferential speeds is considered and validated by
high-speed camera recordings as well as loss torque
measurements. The simulation results show very good
agreement with measurements. Even detailed behaviors
of the oil flow can be identified by the CFD simulation.
Hence, CFD is a very reliable method to expand
knowledge of the lubricant inside geared transmissions.
2. State of the Art
In view of increasing computing power in recent
years, CFD methods have gradually been applied to
geared transmissions to investigate the oil flow behavior
and no-load losses of geared transmissions.
Kvist [8] was one of the first to investigate the oil
distribution of a single dip-lubricated rotating gear using
a CFD model based on the Finite Volume Method
(FVM). The simulated oil distribution correlates well
with photographs of the oil distribution taken at constant
rotational speeds. Based on the simulation results, Kvist
[8] was able to derive a special guide plate to permit a
reduction of the oil fill level. As a result, the hydraulic
power losses can theoretically be reduced by 30%.
By using a two-dimensional finite volume based
model, Li et al. [9] analyzed the influence of rotational
speed, immersion depth and oil viscosity on the oil dis-
World Tribology Congress 2017
Beijing, China, September 17 – 22, 2017
tribution of a single-stage dip-lubricated gearbox. They
asserted that the turbulence of the oil flow increases
with oil viscosity as well as the rotational speed and the
oil fill level. However, a three-dimensional CFD simu-
lation model is required in order to gain detailed infor-
mation on the oil flow behavior inside gearboxes.
Jafarian [6] set up a three-dimensional finite volume
based model with an overset meshing technique to in-
vestigate the oil distribution of a single stage gearbox.
The velocity field of the flow pattern was compared
between CFD simulation results and Particle Image Ve-
locimetry (PIV) measurement results. A good level of
agreement was found.
Klier and Berger [7] presented a work modeling a
gear pair with an overset mesh method capable of rep-
resenting the meshing of two gears. The model gives a
plausible prediction of the oil flow inside a single stage,
dip-lubricated gearbox. However, the simulation results
were not validated due to lack of experimental data.
Concli et al. [1, 2, 3, 4] inducted several studies on
the CFD simulation of oil distribution and churning
power losses of spur gears as well as of planetary gears
by using the Finite Volume Method. Their simulation
results show very good agreement with the experimental
measurements of corresponding churning losses. It was
pointed out that CFD methods provide a reliable tool for
expanding the knowledge on oil flow and hydraulic
power losses in gearboxes.
Liu et al. [10] applied a three dimensional simulation
model based on the Finite Volume Method to investigate
the oil distribution and the churning power losses of a
single-stage spur gearbox. The results for both oil dis-
tribution and churning loss torque agree very well with
high-speed camera recordings and experimental meas-
urements of the loss torque. It was concluded that the
Finite Volume Method is very suitable for determining
the oil flow and the churning power losses in geared
transmissions.
Since the introduction of parallel computing, the
Smoothed Particle Hydrodynamics (SPH) CFD method
has become an interesting alternative to investigate the
oil flow of mainly automobile transmissions.
By applying a SPH model to a realistic automatic
geared transmission model, Groenenboom et al. [5] was
able to give information on the oil flow behavior of a
system consisting of 15 gears in two gearboxes. Despite
a lack of experimental validation, the SPH method
shows great potential for predicting the oil flow inside
complex transmission systems due to its greatly reduced
computing time compared to the Finite Volume Method.
Through the years, various investigations have been
conducted analyzing the oil flow of dip-lubricated geared
transmissions. However, only a few works provide de-
tailed validation of oil flow with experimental meas-
urements. Based on the previous work done by Otto [12],
the FZG no-load power loss test rig has been recon-
structed with two transparent covers made of Plexiglas®,
allowing a more detailed view inside the gearbox. The
high-speed camera Photron FASTCAM Mini AX200
provided by VKT Video Kommunikation GmbH was
used to capture the oil flow with an image frequency of
about 5000 frames per second. Some of these frames are
included in the results section of this work.
3. Object of Investigation and Operating Conditions
3.1 FZG No-Load Power Loss Test Rig
The scheme of the FZG no-load power loss test rig
is shown in Figure 1. The main parts are the electric en-
gine, the slave gearbox, the loss torque meter and the test
gearbox.
test pinion
test wheel
loss torque
electric engine slave gearbox meter test gearbox
Figure 1: Scheme of the FZG no-load power loss test
rig
The speed of the electric engine is transmitted
through the slave gearbox and the pinion shaft to the test
gearbox. The no-load losses produced by the test gear-
box are measured by the torque meter on the pinion
shaft and include the gear losses as well as the losses
caused by the four bearings and two seals. In order to
separate the no-load gear losses from the total no-load
losses, the no-load loss torque caused by the two bearings
and one seal on the test pinion shaft is measured sepa-
rately by a free shaft without the pinion and subtracted
from the total no-load loss produced by the test gearbox.
