Monitoring of Grease Lubrication: Risto Parikka, Research Scientist and Aino Helle, Senior Research Scientist

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

NT2006 -10-73

MONITORING OF GREASE LUBRICATION

Risto Parikka, Research Scientist and Aino Helle, Senior Research Scientist
VTT Technical Research Centre of Finland
firstname.surname@vtt.fi, P.O.Box 1000, FI-02044 VTT, Finland

ABSTRACT

There are several interacting factors in grease lubrication of bearings. They include e.g. the grease
quality, grease supply quantity and interval, bearing housing construction, rotational speed, load,
temperature, sealing as well as grease removal. In order to find out the major problem areas with
respect to grease lubricated bearings, a survey was carried out within industrial companies
representing mainly pulp and paper industry. The most common problems experienced with grease
lubrication included poor grease removal, overlubrication, deficient lubrication and contaminants.
The most prominent reason for bearing failures according to the survey was lack of lubrication or
lubricant.
Monitoring of the lubricant film thickness in grease lubricated bearings is more demanding than in
the case of oil lubrication. It is commonly accepted that in grease lubrication the viscosity index or
κ-ratio does not give a reliable estimate of the lubricant film thickness. Whereas acoustic emission
can be used to determine the critical rotational speed below which boundary lubrication conditions
apply in an oil lubricated bearing, it is possible in grease lubricated bearings that the lubrication
conditions remain good at far lower speeds or that lubrication becomes instable such that it varies
between properly functioning and boundary lubrication. The grease contained within the bearing
also has a vibration damping effect. On the other hand, if there is insufficient amount of lubricant
available for the roller path and the necessary pressure formation required for elastohydrodynamic
(EHL) lubrication cannot be realised, the contact area may become deficient of grease even in
conditions where oil lubrication should theoretically be good.
In this paper results are reported from experimental studies with a rolling element bearing test rig
using vibration based monitoring methods, with the objective of detecting the condition of deficient
lubrication and optimising the lubrication. The test rig enables also experimental research on the
effects of overgreasing, different grease supply and removal methods or the effects of contaminants
in grease.
1 INTRODUCTION

There are several interacting factors in grease lubrication of bearings. They include e.g. the grease
quality, grease feeding quantity and interval, bearing housing construction, rotational speed, load,
temperature, sealing as well as grease removal.
Monitoring of the lubricant film thickness in grease lubricated bearings is more demanding than in
the case of oil lubrication. It is commonly accepted that in grease lubrication the viscosity index or
κ-ratio does not give a reliable estimate of the lubricant film thickness [1]. Whereas acoustic
emission can be used to determine the critical rotational speed below which boundary lubrication
conditions apply in an oil lubricated bearing, it is possible in grease lubricated bearings that the
lubrication conditions remain good at far lower speeds or that lubrication becomes instable such that
it varies between properly functioning and boundary lubrication. The grease contained within the
bearing also has a vibration damping effect. On the other hand, if there is insufficient amount of
lubricant available for the roller path and the necessary pressure formation required for
elastohydrodynamic (EHL) lubrication cannot be realised, the contact area may become deficient of
grease even in conditions where oil lubrication should theoretically be good [2].
In order to find out the major problem areas with respect to grease lubricated bearings, a survey was
carried out within industrial companies representing mainly pulp and paper industry. Some
commonly known methods for detecting deficiencies in grease lubrication were reviewed and a
rolling element bearing test rig was equipped and instrumented for testing grease lubricated
bearings. In this paper results are reported from experimental studies with the test rig using
vibration based monitoring methods, with the objective of detecting the condition of deficient
lubrication and optimising the lubrication. The test rig enables also experimental research on the
effects of overgreasing, different grease feeding and removal methods or the effects of contaminants
in grease.

2 PROBLEMS IN GREASE LUBRICATION

In order to find out the major problem areas with respect to grease lubricated bearings, a survey was
carried out within industrial companies [3-4]. The industrial companies represented mainly pulp and
paper industry, but included also one steel works and one mine.
At one paper mill, tens of thousands of lubrication points may exist, and the annual person hours
used for lubrication may present several person years. The consumption of lubricating grease may
typically amount to 3000 - 8000 kg at one works, and the major part of this is used for lubrication of
rolling bearings. A single central lubrication unit may supply lubricating grease for hundreds of
bearing, and the distance between the lubricating unit and the most remote bearing may be several
tens or even hundreds of meters. Also, the difference in altitude may be tens of meters. In addition,
the bearings lubricated by the same lubricating system may operate at different temperatures,
vibrational conditions and rotational speed, which sets different and great demands for the grease
and the lubrication system.
a) The m ost com m on problem s w ith grease lubricated rolling
bearings

