WHITE PAPER
How to Get Early Warnings of
Rolling Element Bearing Faults
Energy & Power Generation
Programmable Sensor - Bearing Condition Transmitter
Visit us online at www.imi-sensors.com
Toll-Free in USA 800-959-4464
716-684-0003
PROGRAMMABLE SENSOR - BEARING CONDITION TRANSMITTER
George Zusman, Ph.D., D.Sc.
Director of Product Development – PCB Piezotronics
3425 Walden, Depew, NY 14043
gzusman@pcb.com
Abstract: This paper introduces the new Bearing Condition Transmitter, a
programmable 4-20 mA loop powered device-sensor contained in typical
accelerometer housing. It is specially designed to provide early warning of typical
ball/rolling element bearing faults such as cracked races, spalling, brinelling, and
looseness. It has five modes of detection that are user selectable by a simple
software program through a PC’s USB port. The options include RMS acceleration,
true PK acceleration, compensated peak (using bearing diameter and speed to
normalize output), crest factor, and crest factor plus based on an original
combination of the PK, RMS and crest factor for improved detection on variable
speed machinery. A structure diagram of the transmitter and test results are
discussed.
Key Words: Rolling/Ball Bearing, Vibration, Transmitter, Condition Monitoring
Introduction: Machinery and mechanical systems face potential failure when the
ability to function normally is compromised due to worn components or when
operating conditions diverge from normal. Continuous monitoring of vibration
levels help avoid expensive unplanned shutdowns by detecting machinery faults
before they become catastrophic events. It is known that machinery vibration
changes when problems such as worn bearings occur. Bearings are needed whenever
one part of a machine slides against another and can be classified as either sliding or
rolling contact bearings. For rolling contact bearing condition monitoring, a new
field programmable and cost-effective sensor has been designed, the Rolling/Ball
Bearing Condition Transmitter. The sensor units are constructed with a two pin
independent polarity connector and can work directly with PLC or DCS systems. All
parameters and configurations of the units are USB programmable through a PC.
Methods: Two groups of methods are widely adopted for the determination of
rolling bearing health and presence of faults. Although not always possible, the best
results are obtained when both methodologies are adopted.
The first group of methods, which are diagnostics orientated, is based on the
separation and analysis of discrete components of certain frequencies which make
up excited oscillations in the bearing. The diagnostic features are frequency
components of the spectrum and characteristics of the signal pulse shape associated
with the characteristic frequencies of the bearings: the pulse peak value (usually
harmonic amplitude), the ratio of the harmonic energy to the noise level, and the
amplitudes of spectrum components at the pulse repetition frequency. To analyze
these parameters use is made of vibration signal spectra, spectra of AM-envelopers
of narrow-band high frequency components of the vibration signal in a range of 0.1
to 40 kHz and vibration time waveforms.
The second group of methods is based on the determination of the technical
condition of the bearing as a whole. In a case of loss of serviceability, it is
paramount to determine the necessity for the bearing replacement (i.e. determine its
health). The cause of failure may be determined later, if required, by visual
inspection of the bearing. The condition of the bearing is evaluated by the degree of
development of degradation; a process that may be separated into four known
stages. The following is a list of possible diagnostics parameter: RMS and PK of
acceleration, characteristics of amplitude distribution, moment characteristics
(dispersion, excess), correlation and regression variances, amplitude discriminates,
various parameters with use of peak (crest) factor and it combination with RMS and
PK of vibration and comparison of vibration parameters in various frequency bands.
One important property of field applicable methods is a strong ability to separate
current process characteristics, such as speed and loading, from bearing defects. The
RMS, Shock Pulse, Crest Factor, Kurtosis, High Frequency Resonance Technique
(HERT), Spike EnergyTM , gSE, HFD etc are the main methods of bearings
condition related to the second group. Several popular methods for determining
bearing condition are listed bellow.