The oil temperatures are monitored by sensors in the
slave and test gearbox in the sump of the oil. Figure 2
shows photographs of the test gearbox without and with
the transparent covers.
Figure 2: Test gearbox without and with transparent
cover of the FZG no-load power loss test rig
3.2 Test Gears
The considered test gears are of FZG type C
(a = 91.5 mm, b = 14 mm, mn = 4.5 mm, αn = 20°, no
flank modification) and made of case-carburized steel
(16MnCr5E).
3.3 Lubricants
For the experiments, three mineral oils with consid-
erably different oil viscosities are used. The mineral oil
FVA3 with a viscosity grade of 100 (ISO VG) is used as
reference. In addition to that, the significantly more
viscous oil FVA4 (ISO VG 460) and the low viscous oil
FVA2 (ISO VG 32) are used as variant lubricants. The
main data of the considered lubricants is listed in Table
1.
Table 1: Properties of considered lubricants
Parameter Symbol
Oil type
FVA3 FVA4 FVA2
ISO VG - 100 460 32
Density at 40 °C in
ρ(15°C) 864 882 855
kg/m3
Kinematic viscosity
ν(40°C) 95 480 32
at 40 °C in mm2/s
Viscosity index VI ≥94 ≥94 ≥96
3.4 Operating Conditions
Within the framework of this work, selected operat-
ing conditions with regard to different circumferential
speeds and different oil viscosities are investigated. In
summary, three test series consisting of a total of 18 op-
erating points are investigated.
The circumferential speed is varied from
vt = 0.92 m/s (n1 = 240 rpm) to vt = 10.48 m/s
(n1 = 2734 rpm). A range of six different circumferential
speeds is investigated for each lubricant.
The viscosity of the lubricant can primarily be in-
fluenced by the lubricant itself and by its temperature.
In this work, the lubricant viscosity is varied by chang-
ing the lubricant (cf. Table 1). Hence, all measurements
are carried out at an oil sump temperature of 40 °C.
The oil fill level is chosen so that 9 mm (2·mn) of the
pinion dips into the oil sump.
4. Simulation Model
During a finite volume CFD analysis, the conserva-
tion equations are solved iteratively on the finite vol-
umes. Only the main physical and numerical character-
istics of the simulation model are described below.
Therefore, the employed explanations and formulations
are partly based on Liu et al. [10] where the reader is
referred to for more detailed information.
The mesh of the single-stage gearbox of the FZG
no-load power loss test rig essentially consists of three
domains: the pinion and wheel domain, the gearbox
domain and the domain of the remeshing zone. In con-
trast to structure analysis, the entire model represents a
negative model of the gearbox. The mesh of each single
domain is connected with the other domains as shown in
Figure 3. The mesh of the pinion and wheel domain is
discretized with inflation layers consisting of prisms and
does not undergo any mesh deformation. During the
simulation, the meshes of the pinion and wheel domain
rotate inside the gearbox domain at predefined circum-
ferential speeds.
The gearbox domain is meshed with tetrahedral el-
ements and also remains undeformed. The domain of
the remeshing zone fills the cavity between the gearbox
domain and the pinion and wheel domain. During oper-
ation, the meshing zone of the pinion and wheel is a
transient area that changes at every gear meshing posi-
tion. Therefore, the remeshing zone consists of a de-
2
formable meshing structure that changes with every
time step of the rotating pinion and wheel domain. The
domain of the remeshing zone is discretized with de-
formable tetrahedral elements that follow every rigid
movement.
oil tracks starting at the tooth tips. When the prescribed
circumferential speed is reached, the form of the oil
tracks finally takes characteristic shape.