5
Number of replies

n
e
n

ion
n

an

es

y
s io
ur
t io

t io
.. .

t it
in

nc

ct
t
ic a

an
ra

r ro
r ly

ic a

am

le
ba
pe
pe

qu
br

Co

se
br

nt

ur
rl u

m
lu
ro

ed
Co

e
t
Te

dis
tp

as
ve

nt

fe
ie

re
no

e
f ic

es

th
al

De

oc
ov

of
Pr

n
m

io
re

at
e

t im
as

Es
re

Problem
G

b) Failure reason of grease lubricated rolling bearings

14
12
Number of replies

10
8
6
4
2
0
g
re

t
se
nt

re
t

r
ul
n

in
ul

g
n

ea
t

io
an

ilu

tu

in
io

ica

fa

ad
ea
fa
at

lw
...

ar
ra
s

fa
ic

n
gr

rlo
br
ro

g
in

be
br

pe

sig
g

ca
in
al
r

lu
am

ve
in

d
co

nt
lu

Se

ni
fie

De
at
ng

O
ou
e/

i
or

te
nt

ha
ni
ric

ct
ro
ur

co

gh

ec
n

po

fe
b
W
st
io

lu

de
er

Hi

M
Sa
oi
at

th

ic
/m
ic

g
at
O

ri n
br

er

m
lu

tu
at

to

ac
of

au

uf
ck

of

an
La

lt

M
au
lf
na

Reason for failure


t io
nc
Fu

Figure 1. a) Problems experienced and b) reasons for failure of grease lubricated rolling bearings
according to a survey carried out within Finnish industrial companies [4].

Figure 1 shows a summary of the type of problems encountered and their prevalence according to
the survey as well as reasons for bearing failures [4]. The most common problems experienced with
grease lubrication included poor grease removal, overlubrication, deficient lubrication and
contaminants. The most prominent reason for bearing failures according to the survey was lack of
lubrication or lubricant.
3 METHODS FOR DETECTING DEFICIENCIES IN LUBRICATION

Vibration based methods, particularly at high frequencies, can be used for detecting when the
contact in rolling element bearings becomes deficient of lubricant. A short description of some of
the most common methods is given below.

3.1 Haystack-effect
The maintenance personnel of some of the industrial companies in the afore mentioned survey had
experience about such cases where deficient lubrication of bearings had been visible in a relatively
high frequency acceleration spectrum (max. 10 kHz) as an increase in the vibration level. In some
published articles the phenomenon is known as the "haystack-effect" (Fig. 2). According to ref. [5]
the haystack effect is a phenomenon which normally appears when the lubrication quality or
effectiviness deteriorates and results in metal-to-metal contact in bearings. If the situation cannot be
remedied for example by lubricant addition, this may result in a bearing failure. In the articles
published this phenomenon has been observed particularly with rolling element bearings, for
example in electric fan motors. Physically, the phenomenon is probably due to the activation of the
natural frequency of some structural components as a result of the metal-to-metal contacts between
the bearing raceways and the rolling elements.

Figure 2. "Haystack effect" due to deficient lubrication [5]. Spectrum before (left) and after (right)
lubricant addition.

3.2 Other high frequency vibration measuring methods

Based on literature and published research it can be noted that besides the afore mentioned high
frequency vibration measurement, a number of other vibration analysis methods exists for detecting
lubrication deficiencies, such as for example PeakVue (Figure 3) and SPM. PeakVue is a method
developed and patented by CSI (Computational Systems, Incorporated). The PeakVue -meathod utilizes
high-pass filtering in order to enable an accurate analysis of higher frequencies and small amplitude
phenomena occurring at high frequencies. SPM or Shock pulse method is a vibration measurement
method patented by SPM Instruments AB. It is based on a sensor having a resonance frequency of
32 kHz. These methods are described in more detail e.g. in references [6-8].
In some cases deficient lubrication in industrial machinery has been detected by condition
monitoring measurements as an increase in the level of envelope spectrum without the appearance
of peaks typical to bearing failures. Changes detectable in high frequency vibration can also be
emphasized by derivation of the acceleration signal.

