Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Agenda:
Introduction to maintenance and asset management
Machinery Monitoring and Diagnostics
Machinery Vibration System
Data Acquisition and Signal Conditioning
Signal Processing
Technical Data Management
Example Real-World Installations
Architectural Summary
Conclusions
2
The Economics of Maintenance
Globalization of economies
Pressure to reduce manufacturing costs
Higher volume markets (machine uptime)
Geographic specific factors
Aging workforce and equipment
Reductions in capital investment for newer machinery
Increased machinery automation technologies
Programmable Automation Controllers
With the globalization of our world economy, competition for our customers
exists from every reach of the globe. As a manufacturing company, we are
able to reduce the cost of our products in part by improving the reliability of
our manufacturing machinery. Further, with a world wide market, we can
produce products in higher volumes. Higher volumes in turn imply higher
utilization and uptime of our machinery.
In the USA, manufacturing suffers from an aging workforce and from aging
equipment relative to newly mobilizing economies. Cost pressures are even
higher in the USA as newly emerging economies introduce a competitive cost
structure. As our corporations debate adding manufacturing operations
overseas, we delay capital investment in newer machinery.
3
Benefits of Intelligent Maintenance Systems
(Asset Management)
Reduce scrap and raw material consumption
Lowers Cost of Raw Materials
Reduce downtime
Increases Output and Revenue
Lengthen maintenance cycle
Lowers cost of maintenance and service
As machinery runs more smoothly and reliably, we have less waste of raw
materials. Often times this comes from the prevention of machinery failure in
the middle of a production run. In addition, vibrations from unhealthy
machinery can actually impact product quality as well. For example, in rolling
mill applications such as paper and steel, bearing vibrations on rollers can
create chatter on the rollers thus leaving a washboard like imprint on the end
product.
4
Asset Management
Sensors
Is the sensor working (bias voltage)
Calibration information and health
Machinery
Scrap rates, energy consumption
Mechanical Health (gears, bearings, etc.)
Control Systems
PID loop tuning
I/O Functionality Health
On the other end, the control systems contain both analog and digital inputs
and outputs (I/O) as well as control logic. The I/O calibration information, and
operational status is another element of asset management in a production
environment. Further, with changing suppliers of raw materials, control
systems may need to be re-tuned. An out of tune process creates waste in
raw materials, and can even damage or increase wear rates on machinery.
5
What is a Maintenance System?
Service Personnel
Spare Parts Inventory
Machinery Monitoring and Diagnostics
THE DATA
Computerized Maintenance Management Systems
Personnel, spare parts inventory, and corrective actions are all driven by
data. This data is collected from machinery, and in particular from machinery
monitoring and diagnostics hardware and software applications.
6
Machinery Monitoring and Diagnostics
Electrical Power
Machinery Vibration
Thermography
Tribology
Operational Monitoring
7
Machine Condition Monitor System Elements
Software
Application
Development Tools
Development
Sound and Vibration,
Signal Sources/ Environment
Order analysis,
LabVIEW, C/C++,
Signal processing
Unit Under Test LabWindows/CVI, etc.
Acoustics, Sound,
Vibration, Audio
Computer
PCI or PXI-based
Hardware
Transducers
Simultaneous
Accelerometers,
Acquisition/DSA
Proximity Probes,
NI 4472, NI4451,
Tachometers,
SCXI, SCC
Pressure probes
8
Vibration Sources
Loose Mechanical Unbalance
Components Bent Shaft Blade Pass /
Gears Fluid Related
Slot Frequency /
EM related
Alignment
Motor
Journal (Fluid Film)
Bearings
Mechanical Rolling Element
Resonances Couplings
Bearings
9
IEPE Accelerometers
Direct Connection
Advantages
Simple and easy to use
Built-in microelectronics
Simple constant current signal conditioner
(18-30 VDC ; 2 mA)
Limitations
Temperature range: max 250 F (some 325 F)
Fixed sensitivity
IEPE refers to a type of transducers that are packaged with a built-in charge
amplifier or voltage amplifier. IEPE is an acronym for Integral Electronic
Piezoelectric. Because the charge produced by an accelerometer is very
small, the electrical signal produced by the transducer is susceptible to noise,
and sensitive electronics must be used to amplify and condition the signal.
IEPE makes the logical step of integrating the sensitive electronics as close
as possible to the transducer to ensure better noise immunity and more
convenient packaging.
The major advantage of the IEPE accelerometer is that it is simple and easy
to use due to the built-in microelectronics and the simple conditioner required.
10
Tachometers
Measure
Angular position
Angular speed (rpm)
Used for
Order analysis
Order tracking
Angular averaging
More than providing this angular speed information, the signal from the
tachometer is needed to perform advanced analysis such as order
tracking, order analysis, and angular averaging.
