Introduction
Welcome to Level 2, Lesson 3 – Monitoring Machinery
Vibration Using Dynamic And Process Control Signals, part
of CTC’s free online vibration analysis training series.
We hope you enjoyed and benefitted from the previous course
and will continue to build your vibration analysis knowledge as
you progress through Level 2.
Monitoring Machinery Vibration Using Dynamic And Process Control Signals is
created and presented by CTC for complimentary educational use only. This
training presentation may not be edited or used for any other purpose without
express written consent from CTC.
Training
Objectives
This training will focus on the following concepts:
Dynamic Signal Process Signal
Upon completion of this lesson, you will:
1 Understand the difference between a dynamic signal and a process signal
2 Understand the strengths and limitations of each type of signal
Understand how to use dynamic signal and process control signal together in
3 some instances to offer an optimal level of protection for critical machinery
Traditional
Vibration Monitoring
Traditionally, industry has been monitoring machinery vibration with various specialized vibration
analyzers. The most widely used systems are designed to trend the vibration levels of various critical
capital equipment.
Typically, a vibration sensor sends a dynamic signal (an mV output of the time waveform) to the analyzer.
The analyzer can then convert the time waveform into an FFT which displays the amplitudes of the
individual frequencies which make up the complex time waveform.
A trained analyst can study the FFT and time waveform trends to determine what types of problems a
piece of machinery might be experiencing and further identify the root cause of such problems.
Vibration analysis systems typically take the form of online monitoring systems which continually
poll a series of measurement points; or portable data collectors which can be used for periodic
measurements. Either type of equipment can be very effective for analysts to trend vibration levels, set
alarms and monitor a wide range of faults on a wide range of applications.
Vibration Monitoring –
Traditional Methods For Vibration Alarms
Several methods have been used to
establish vibration alarms. At their most A yellow alarm is the first alarm that warns
basic, all methods attempt to analysts or end users that vibration levels
establish a baseline at which a piece have increased and in-depth analysis of the
machine is in order.
of machinery operates in a normal or
healthy state.
Analysts will typically set two alarm
levels (although some utilize more) at A red alarm is triggered at higher vibration
some level of increased vibration over levels, indicating catastrophic failure is a
this normal state. The alarms are serious near-term threat. Sometimes red
designed to alert an analyst or end user alarm levels can trigger an automatic or
to a potential problem with the manual machine shutdown.
machinery.
Vibration Monitoring –
Typical Alarm Methods
Some examples of typical alarm methods are:
Trending Time Waveform Values FFT Monitoring
Overall or peak values, Peak, or peak to peak Masking, banding, or
or the crest factor enveloping
Vibration Monitoring –
Trending In this trend display, the overall value has been recorded
over a period of 5 years.
Warning, alert, and fault alarms were set but never
exceeded.
The peak value or crest factor could also have been
trended.
Trending alarms based on overall, peak, or crest factor
are very useful for alarming for a general vibration level.
It is important to keep in mind that an individual fault
frequency could be overshadowed by the increase or
decrease in amplitudes of other fault frequencies; or
it could be ignored or disproportionately weighted
relative to other frequencies contributing to the trend
value.
Vibration Monitoring –
Time Waveform
In this waveform, the Peak, or Peak to Peak values of
the time waveform can be alarmed.
This is a much more instantaneous type of alarm, and
the impacts in the time waveform are exceeding the
alert and fault limits.
Again, a specific fault frequency which could
indicate a bearing fault (for example) could be
overshadowed by lower frequency, higher
amplitude vibrations. This tends to be a good
alarm method for catching lower frequency
transients and impacting.
Vibration Monitoring –
FFT
In this FFT, a process of Masking, Banding, or Enveloping
specific frequency regions is being used.
This method works very well when you want to have
different alarm levels at different frequency ranges
so that you can monitor for specific faults.
Traditional
Vibration Monitoring
Historically, online monitoring and
portable data collection have required
specialized vibration instrumentation.
However, in today’s environment of
Process Control, 4-20 mA current loops
can be used with PLCs or DCS systems
to generate vibration alarms, primarily
using systems which many industrial Online Monitoring Portable Data Collection
plants already have on site.
Traditional Concepts -
Added Value Of Process Control
A process control signal (such as 4-20 mA) can assist traditional vibration analysis in protecting:
Critical applications Applications which can Applications that can Applications which are
fail quickly fail dramatically remote, or where no
operators can monitor
the machinery
The PLC or DCS can constantly monitor for catastrophic failure and alarm analysts for potential
problems.
This allows analysts to spend more time on analysis and less time putting out fires, or worrying
about a catastrophic failure due to operator error or environmental issues between route
measurements.
Process Control –
For PLCs Or DCS Systems
The 4-20 mA signal can be provided to the PLC or DCS by any of the following methods:
4-20 mA Loop Power Dual Output 4-20 mA Transmitters
Sensors Loop Power Sensors which convert or
that provide 4-20 mA conditions a standard
and temperature mV (dynamic) signal to
output an mA signal
Traditional Concepts –
How Does It Work?
The loop power (4-20 mA) sensor or transmitter generates a current
signal which is scaled to the maximum load (the highest vibration level
the user would plan to alarm for) and is represented by a 20 mA output.
An output of 4 mA represents no load (no vibration).
It is important to understand that the overall 4-20 A signal output is
proportional to the overall amplitude generated with a defined
frequency band.
Therefore, the signal does not include data from frequencies outside
the frequency band and includes all vibration (critical faults and non-
critical) within that band.