3.3 Acoustic emission


The use of acoustic emission (AE) for estimating the lubrication condition has been mentioned in a
number of references, for example refs. [9-10]. As examples of the various AE-methods the
following can be mentioned: counting the number of pulses exceeding a certain treshold level
(Pulse Count), AE parameter monitoring of Holroyd Instruments (Distress and dB-level), the SEE
method of SKF AB which utilizes enveloping, as well as the continuous monitoring of the AE level
using certain parameters (e.g. rms, peak, kurtosis etc.). The last mentioned monitoring method has
been used also at VTT for monitoring the oil film thickness and the running clearance in rolling
element bearing tests [11]. By using acoustic emission it is also possible to detect solid
contaminants in lubricating greases as well as various changes in the lubrication parameters [12].

3.4 Lubricant analysis

Information concerning the condition of both the lubricant and the lubricated machine can be
obtained also by lubricant analyses in addition to vibration based methods. The selection of the
lubricant analysis methods to be used is made on the basis of the requirements to monitor for
example the metal content, water content, oxidation, elemental analysis or other changes in the
grease. The various analysis methods have been reviewed e.g. in references [13-15]. One of the
shortcomings of monitoring grease lubrication is that the current techniques available are not
particularly suitable for on-line monitoring of the lubricating grease. The development of methods
suitable for rapid determination of the condition and properties of the lubricating grease in field
near the machinery could be the next step forward in grease condition monitoring.

4 TESTING EQUIPMENT

The tests were performed using the rolling bearing test rig of VTT shown in Fig. 6. For the grease
lubrication tests, a replaceable bearing housing fitted with a grease supply was installed to the test
rig and instrumented with a number of sensors. The bearing housing was of type SNL 507-606
made by SKF. The grease can be supplied either from the side of the bearing or directly into the
bearing through a hole in the outer ring. The test rig was also equipped with an automatic grease
lubrication unit made by SKF. The test bearing was a spherical roller bearing of type 22207 EK.
The outer diameter of the outer ring of the bearing was D = 72 mm and the inner diameter of the
inner ring was d = 35 mm.
During testing a radial load is applied on the test bearing using a hydraulic power unit and a
hydraulic cylinder. The bearing load can be varied within the range from 0 to15 kN and the
rotational speed within the range from 0 to 2500 rpm.
Figure 6. A schematic presentation of the rolling element bearing test rig (left) and a photograph of
the test rig equipped with a number of sensors and an automatic grease supply unit (right).

Instrumentation of the bearing housing includes accelerometers for horizontal and vertical vibration
measurement and acoustic emission (AE) sensors. Two thermo elements are used for temperature
measurements, one for bearing temperature close to the loaded zone of the bearing outer ring and
another one for ambient temperature outside the housing beside the test rig. In addition, the sound
level can also be measured and an MHC-Memo AE sensor and datalogger of Holroyd Instruments
is available for continuous measurement of the Distress and Desibel parameters of Holroyd.
A PC equipped with an AD-card is used for data acquisition and the parameters calculated from the
signals are stored in a database in the computer. Additionally, raw signals and spectra are stored for
possible further actions in text files on the basis of a separate command. Several time domain
parameters of the vibration acceleration and AE signals are stored in the database, including e.g.
root mean squeare (rms), peak value (peak), crest factor, skewness and kurtosis. The vibration
acceleration parameters in frequency domain are stored as the rms-value of the spectrum in five
different frequency bands, i.e. 0-2 kHz, 2-4 kHz, 4-6 kHz, 6-8 kHz and 8-12 kHz.

5 RESULTS

Grease lubrication tests were run using the testing equipment with constant test run parameters for
such a time that a similar type of increase in the vibration level as the haystack-effect described in
chapter 3.1 was observed. After this regreasing of the bearing was carried out and it's effect on the
measured parameters was observed. The grease supply was applied from the side of the bearing.
Three similar test series were run and each gave comparable results.
The test results are illustrated as the trend of the bearing outer ring temperature, the trend of the rms
value in time domain of the vibration acceleration within 0-12 kHz, the trends of the vibration
acceleration in the different frequency bands of the spectrum and the trend of the acoustic emission
rms value obtained in the test run performed between 26.9. and 30.9.2005.
Bearing temperature
80

75
Temperature [ C]
o

70

65

60

55

50
26.9.2005 26.9.2005 27.9.2005 27.9.2005 28.9.2005 28.9.2005 29.9.2005 29.9.2005 30.9.2005 30.9.2005 1.10.2005
0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00
Date/time

Figure 7. The temperature trend measured from the bearing outer ring during the test run on 26.9.-
30.9.2005.