•
Real or true PK of Acceleration and HFD
•
True RMS of Acceleration
•
Crest (peak) factor CF = Apk/Arms (trend is required) [1]
•
Shock pulse method (rpm is required) [2, 3]
•
W1x (Apk/Arms) + W 2 x Arms [4]
•
Enveloping with Spectrum Analyses, Cepstrum and others
•
K – factor K= ApkxArms and other high order moments (rpm is required)
•
Strum factor K0 /Kt = Apk0xArms0 / Apkt x Armst
•
Kurtosis (rpm is required)
•
Modulation deepness of envelop signal (rpm is required)
Operation: The new Bearing Condition Transmitter is designed to use the first five
methods (from the second group) listed above. It is a programmable 4-20 mA loop
powered device-sensor contained in a typical accelerometer style housing. The view
of the sensor Bearing Condition Transmitter is presented at Figure 1. Five modes of
detection are user selectable by the PC software by a USB port- RMS acceleration,
true PK acceleration, compensated peak (using bearing diameter and speed to
normalize output), crest factor (CF), and crest factor plus (CF+) based on original
combination of the PK, RMS and CF for improved detection on variable speed
machinery. The two operation pins used for sensor field programming as well as
factory calibration by a special programmer shown on Figure 2.
Figure 1 - Bearing Condition
Transmitter view
Figure 2 - Sensor USB Programmator
Figure 3 - Customer Software Screenshot
The customer software screenshot is shown on Figure 3. The left part of the screen
provides the setup of the sensor and consists of two sections. Top one is “Constant
Speed Machinery” section includes follow choices - true PK of acceleration (A pk),
true RMS of acceleration (A rms), both with user programmable full scale output 2
to 50 g, and true PK with compensation (A pk with compensation) which is
modified shock pulse method [2, 3] for non resonance measurements and require
user to insert the bearing diameter and rotation speed. The bottom is “Variable
Speed Machinery” section includes crest (peak) factor (CF) which is ratio of
acceleration PK and RMS and represented at the sensor output as number of mA
equal to the CF. The crest factor plus (CF+), which is combination of the PK, RMS
and CF, is modified known technology [4] for improving bearing defect detection
on variable speed machinery. The output is useful for detecting high level defects
which can lead to decreasing CF.
PiezoSensor
Filter 1
Vibration
Ampl.
PK
Detector
A/D
Filter 2
RMS
Detector
CPU
Switch
Digital
Interface
4-20mA OUT/Power
V/I
Power
Supply
Figure 4 - Simplified block diagram of the Bearing Condition Transmitter
D/A
Block Diagram and Test Results: Simplified block diagram of the Bearing
Condition Transmitter is shown on Figure 4.
The Bearing Condition Transmitter includes an embedded accelerometer (piezosensor) that generates a voltage output that is proportional to the shock and
vibration sensed on the machine bearing. The output of the sensor is passed
through a band pass filter (250 Hz to 10+ kHz) and coupled through an amplifier
to a high speed peak detector and thru other band filter (2.5 kHz to 10+kHz) or
directly to a true RMS detector.
The PK detector and RMS detector monitor the continuous vibration signal and
hold the highest values seen within the sample window. The PK and RMS
values are passed through the analog-to-digital converter (A/D) to the CPU. The
sensor output is a 4-20 mA current proportional to measured parameter based on
USB programmed settings.
CF+ vs. RPM
14
12
CF+
10
8
Bearing w/o defects
6
Bearing with combine
defect
4
2
0
0
500
1000
1500
2 000
2500
300 0
RPM, 1/min
Figure 5 - Output current (CF+) of Bearing Condition Transmitter vs.
rotation speed for bearing without defect (bottom line) and for bearing with
combine defect (top line).
The plot shows the comparison of normal and defective bearings by using the
Bearing Condition Transmitter CF+ mode at different rotation speeds represented at
Figure 5.
References:
1.