vt=0.92m/s vt=1.38 m/s FVA3

vt=0.92 m/s vt=1.38 m/s FVA2

Figure 3: Pinion and wheel domain (top left); domain of vt=0.92 m/s vt=1.38 m/s FVA4
remeshing zone (top right); outer domain of the gearbox
(bottom left); entire mesh model (bottom right) Figure 4: Recorded oil distributions by high-speed
camera for circumferential speeds of vt = {0.92;
In order to avoid numerical singularities due to bad 1.38} m/s and mineral oils FVA3, FVA2 and FVA4
element quality in the very small gap between the tooth (ϑoil = 40 °C)
flanks i.e. backlash, both the pinion and the wheel are
scaled to 98% of their actual size. It is not expected that
the results with respect to oil distribution and loss torque
will be significantly influenced by this simplification.
According to Mauz [11], the squeezing losses caused by
sqeezing the oil particles between tooth flanks can be
neglected for the considered direction of rotation con- vt=0.92m/s vt=1.38 m/s FVA3
sidered in this study (cf. Figure 3).
5. Results
As an example of the considered operating conditions
of this work (cf. section 3.4), six different operating
conditions with two circumferential speeds (vt = {0.92;
1.38} m/s) and three oil viscosities (FVA2, FVA3, FVA4) vt=0.92 m/s vt=1.38 m/s FVA2
are considered in this extended abstract. In order to
check the performance of the considered finite volume
CFD simulation model, the operating conditions are
chosen to obtain conspicuous oil flow behaviors. The
quality of the CFD simulation results can thereby be di-
rectly validated with the recorded oil flow. Thus, the vt=0.92 m/s vt=1.38 m/s FVA4
circumferential speed has been set to low values so that
characteristic oil tracks remain behind both the pinion Figure 5: Simulated oil distributions by CFD simula-
and the wheel. tion for circumferential speeds of vt = {0.92; 1.38} m/s
All simulated oil distributions are directly compared and mineral oils FVA3, FVA2 and FVA4 (ϑoil=40°C)
with recordings made by a high-speed camera (Photron It can clearly be seen that the oil flows reveal major
FASTCAM Mini AX200 provided by VKT Video differences in comparison with the considered mineral
Kommunikation GmbH) positioned at the front of the oils. Hence, the high-speed camera recordings clearly
test gearbox. Figure 4 shows the recorded oil distribu- highlight the difference in oil viscosity. Due to the small
tions by the high-speed camera, whereas Figure 5 illus- differences in oil density between FVA2, FVA3 and
trates the corresponding CFD simulation results. FVA4 (cf. Table 1), its influence is expected to be sub-
After the start-up of the engine, the oil is gradually ordinate.
taken out of the oil sump and carried along with the When the reference oil FVA3 is used, the centrifugal
teeth, until it reaches the meshing zone, in which it is force swings the oil off the tooth flanks and leaves
squeezed from the narrowing tooth gap. As the circum- characteristic oil tracks in the upper half of the gearbox.
ferential speed increases, the oil dragged along the tooth On the pinion side, the approximately 1.5-fold centrifu-
flanks is gradually swung off the teeth forming visible gal force compared to the wheel is able to swing a
3
 greater amount of oil off the tooth flanks, so that pro- to investigate the oil flow inside a single-stage gearbox
nounced oil tracks fill the right half of the gearbox. On of the FZG no-load power loss gear test rig. The inves-
the wheel side, due the weaker centrifugal force, the oil tigations show that the finite volume CFD method is a
swung off the tooth flanks does not reach the space valid tool for predicting the oil flow inside a gearbox.
range of the oil thrown off by the pinion. When the cir- The results are very consistent with the high-speed
cumferential speed is increased to vt = 1.38 m/s, the camera recordings. The investigations show that the in-
centrifugal force is approximately 2.3-fold compared to fluence of both oil viscosity and circumferential speed
vt = 0.92 m/s. This leads to more expanded oil tracks in can be captured very well.
space as the oil is thrown a further distance from the Based on the presented CFD simulation model, the
gears. impact of various parameters like the influence of gear
In order to analyze the influence of the viscosity, the geometry, guide plates, immersion depth and others on
less viscous mineral oil FVA2 is used. In general, the oil both the oil distribution and the churning power losses
flow is similar to FVA3. However, due to the lower lev- will be investigated in further studies.
el of resistance to shear stress of FVA2, the oil particles
7. References
are thrown a further distance. Furthermore, the formed
[1] Concli, F.: Efficiency of gear transmissions and
oil tracks are thinner than for FVA3. Instead of refined
CFD analysis of the load independent power loss-
oil tracks on both sides like FVA3, the oil tends to turn
es. Dissertation, Politecnico di Milano. (2012).
into irregular larger oil droplets on the side of the wheel
[2] Concli, F., Conrado, E., Gorla, C.: Analysis of
as well as at the region above the meshing zone, partic-
power losses in an industrial planetary speed re-
ularly when the circumferential speed is increased to
ducer -Measurements and computational fluid dy-
vt = 1.38 m/s.
namics calculations. Engineering Tribology, V.