Vibration acceleration 0-12 kHz, rms

18,38

14,71
Acceleration [m/s ]
2

11,03

7,35

3,68

0,00
26.9.2005 26.9.2005 27.9.2005 27.9.2005 28.9.2005 28.9.2005 29.9.2005 29.9.2005 30.9.2005 30.9.2005 1.10.2005
0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00
Date/time

Figure 8. The trend of the time domain vibration acceleration rms value at 0-12 kHz during the test
run on 26.9.-30.9.2005.

RMS values of vibration acceleration spectrum

5,70
5,06
4,43
Acceleration [m/s ]
2

3,80 0-2 kHz


3,16 2-4 kHz
4-6 kHz
2,53
6-8 kHz
1,90 8-12 kHz

1,27
0,63
0,00
26.9.2005 26.9.2005 27.9.2005 27.9.2005 28.9.2005 28.9.2005 29.9.2005 29.9.2005 30.9.2005 30.9.2005 1.10.2005
0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00
Date/time

Figure 9. The trends of rms values of the vibration acceleration spectrum in the different frequency
bands during the test run on 26.9.-30.9.2005.

Acoustic emission, rms

0,0140

0,0120

0,0100
Voltage [V]

0,0080

0,0060

0,0040

0,0020

0,0000
26.9.2005 26.9.2005 27.9.2005 27.9.2005 28.9.2005 28.9.2005 29.9.2005 29.9.2005 30.9.2005 30.9.2005 1.10.2005
0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00
Date/time

Figure 10. The trend of the rms value of acoustic emission during the test run on 26.9.-30.9.2005.
6 DISCUSSION

During the first test run a sharp increase in the vibration level was observed about 58 hours from the
start of the test. At the higher frequency bands (8-12 kHz) the change was particularly distinct but it
was also clearly noticeable in the overall level of vibration. The change was detectable at all
frequency bands but at the lowest band of 0-2 kHz it was very faint and the vibration level here was
lower than the level during the test start up. During the early stages of the measurement two short
periods of a lower vibration level were also observed at all frequency bands. The origin of these is
believed to have been in the measurement equipment (e.g. the amplifier), since similar phenomena
could not be observed in any other test runs.
A decrease in the temperature was observed right from the beginning of the test and it continued
even when the vibration level started to increase. The decrease in the temperature was 11-12 oC
from the temperature measured in the beginning of the test. As about 10 g of grease was added to
the bearing, the vibration levels were soon observed to normalize again. At the same time a steep
increase in the temperature was observed until the temperature became even remaining at a lower
level than it had been right in the beginning of the test.
The level of acoustic emission decreased during the first part of the test and it appeared to follow
the change of the temperature to some extent. The change occurring in the vibration level was also
detected in the acoustic emission measurement but it was clearly smaller. In the rms-value the
change was visible as a slight increase in the level but the AE amplitude was then much lower than
in the beginning of the measurement. After regreasing the AE level did not decrease as clearly as
the vibration signal.

Figure 11. Vibration acceleration spectra measured with the OROS spectrum analyzer during test
run #2 at a situation when the vibration levels are at their highest and the lubrication condition
apparently deteriorated. The spectrum at the top has been measured vertically from the top of the
bearing housing and the one below horizontally from the side of the housing.
Figure 11 shows acceleration spectra in the frequency range of 0-20 kHz measured using the
OROS-spectrum analyzer. The spectra in Fig 11 are from the test run #2 from such a situation when
the vibration levels were at their highest. The spectrum at the top has been measured vertically from
the top of the bearing housing and the one below horizontally from the side of the bearing housing.
The frequency range corresponding to the frequency with the increased vibration level in the trend
curves is clearly visible in the spectra. This "haystack" corresponds to frequencies of 10-16 kHz
when measured from the top of the bearing, with a maximum at a frequency of about 13 kHz. When
measured from the side of the bearing the corresponding frequency range and its maximum is
located at somewhat lower frequencies.

7 CONCLUSIONS

The vibration acceleration rms at the frequency band of 8-12 kHz was the most sensitive of the
measured parameters for detecting deficient lubrication condition in the current test arrangement
and bearing type used. When examining the vibration spectrum in the frequency range of 0-20 kHz
using the OROS spectrum analyzer it was observed that the increase in the vertical vibration level
took place at the frequencies of 10-16 kHz, with a maximum at about 13 kHz. Some natural
frequency which is very sensitive to changes in the lubrication condition is located in this frequency
range. The frequency ranges of the higher vibration levels are slightly different when measured
horizontally as compared to that measured vertically. This indicates that the stiffness of the
resonating structure is different in different directions. On the basis of the measurements made it is
not possible to say for certain whether it is a question of the natural frequency of the bearing, the
bearing housing or some structural components, or possibly the natural frequency of the sensors or
their fastening. Further measurements will be made in order to clarify this natural frequency and the
respective component, using for example the shock pulse method or different sensors and ways of
fastening them.
The bearing temperature seems to have a direct correlation with the change of the lubrication
condition. In the tests where the temperature could be measured, it decreased by more than 10 oC as
compared to the start of the run, and the lowest temperatures were measured at those situations
when the vibration measurements indicated deficient lubrication conditions. This correlation may
possibly be used in the determination of the optimal regreasing interval and quantity.