11
Harmonics of a Fan
4th order
7th order
1st order
Hz
55
385
220
3,300 RPM = 55 RPS
12
Data Acquisition Considerations
Number of Input Channels
Dynamic Range
Sampling rate
DSA Digital Filter vs. 8th Order Analog Bessel
Filter Response(dB)
in band of interest
-40
Anti-Aliasing Filters
-60
-80
Signal
Bandwidth
-100
Analysis of data is only as accurate as the data that is collected. Attention to data acquisition system
specifications is equally as important as selection of sensors and analysis.
Accelerometers and other transducers convert vibration into a voltage. Modern MCM systems
digitize this voltage for transfer to a computer for analysis. This brings us to the next element of a
machine monitoring system that works with dynamic vibration.
This acquisition hardware might be a plug-in instrument that resides in an industrial computer
(PXI/CompactPCI), a desktop PC, or an industrial distributed I/O system such as CompactRIO.
National Instruments offers a number of options for hardware that performs this function. We will get
to these in a moment. For now, lets consider some common requirements.
Dynamic range characterizes how well you can distinguish a low-amplitude signal in the presence
of high-amplitude signals. A high dynamic range ensures that you will not miss low amplitude
components that are sometimes key to successful analysis. The latest dynamic signal acquisition
board from National Instruments offers 110 dB spurious free dynamic range.
Anti-aliasing filters are required to avoid aliasing that could potentially ruin your measurement.
Anti-aliasing filters are low-pass filters, meaning that they reduce errors caused by the resonance
frequency of the accelerometer, as discussed earlier.
The sampling frequency is usually driven by the maximum rotational speed of the machine and
the highest order of interest. As a rule of thumb, min sampling rate = 11 * max rotational speed *
max order or harmonic of interest.
Simultaneous sampling ensures that you gather useful cross-channel phase information
Multiple input gains allow the user to take advantage of the full dynamic range of the analyzer, on
a per channel basis
Finally, built-in signal conditioning for accelerometers and microphones (often a polarization
voltage) is a nice feature, especially where space is a concern.
The NI 4472 data acquisition device is available for CompactPCI and PXI computing platforms. This
device provides eight simultaneously sampled inputs with 110 db dynamic range. It offers sampling
rates to 104.2 kS/s with a 47 kHz bandwidth using built in hardware tracking anti-aliasing filters. The
hardware also provides piezoelectric power for accelerometers and offers both AC and DC coupling.
Any number of NI-4472 devices can be utilized in a single or in multiple chassis to provide the
channel count needed for the application. A similar device, the NI-9233 is available for CompactRIO.
13
Anti-Aliasing
14
Signal Conditioning
DC and AC Coupling DC Coupling = AC+DC
AC Coupling = AC AC
5 volts
DC Coupling = AC+DC
DC
IEPE = Integrated
Circuit Piezoelectric 0 volts
built-in charge amplifier
or voltage amplifier AC Coupling = NO DC
15
Signal Processing
Extracts features from vibration signals
Fourier Transform is most common tool
Order Analysis identifies harmonics of rotation
Other Analysis techniques
Envelope Analysis
Modulation analysis
Wavelets
Now that we have a solid data acquisition system in place, we can focus our
attention on the mathematics used to extract mechanical health indication
features from the vibration signal.
The most common signal processing tool used to extract machinery health
indicators from machine vibration signals is the Fast Fourier Transform or
FFT.
16
Typical Spectrum Showing Basic Faults
0.01
Velocity (ips)
unbalance
misalignment
0.001
Courtesy of
The workhorse analysis tool for vibration monitoring is the Fast Fourier
Transform, or FFT. This tool allows us to focus attention on particular
frequency space to completely understand the vibration characteristics
of the mechanical components, including amplitude, frequency, phase,
harmonics, sidebands, and so forth.
17
Gear Mesh Example
Gear mesh amplitudes
much higher than
running speed 1XGMF
2XGMF
spectrum
Number of Sidebands
and distance from center 5 10 15 20
Frequency (orders
25 30
Courtesy of
An extension to the FFT is the zoom FFT. The resolution of the FFT
result is dependent on the amount of data collected. The zoom FFT
stores up consecutive blocks of data until the desired frequency
resolution of the X-axis is obtained. Further more, the zoom-FFT only
analyzes the raw signal for repetitions in the desired frequency range.
Gear frequencies are an example use case for zoom-FFT. Gear mesh
amplitudes are at much higher frequencies than running speed. The
gear mesh frequency and sidebands are visible in the spectrum. Here,
a zoom-FFT will help to further understand the sideband peaks, their
relationship to the center frequency, amplitude ratios, and so forth.