When the far more viscous lubricant FVA4 is used,
228, p.11–21. DOI: 10.1177/ 1350650113496 980
the oil flow appears completely different. The resistance
(2013).
to shear stress of FVA4 is higher compared to FVA3 and
[3] Concli, F. and Gorla, C. Numerical modeling of
FVA2. Thus, especially when the circumferential speed
the power losses in geared transmissions: Windage,
is only vt = 0.92 m/s, the oil dragged by gears “sticks”
churing and cavitation simulations with a new in-
together and is not swung away. Hence, a sort of belt is
tegrated approach that drastically reduces the
formed around both gears instead of refined oil tracks.
computational effort. Tribology International, V.
When the circumferential speed is increased to
103, p.58–68. DOI: 10.1016/ triboint.2016.06.046.
vt = 1.38 m/s, the higher centrifugal force at the tooth tips
(2016)
of the pinion results in higher momentum of the oil par-
[4] Gorla, C., Concli, F., Stahl, K., et al.: CFD Simu-
ticles, which is great enough to form refined oil tracks
lations of Splash Losses of a Gearbox. Advances
instead of oil belts. As compared to FVA3 and FVA2 less
in Tribology, V. 2012, DOI: 10.1155/2012/616923
oil is swung off from the tooth flanks along the way to the
(2012).
gear meshing zone, a higher amount of oil is dragged into
[5] Groenenboom, P. H. L. Mettichi, M.Z., Gargouri,
the gear meshing zone and squeezed out. It can also
Y.: Simulating Oil Flow for Gearbox Lubrication
clearly be observed that a greater amount of oil is present
using Smoothed Particle Hydrodynamics. Interna-
on the front surface of the gears. This is also due to the
tional Conference on Gears, Garching (2015).
high resistance to shear stress of FVA4. Another notable
[6] Jafarian, P. Gear interlocking effect study using
aspect is the dynamic formation of oil sump, which dif-
CFD. CTI Symposium, Berlin. (2014).
fers significantly from FVA3 and FVA2. Especially on
[7] Klier C., Berger L., Stock K.: New prospects for
the side of the wheel, the oil sump is deflected to a con-
oil flow simulation in rotating spur-gear systems.
siderable angle.
Internat. Conference on Gears, Garching (2015).
The simulation results for the oil flow in Figure 5
[8] Kvist, T. Oil splashing simulation using CFD,
are generally in very good agreement with the
Trollhättan, Schweden. (2012).
high-speed recordings in Figure 4. The influence of the
[9] Li L., Versteeg H.K., Hargrave G.K., Potter T.,
circumferential speed and the lubricant viscosity is cap-
Halse C. Numerical Investigation on Fluid Flow of
tured well and even the characteristic oil tracks are pre-
Gear Lubrication. SAE International Journal of
sent in the simulation results. Only the resolution of the
Fuels and Lubricants, 1 (1) , p.1056-1062, DOI:
oil tracks and of small dispersed oil droplets (especially
10.4271/2008-01-1650 (2009).
for FVA2) as well as the results with FVA4 leave room
[10] Liu, H., Jurkschat, T., Lohner, T., and Stahl, K.:
for improvement.
Determination of oil distribution and churning
In this work, a CFD simulation model with a rela-
power loss of gearboxes by finite volume CFD
tively small number of elements and therefore a compa-
method. Tribology International, V. 109, 346–354,
rable low resolution was used. The element size of the
DOI: 10.1016/j.triboint.2016.12.042 (2016).
model as well as the kinematics, i.e. acceleration time of
[11] Mauz, W.: Hydraulische Verluste von Stirnradge-
the start-ups of the gears, may be optimized to refine the
trieben bei Umfangsgeschwindigkeiten bis 60 m/s.
simulation results. The results on the churning losses
Dissertation, Universität Stuttgart. (1987).
were not presented in this extended abstract.
[12] Otto, H.-P. Flank Load Carrying Capacity and
6. Conclusion Power Loss Reduction by Minimised Lubrication.
In this study, a CFD simulation model was applied Dissertation, FZG, TU München. (2009).

View publication stats