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

net/publication/229018389

Gear test rig for noise and vibration testing of cylindrical gears

Article · January 1999

CITATIONS READS

4 801

1 author:

Mats Åkerblom
AB Volvo
7 PUBLICATIONS   46 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Mats Åkerblom on 21 November 2015.

The user has requested enhancement of the downloaded file.

 GEAR TEST RIG FOR NOISE AND VIBRATION
TESTING OF CYLINDRICAL GEARS

Mats Åkerblom

mats.akerblom@volvo.com
Volvo Construction Equipment Components AB
SE–631 85 Eskilstuna, Sweden

Abstract

Gear noise is sometimes the dominating noise in commercial vehicles. Noise testing of com-
plete gearboxes is very time-consuming and expensive. A test rig has been designed for test-
ing gears under controlled conditions. The test rig is of the recirculating power type. Finite
element analysis has been used to predict the dynamical properties of the gear test rig. Ex-
perimental modal analysis has been carried out on the gearbox housing to verify the theoreti-
cal predictions of natural frequencies. The test rig can be used for noise and vibration testing
of gears with different manufacturing errors and different design parameters. In addition to
noise testing, the rig can be used for gear life testing and measurement of efficiency.

Keywords: gear, noise, vibration, test rig, modal analysis

1. Introduction

In commercial vehicles and construction machines, for example wheel-loaders, gear noise
from the transmission is sometimes the dominating noise. Even if the gear noise from the
transmission is not the loudest, the pure high frequency noise can easily be distinguished from
other noise sources, and noise of this kind creates an impression of bad quality. In order not to
be heard, gear noise must be 10–15 dB below other noise sources, for example engine noise.
One approach to reduce noise from gears is to produce more accurate gears, usually this
means that the manufacturing method must be changed and the cost increases. Besides the
gear accuracy, also the dynamic properties of the gears, shafts, bearings and the gearbox hous-
ing are important for the noise level from a gearbox, [2] and [3].

Since noise testing of complete gearboxes is very time-consuming and expensive, a test rig
has been designed and built to allow noise testing of gears in a controlled environment. The
test rig can be used for noise and vibration testing of gears with different manufacturing errors
and different design parameters. In addition to noise testing, the rig can be used for gear life

1
testing and measurement of efficiency. The experimental noise and vibration measurements
will be used to verify different theoretical prediction methods, for example calculations of
transmission error, finite element models and multi-body system dynamic models.

2. Design of the test rig

The design process started with evaluation of different test rig principles. Due to that the test
rig will also be used for gear life testing, which requires high torque and high speed during
long time, some kind of power recirculation was desirable. The principle for recirculation of
power usually used in test rigs of this kind is electrical, hydraulic or mechanical recirculation.
The mechanical principle was chosen because of the relatively low cost and high perform-
ance. It is possible to have a high power circulating between the gearboxes and use a rela-
tively small electric motor for driving the rig, replacing power due to power dissipation in the
gearboxes. The test rig consists of two identical gearboxes, connected to each other with two
universal joint shafts. A torque is applied by tilting one of the gearboxes around one of its
axles. This tilting is made possible by bearings between the gearbox and the supporting
brackets. A hydraulic cylinder creates the tilting force. The torque is measured with a load
sensor placed between the cylinder and the gearbox.

Hydraulic Cylinder
Load Sensor
Slave or Master Gearbox Microphone
Test Gearbox

Electric Motor

Accelerometer
Articulated Attachment
Figure 1. Sketch of test rig.

The noise will be measured with one or more microphones and vibrations will be measured
with accelerometers on the gearbox housing. Free shaft ends on the test gearbox makes it pos-
sible to measure transmission error with optical encoders or measure rotational vibrations
with accelerometers. There might be a risk that the slave gears interfere with the test gears
when noise is measured [1]. To reduce this risk, the slave gearbox will be equipped with extra
wide precision ground gears and the test gearbox can be shielded off from the rest of the test
rig. The gearboxes are mounted on a concrete bed, which is placed on rubber blocks to insu-
late from vibrations. Each gearbox has its own oil-system with the possibility to filter and
control the temperature of the oil.

The vibrations that cause noise are induced in the gear-mesh and then transmitted through the
gears, shafts and through the bearings to the housing, which vibrates and emits noise. To re-
duce the influence of the housing, test rig gearboxes are often made of thick welded steel
plates and very rigid. In this test rig the intention is to include the influence of the housing in
the investigations. Therefore the shafts, bearings and housing have been designed to be as

2
similar in character as possible to a wheel-loader transmission. This is achieved by using
gears, shafts and bearings from an existing gearbox and making the housings of the same
material (nodular iron) and thickness similar to the housing of a wheel-loader transmission.

Figure 2. Test gearbox (CAD model).

3-D CAD was used for the design of the housing and “rapid prototyping” was used to create a
model in plastic, which was used as a casting model. With this method it is possible to make
small series of cast parts at a relatively low cost.