Often, shaft sidebands around gearmesh suggest input shaft
problems.
18
Orders
.5X 4X 5.5X
3X
600
1 1200
2 1800
3 2400
4 3200
5.5
600 RPM frequency (CPM)
frequency (orders)
Express frequency in
terms of multiples of
the shaft running speed
Another extension to the FFT is Order Analysis. With order analysis, the
rotational speed of the machine is used to normalize the x-axis of the
resulting FFT to show harmonics or rotation or angular position rather than
harmonics of time, or frequency.
In this example, the tachometer signal us used to identify the rotational speed
of the fan, and to normalize the frequency axis of the resulting FFT.
A more complete order analysis technique is to transform the data from the
time domain to angular domain with a tool called Resampling. With
Resampling, any FFT skew in the graph is removed by analyzing equal
angular sampled data. This signal processing technique is very common for
machinery which changes speed such as variable speed drives, turbo
machinery, and so on.
19
Synchronous & Non-Synchronous Frequencies
mm/s
Synchronous
frequencies are 1X
whole number 2X
multiples of the .5X 3X
4X 5.5X
shaft speed
600 1200 1800 2400 3200
frequency (CPM)
Gear Mesh, Balance, Mechanical Looseness, and blade pass frequencies are
repetitive vibrations that are synchronous.
Roller element bearing vibrations are repetitive, yet are typically non-
synchronous due to their mechanical anti-friction design.
20
Bearings
21
Bearings
When a machine changes speeds, the FFT, combined with zoom-FFT, and
order analysis becomes a key signal processing element in the machinery
monitoring and overall maintenance programs.
22
These are tabulated in
a bearing atlas
Pd = Pitch diameter
n = number of balls
Bd = Ball diameter
= Contact angle
As a side note, the bearing fault frequencies are related to the mechanical
design of the bearings.
23
Trending over time key to changes
Collecting data over time provides trends in
vibration parameters
Technical Data Management allows storage of
signatures for off-line signal analysis
Both are key elements of an intelligent maintenance
systems
Now that we have a tool to track the magnitude or level of vibration from key
mechanical components, we need to use this information for reports and
predictions of failure.
Using a trend plot of say bearing vibration, we can monitoring the trend over
time and determine whether the vibration is rising, or remaining the same.
Rising vibration levels tend to indicate a failure waiting to happen. Using
vibration severity charts, past experiences, design knowledge from the
supplier, and so on, we can set alarms on these vibration levels to tell the
maintenance team when to order replacement bearings and to schedule
maintenance.
Going further, the maintenance team often chooses to store the raw vibration
signal such that it can be further analyzed off-line and compare to periodic
recordings over time. These raw signals can be used for further diagnostics
activities, and to double check the alarm trend levels from above.
24
Datalogging Features in LabVIEW 8.0
Using a historical trending and alarming software, we can track the results of
the FFT for any and all machines and mechanical components of the
machine. The alarming system will notify the maintenance team if a vibration
level has risen above a specific level. The historical trending system will
provide a history of the vibration levels, and allow for vibration level reviews
both before and after a repair has been completed. The historical trending
system also serves then to verify that the maintenance activity is successful.
25
Configuring Alarms and Trends
With an alarming tool, the maintenance team has access to the alarm
settings, and programmatic actions when alarms occur.
26
Locate Your Test Data and Load All of It
Manage
Inspect
Analyze
Report
Connect
Generic Plug-In Interface
Automate
When storing raw vibration signals from a large number of machines and
sensors, we create a technical data management problem.
This allows the maintenance team to keep track of and utilize all the data
collected from the vibration measurement devices employed in the plant.
27
Inspect Your Test Data
Manage
Inspect
Analyze
Report
Connect
Now that the data is accessible, the maintenance team will want to preview
the data prior and interactively analyze the data prior to making formal
reports.
Visual data inspection helps you investigate why certain events happen
during your tests. You can use DIAdem to correlate several test runs and
visually compare results. You also can zoom in, scroll across, and
interactively analyze graphs of your measurements to quickly and fully
visualize your test results.
28
Report your Test Data
Manage
Inspect
Analyze
Report
Connect
Now that the data is accessible and analysis completed, reporting is the next
step in the vibration data aspect of maintenance. Reports are used to both
justify repairs and to report the success of a maintenance activity.
29
Putting it all together
Multiple machine
monitoring nodes
integrated together over a
network
Local signal processing
Data logging at a host
engineering workstation
Signal Processing to
extract features
Using these tools in a networked environment allows for data sharing and
automatic communication to other systems in the plant, including the
computerized scheduling systems for parts, personnel, and production.