Process Control –
Process Scaling
Scaling is an important factor to consider when specifying a loop power sensor.
The scale (or measurement range) you choose should position the vibration levels at normal or
healthy conditions at roughly 8 to 10 mA of output.
This will allow you to establish alarm limits (for example) at 12 to 16 mA and shutdown limits at
18 to 20 mA.
It is important that you remember that the overall amplitude will be based on the frequency band
you have selected for your sensor or transmitter.
Process Control –
Process Scaling
This chart shows the
scaled output for a
0 to 2 G or 0 to 2 IPS
(50.8 mm/s) scale
Process Control –
Process Scaling
In some cases, a broader
than normal scale may
be desired when a
significant increase in
vibration above normal
can be tolerated before
alarm and shutdown is
required.
For example, your
application might operate
normally at 0.75 IPS and
varying loads might make
higher overall levels a
normal occurrence that
would not merit an alarm.
Instead, alarms and
shutdown might not be
desired until 4.5 IPS.
Process Control –
Loop Power Sensor: 4-20 mA
Loop power sensors The acceleration or The frequency Loop power sensors
are normally velocity units can be ranges for loop are great for
available in expressed as Peak power sensors are trending and
acceleration or or RMS values. generally available in alarming the overall
velocity output. two fixed bands, vibration.
Generally, Peak will such as:
Acceleration output provide more info Just remember, that
will give the higher about transient 10-1,000 Hz since there is no
frequencies more variations, while RMS (600-60,000 CPM) dynamic output from
proportional value in (which essentially this sensor, there is
the overall amplitude averages the peaks) 3-2,500 Hz also no time
relative to a will give less (180-150,000 CPM) waveform or FFT,
comparably attention to and therefor is not
specified sensor. transients. intended for
diagnosing what the
alarm might be
caused by.
Process Control –
Vibration Data
This chart shows a 27-minute trend of process
vibration data from a 600 HP Boiler Feed Pump.
Initially, it looks as if the power failed on all four of the
vibration sensors, creating the zero output in the
center of the display.
How can this be analyzed?
What do you think we should start looking for as a
cause for the loss of data?
Could this be a powering issue with our PLC or
sensors?
Process Control –
Vibration Data With Motor Current
Let’s redisplay the data in a 100 second interval and
add the motor current to the display.
The motor current also went to zero.
With the addition of other process control data, we
can see that the motor, not the sensors, was
accidently shut off for 45 seconds, and then turned
back on.
This could be a topic for coaching the end-user, rather
than searching for power issues on the vibration
sensors.
This is a good example of how process
control data can help to paint a fuller picture.
Blended Approach –
Best Of Both Worlds
Dynamic vibration sensors paired with signal conditioners (vibration transmitters) provide a
4-20 mA output and a dynamic signal output. By utilizing both signals, companies have a
solutions that gives the best of both worlds:
Continuous monitoring Alarming for catastrophic Diagnostic analysis and
through the PLC or DCS failure through the PLC or trending for specific fault
system DCS system frequencies via a dynamic
signal analyzer
Convenient access to all Better use of analysts’ time Team approach to
data via permanently since they can spend less protection and monitoring
mounted sensors which time trending perfectly brings more resources to
speeds route collection and healthy equipment and more protecting machinery and
increases safety time analyzing data plant health
Blended Approach –
Dynamic Sensors And Transmitter
The dynamic signal from a permanently mounted vibration sensor can also be converted to a
process control signal by a vibration transmitter (or signal conditioner).
The transmitter receives an mV signal from the sensor, then filters and scales the signal to a
4-20 mA output.
The 4-20 mA signal can then be passed to a PLC or DCS where it can be monitored with other
process control data.
Transmitter systems also allow access to the dynamic signal so that analysts can trend specific
fault frequencies or diagnose the cause of alarms.
Most units offer a buffered output via a BNC on the face of the transmitter, as well as a pass-
through via screw terminals so analysts can run the signal to a traditional switchbox.
The larger size of the transmitter allows manufacturers to offer greater flexibility for filter options
than standard loop power sensors.
The ability for analysts to choose from a menu of high pass and low pass filters provides the
opportunity to target alarms more accurately, rather than using less specific overall values.
Blended Approach –
Dynamic Sensors And Transmitter
Standard sensors and vibration transmitters can also be configured with a local monitoring
system to offer several options which are valuable in a variety of applications, making them an
extremely versatile tool for vibration analysts and process control engineers.
These systems can include:
Relays Digital displays
make it possible to shut down machinery in of vibration levels in the scaled or actual value
the event vibration levels exceed a user-
defined level
Alarm options Retransmission
visual or auditory, such as lights or sirens of a process control signal to a PLC or DCS
Blended Approach –
Cooling Tower Application
A good example of protecting equipment
or processes using diagnostic and
process control signals is a cooling
tower application where the gearbox has
been the primary source of problems.
Dual output sensors (shown in red) are
on the input and output shafts of the
gearbox.
mV/g sensors (shown blue) on all the
bearings.
This allows analysts to trend for bearing
faults and alarm for imbalance or gear
noise.
Process Control –
Integration
Remember, integrating dynamic vibration with process control can make a very successful program for
monitoring your machines.
Why Choose
CTC?
Superior durability Accuracy and performance
Industry’s best lead times Knowledgeable staff
Industry leading warranties
Connect With CTC
We look forward to hearing from you
FAILURE
18-20 mA
www.ctconline.com sales@ctconline.com
Time
Waveform
Connection Technology Center
(585) 924-5900