The test gears can be seen as some kind of “average gears” in a wheel loader transmission.
The test gears have technical data according to table 1.

pinion gear
Number of teeth 49 55
Normal module (mm) 3.5 3.5
Pressure angle 20º 20º
Helix angle –20º 20º
Centre distance (mm) 191.9 191.9
Face width (mm) 35 33
Profile shift coefficient +0.038 –0.529
Tip diameter (mm) 191 209
Table 1. Technical data for test gears.

The centre distance can be changed from 191.9 to 160.0 mm by turning the eccentric covers
(see figure 2 and 4) 180 º. The centre distance 160 mm is chosen to allow testing of another
existing gear-pair. An arbitrary centre distance between 160 and 192 mm can be obtained by
making new covers.

In addition to noise and vibration testing the gear test rig can be used for gear life testing and
measurement of efficiency. The measurement of efficiency is possible by measuring the
torque and rotational speed of the shaft from the electric motor. This power corresponds to the

3
power dissipated in the gearboxes. Possible investigations can be how different oils, surface
finish, surface treatment or gear geometry influence the efficiency.

Figure 3. Gear test rig.

Figure 3 shows the complete test rig and its technical data are listed in table 2.
Torque 0–5000 Nm
Rotation speed 0–2800 rpm
Power of electric motor 110 kW
Maximum power through gearboxes 1400 kW
Oil temperature 20–100 ºC
Table 2. Technical data for the test rig.

3. Theoretical modal analysis

Finite element analysis has been used to predict the natural frequencies and mode shapes for
gears with shafts, the empty housing, the complete gearbox and finally for the complete test
rig.

Figure 4. FE model of the gearbox with gears, shafts and bearings.

4
In the figures 5 and 6 some examples of mode shapes and corresponding frequencies are
shown. Boundary conditions for the housing are free–free, and for the gear and shaft the bear-
ing outside diameter is fixed.

Mode 1, 1895 Hz Mode 2, 2169 Hz Mode 3, 2676 Hz

Figure 5. The first three natural frequencies and corresponding mode shapes for one
of the gears mounted on the shaft and with bearings included
(displacements not to scale).

Mode 1, 900 Hz Mode 2, 1098 Hz Mode 3, 1215 Hz

Figure 6. The first three natural frequencies and corresponding mode shapes for the housing
(displacements not to scale).

The modal analysis of the complete gearbox resulted in about 30 natural frequencies in the
frequency range 900–3000 Hz. It is of course difficult to draw conclusions from these results
but they will be valuable for the evaluation of the measurement results.

The FE-model of the complete gearbox was also used in a harmonic response analysis. A si-
nusoidal varying force was applied in the gear mesh and the corresponding vibration ampli-
tude at a point on the gearbox housing was predicted. The frequency was swept from 1000 Hz
to 2500 Hz. In this frequency range the amplitude has a number of peaks, see fig. 7. The peak
at 1895 Hz corresponds to the first mode of gear on shaft, see fig. 5.

5
 2190 Hz

1540 Hz
Amplitude at X-dir.
Z-dir
this point
1895 Hz
Y-dir

1000 frequiency (Hz) 2500

Figure 7. Example from the response analysis results.

4. Experimental modal analysis

Experimental modal analysis has been carried out on the empty gearbox housing without
shafts and gears. When comparing the results from the experimental modal analysis with the
results from the FE predictions, the five first modes from the FE predictions could be identi-
fied.
Measured FE predicted Error in
Mode # frequency frequency predicted
(Hz) (Hz) frequency
1 861 900 +5 %
2 1086 1098 +1 %
3 1157 1215 +5 %
4 1767 1642 –7 %
5 1977 1819 –8 %
Table 3. Comparison between experimental and FE modal analysis.

The experimental modal analysis shows, see table 3, that it is possible to predict natural fre-
quencies for the housing, with FE analysis, at least in the frequency range 800–2000 Hz, with
an error less than 10 %.

5. Summary

A gear test rig has been designed and built. The test rig will be used for gear noise and vibra-
tion testing. Finite element analysis has been used to predict the natural frequencies and mode
shapes for individual parts and for complete gearboxes. Experimental modal analysis has been
carried out on the gearbox housing and the results show that the FE predictions are in good
agreement with measured frequencies.

6
 6. Acknowledgements

The author would like to thank The Program Board for the Swedish Automotive Research
Program for financial support, Alfgam Optimering AB for performing FE analysis and Inge-
manson Technology AB for performing experimental modal analysis.

7. References

[1] Houser D. R., Blankenship G. W., Methods for Measuring Gear Transmission Error
Under Load and at Operating Speeds, SAE Technical Paper 891869, 1989.
[2] Hellinger W., Raffel H. Ch., Rainer G. Ph., Numerical Methods to Calculate Gear
Tranmission Noise, SAE Technical Paper 971965, 1997.
[3] Campell B., et al., Gear Noise Reduction of Automatic Transmission Trough Finite
Element Dynamic Simulation, SAE Technical Paper 971966, 1997.

View publication stats