Weichbrodt, B. and Darrel, B. “Damage Detection Method and Apparatus for
Machine Elements Utilizing Vibrations Therefrom“, USA Patent 3,677,072
July 18, 1972.
2.
Sohoel, E.O. “Method and Arrangement for Determining the Mechanical
State of Machines”, USA Patent 3,554,012, Feb. 29, 1968.
3.
Sohoel, E.O. “Method and Instrument for Determining the Condition of an
Operating Bearing”, USA Patent 4,528,852, Jul. 16, 1985.
4.
“Defect Factor: A dedicated tool for Rolling Element Bearing Monitoring”,
01dB-Meravib, 2001.
5.
Harris, T.A. “Rolling Element Bearing Analysis”, John Wiley & Sons,
1991.
6.
Yu, J.J., Bently, D.E., Goldman, P., Dayton, K.P. and Van Slyke, B.G.
“Rolling Element Bearing Defect Detection and Diagnostics Using
Displacement Transducers,” ASME Journal of Engineering for Gas
Turbines and Power, 2002, Vol.124, pp.517-527.
WHITE PAPER
What are divisions of PCB Piezotronics?
PCB Piezotronics, a member of the PCB Group families of companies, has five major divisions, all of which offer targeted sensor technologies. These
divisions are supported by an active outside direct sales force of Field Application Engineers, as well as international direct sales offices throughout the world. Individual PCB Piezotronics divisions, locations and their primary product specialties include:
Depew, NY, USA - www.pcb.com – Piezoelectric, ICP®, piezoresistive & capacitive
pressure, acoustic, force, torque, load, strain, shock & vibration sensors.
Depew, NY, USA - www.pcb.com/aerospace – Sensors & Instrumentation for
aerospace & defense applications, including air and spacecraft testing.
Novi, MI, USA - www.pcb.com/auto – Sensors & Instrumentation for automotive testing,
including modal analysis; NVH; component durability; powertrain testing; vehicle dynamics; safety and
regulatory testing.
Depew, NY, USA - www.imi-sensors.com – Industrial vibration sensors, bearing fault detectors, mechanical vibration switches, panel meters, cables & accessories for predictive
maintenance and equipment protection.
Depew, NY & Provo, UT, USA www.larsondavis.com – Precision microphones, sound
level meters, noise dosimeters, audiometric calibration systems.
San Clemente, CA, USA - www.pcb.com – Research & development engineering center for
special technologies.
MEMS ADVANCED DESIGN CENTER
Seattle, WA, USA - www.pcb.com – Process development and fabrication of
MEMS sensors.
PCB® Group Companies
Farmington Hills, MI, USA - www.pcbloadtorque.com – Designs and manufactures high quality,
precision load cells, wheel force transducers, torque transducers, telemetry systems, and fastener
torque-tension test systems.
DYNAMIC METROLOGY
Cincinnati, OH, USA - www.modalshop.com – Global leader in dynamic calibration offering a
complete line of automated calibration systems and recalibration services to support dynamic
vibration, pressure and force sensors in applications such as: national standards, commercial
labs, government/military research, consultancies, and industrial/plant floor operations.
Rochester, NY, USA - www.sti-tech.com – Mechanical engineering consulting firm
specializing infinite element analysis, advance analytical techniques, experimentation,
technology development, &design optimization for turbo machinery, industrial machine
systems & mechanical structures.
Corporate Headquarters 3425 Walden Avenue Depew, NY 14043-2495 USA
Toll-Free in USA 800-828-8840SMn 24-hour SensorLine 716-684-0001
n Fax 716-684-3823 n Email imi@pcb.com
Toll-free in the USA 800-959-4464 n 24-hour
SensorLine
716-684-0003
E-mail
info@pcb.com
n www.pcb.com
n ISO 9001 CERTIFIED n A2LA ACCREDITED to ISO 17025
AS9100 CERTIFIED
ISO 9001 Certified n AS9100 Certified n A2LA Accredited to ISO 17025
SM