8 ACKNOWLEDGEMENTS

The work reported in this paper is part of the grease lubrication case study of the project
Prognostics for Industrial Machinery Availability. The National Technology Agency of Finland and
participating industrial companies are gratefully acknowledged for their financial and technical
support. Thanks are also due to Mr Jari Halme and Dr Erkki Jantunen for helpful discussions and to
Mr Hannu Sainio for assistance in constructing the test rig and in performing the tests.
9 REFERENCES

1. SKF 2003. SKF General Catalogue 5000 E. 1120 p.


2. Cann, P. & Lubrecht, T. 1998. Mechanism of grease lubrication in rolling element bearings.
COST 516 Tribology Symposium. Espoo, Finland, 14 - 15 May 1998. Espoo, VTT, 457 p.
VTT Symposium 180. ISBN 951-384573-7.
3. Hynönen, P., 2005. Centraliced grease lubrication of rolling bearings. Master's Thesis.
Tampere University of Technology, Department of Mechanical Engineering. 102 p.
4. Parikka, R., 2005. Methods and requirements regarding identification and remedies of faults
in grease lubrication of rolling bearings. VTT Report BTUO43-051355. 15 p. (in Finnish)
5. Stevens, D, 2004. Vibration Analysis Pinpoints Inadequate Motor Bearing Lubrication.
Practicing Oil Analysis Magazine. May 2004. Available also in Internet at
http://www.practicingoilanalysis.com/article_printer_friendly.asp?articleid=603 and at
http://www.noria.com/learning_center/category_article.asp?articleid=603&relatedbookgrou
p=OilAnalysis. Cited on 18.3.2005.
6. Jones, R.M., 1994. A guide to the interpretation of machinery vibration measurements - Part
2. Sound and Vibration, No: 9. p. 12 - 20.
7. Sundberg, A, 1997. Shock pulse method for condition monitoring of bearings. In Jantunen,
E., Holmberg K., Rao, R.B.K.N. (Ed.). COMADEM ’97. 10th international congress and
exhibition on condition monitoring and diagnostic engineering management, Vol. 2. Espoo:
VTT Technical Research Centre of Finland. p 554 - 563. (VTT Symposium 172).
8. Kunnox 2004. A case example of the use of PeakVue-measurement for condition
monitoring of low speed rotating machinery. Internet page: http://www.kunnox.fi. Cited on
18.10.2004.
9. Li, Y., Billington, S., Zhang, C., Kurfess, T., Danyluk, S., Liang, S., 1997. Dynamic
prognostic prediction of defect propagation on rolling element bearings. Tribology
Transactions, Vol 42, No. 2. p. 385 - 392.
10. Shiroishi, J., Li, Y., Liang, S. Kurfess, T., Danyluk, S., 1997. Bearing condition monitoring
diagnostics via vibration and acoustic emission measurements. Mechanical Systems and
Signal Processing, Vol. 11. No. 5. p. 693 - 705.
11. Parikka, Risto; Vaajoensuu, Eero; Vidqvist, Ville; Tervo, Jyrki. 2002. Test equipment and
diagnostic system for determining the lubrication condition and running clearance of the
rolling bearing. Comadem 2002. Birmingham 2 - 4 Sept. 2002. Comadem International, p.
153 - 162.
12. Miettinen, J., 2000. Condition monitoring of grease lubricated rolling bearings by acoustic
emission measurements. Tampere University of Technology, Publications 307. Doctoral
Thesis, Tampere. ISBN 952-15-0477-3.
13. Parikka, Risto. 2004. Basic measurements in oil condition monitoring. Kunnossapito, vol.
18, 5, p. 24 - 26. (in Finnish)
14. Parikka, R. & Sainio, H. 2004. Fundamentals of grease lubrication of rolling bearings. VTT
Industrial Systems. 31 p. + App. 4 p. Research report BTUO43-041258. (in Finnish)
15. Evans, J.S. & Hunt, T.M. Oil Analysis. Machine & Systems Condition Monitoring Series.
Coxmoor, 2004.

You might also like