30
Instituto de Investigations Electricas SICAD
for PEMEX and CFE
Challenge:
Develop a cost effective and advanced
machine monitoring and diagnostic tool
for Oil Platforms and Power Generation
Flexible, Cost Effective, Advanced
Process and Vibration Monitoring
NI Tools
PXI, DSA, Counter Timer
LabVIEW, S&V, OAT, RT
Results
12 systems deployed
Pemex asking for more
Adapted for portable solution
31
Dresser-Rand VAM System
Low Speed Shaft Key Phaser
x x x
Turbine/
y Gear Box y
Electric Motor y x x
x
x Compressor
y y y y
Key
Phaser High Speed Shaft
Vibration Probes/ 2 leads:
Acceleration/velocity
Challenge: To develop a comprehensive, cost effective, and advanced compressor remote
monitoring, diagnostics, and data capture system. Systems to be deployed for Oil and Gas
Industry Installations.
NI Tools Used: PXI DSA, 6602, LabVIEW Pro, S&V, OAT, Internet Tools
Results: RECON Total is Dresser-Rands flagship condition monitoring system that provides on-
line, fully integrated, fully remote, core knowledge base capabilities with comprehensive
vibration, performance, process, and stress wave analysis.
32
Duke Energy Reactor Pump Vibration
Monitor
Challenge
Provide a combined system for 32
vibration channels in four machinery
locations
NI Tools
PXI, LabVIEW, Order Analysis Tools
Results
Real-Time Monitoring
Historical Data
Transient Data
Alarming on Vibration parameters
Adaptable to other process and
control variables
33
Vibdaq Solutions for TurboMachinery
Challenge
Develop a portable vibration
diagnostics tool for turbo machinery
applications
Needs flexibility, scalability, graphics,
and advanced signal processing
NI Tools
LabVIEW, PXI, DSA, Order Analysis,
USB DAQ
Solution
Leveraging off the shelf technologies
Vibedaq now has a strong competitive
posture in the portable machine
diagnostics applications market
34
Power Systems Manufacturing DDAQ
Challenge
Develop a dynamic pressure monitoring
solution for Gas turbine pressure pulsations
Interface the on-line system to control systems
for operator use
Tools
NI DSA, LabVIEW, Signal Processing
Results
A superior early warning device with multiple
screen views to monitor overall turbine
performance or focus on individual combustors
35
National Renewable Energy Labs (NREL)
Challenge: To develop a power,
vibration monitoring, and control
solution for study of Wind Turbine
mechanical systems. System
needs to be advanced, flexible,
and cost effective
NI Tools Used: PXI AI, AO, DSA,
LabVIEW RT, OAT, S&V
Results: Vibration Systems on par
with typical systems, plus
integrated with electrical power
quality/metering, and control
systems
36
The NI Computer-Based Monitoring Option
Frequency Analysis
Networked Access
Flexible Configuration
Time Synchronized
Data Logging
Orbit Plot
Bode Plot
Polar Plot
Order Spectrum
Shaft Centerline
Trend Charts
Data Archives
Alarming
Transient Recording
Sequence of Events
37
Sound and Vibration Platform
Embedded Controllers
LabVIEW RT and FPGA
Programmable Automation
Controllers RMS levels
PXI Industrial Controllers
Portable PCs
Desktop PCs
38
Solutions
www.ni.com/industrial/machine_vibration.htm
Cal-Bay Systems: REMS100
http://www.calbay.com/REMS.htm
Omnia Technologies: OVT1000
http://optimation.us/omnia/index.asp, http://vibdaq.com
Nexjen: VTS
http://www.nexjen.com/nexjen/portfolio/customerdefined.htm
FAG Industrial Services: X1
http://www.fag-industrial-services.com/site/en/
Advanced Engine Technologies: TraceVIEW
http://www.aetco.com/documents/TruTrace.pdf
Dynalco: RT-Online
http://www.dynalco.com/prod_details.cfm?product_id=168
Spectraquest: Vibquest
http://www.spectraquest.com/products/vibraquest/index.html
Power Engineering: Hydro Power Plant Monitoring
http://www.powerengineering.com/
Structural Integrity: Vibration Consulting
http://www.structint.com/analytical/vibration.htm
PCB: Vibration Sensors
http://www.pcb.com/
39
Conclusion
Vibration is a key parameter for asset management
of industrial machinery
Distributed intelligent solutions provide an excellent
means of local monitoring and remote data logging
Call on NI for any of your machinery vibration needs
www.ni.com/industrial/machine_vibration.htm
40