In selecting a transducer, the most important priority
would be sensitivity and 'Frequency range'.
To
explain the vertical axis in the graph of amplitude and frequency, it would be
nice if one sensor could measure all amplitudes, but it can never be done
because it depends on the sensitivity. To explain the horizontal axis in this
way, if all sensors can measure all frequencies, there is no reason to select a
sensor. This is because each type of vibration sensor, further subdivided, each type of vibration sensor
(including an acceleration sensor) has its own frequency range of an area where
measurement is accurate. This is called the frequency range.
Frequency range
The sensor indicates the range of measurable frequencies
(e.g., acceleration sensor: 0.5 to 10 kHz), and the definition of this range is
slightly different depending on the user, so that the maximum and minimum frequency
ranges corresponding to the 'accurate zone' of the measurement can be selected.
This
frequency range is 'a reliable area where the sensor can output a properly
matched signal; It means 'the range of frequency response accuracy according to
the frequency sensitivity deviation', and although the meaning is slightly different from the
non-linearity of the sensor, this reliable area can be expressed as a linear
area.
Looking at the criterion of error amplitude related to
the frequency domain, select the rate of change of the amplitude limit that can
be judged by the presence or absence of an error, eg) ±3dB, ±5%, ±10%, etc.,
and the applicable frequency at this time is ..........
If
the desired 'unit' of vibration is selected as a result for the evaluation, the
next step is to select an appropriate 'sensor'. First of all, it is necessary
to check whether contact attachment is possible at the location to be
measured... or whether the attachment method is appropriate. For example, in
the case of high frequency (more than 5000Hz), the magnetic attachment method
is not appropriate, and if you want to measure the behavior of an axis, it is
difficult to use a contact sensor. Also, if you try to measure the speed and
measure 5 Hz using a coin type speed sensor, you will get an amplified error
signal. This is because the characteristics of each sensor are different depending
on the amplitude band, frequency band, resonance band, etc.
Selection of displacement, velocity, and acceleration sensors
such as amplitude units
Since the output voltage or current is proportional to
each unit, the selection of the unit is not very different from the selection
of the sensor. Sensors mainly used for diagnosing and monitoring equipment
1. Eddy current type displacement sensor (Proximity) that
directly measures the behavior of a shaft supported by a sleeve bearing in a
non-contact manner.
2. Accelerometer, which propagates shaft vibration to
rolling bearing and measures it indirectly by contact method outside the
bearing housing (indirectly transmitted to the housing by impact of the bearing
connected to the shaft)
3. There is a velocity transducer that works without power.
However, among these, the speed sensor is very precise,
but has a weakness limited to the range of 10 to 1000 Hz because it has a
natural frequency in the upper and lower frequencies, which is why displacement
sensors and acceleration sensors are widely used. (If it is out of this frequency
range, an erroneous or amplified value is output.)
A sensor (transducer) is one of the components
of a system that is mainly used by companies that use sensors to research,
diagnose or manufacture monitoring equipment. Since the manufacturer has
selected a sensor that fits a specific principle, the manufacturer has
accumulated a lot of engineering grounds for this. In many cases, the level of
engineering is considerably deeper than that of academia because it must be
required and the reliability of the measurement needs to be verified.
Above all, this principle can be considered as the most
basic sensor selection method. The reason why a displacement sensor is
called a displacement sensor and an acceleration sensor is called an
acceleration sensor is that each sensor generates an electrical output
'proportional to the amplitude unit'. It's because you do it. For example, since the value
converted to displacement by outputting acceleration vibration with an
acceleration sensor and integrating twice is not very accurate (especially when
it is not a sine wave), it is better to measure acceleration vibration with an
acceleration sensor, and displacement vibration It is basic to measure with a
displacement sensor. However, there are cases where it is absolutely necessary to
evaluate the health of a machine or the vibration of a building with the
'velocity' value, which is used as the most evaluation unit in academia and
industry, so this only allows integration from the acceleration sensor once.
Because the speed sensor isn't cool...
Whatever
it is, power (#Power, output) itself is constant, but energy (#Energy,
work) gets smaller as the distance increases; As you move, the energy gets
smaller and smaller.
Noise and vibration are lost as kinetic energy and
potential energy are transformed into thermal energy or other energy. That is,
the wave energy must go through a process of attenuating and disappearing. If
you apply this, ............................
'dB (dB, decibel)', which is used as the most common way
to determine the level of amplitude, is often used when trying to explain 'to
what extent to be compared', especially for the evaluation of linearity, which
is the usable frequency range of vibration and noise, measuring instruments or
sensors. there is. Explain again how this differs from '%'.
The
reason why the linear section of the graph is selected as dB or % is used when
setting standards for reliability in various applications (mathematics,
physics, measurement, medicine, statistics, etc.). In particular, it is an
essential confirmation condition for accurate measurement and selection in the
sensor field. At this time, dB and % can be understood in terms of mathematical
principles, but they can be difficult to compare in practice, so they are
compared and explained.
dB and % are comparison methods.
dB is an amplitude value expressing log (ratio of change
rate), and % is a method of expressing the size of a linear (arithmetic) value. If % has changed by 50%,
this is self-explanatory and easy to understand. But what does it mean that dB
'changed by 3dB'?............
What is Frequency? The number of cycles per second, that
is, the number of repetitions per second, is used as a unit called ‘# Hertz (Hz)’. On the other hand, there
is ‘#Octave’ as a filter that goes well with human
listening ability, and this #filter is used to estimate the speed of sound that
can be roughly distinguished by human ears. Therefore, if the horizontal axis
is the octave-filtered Hz and the vertical axis is expressed as sound pressure
or amplitudes, it is called a #frequency analysis graph (#spectrum analysis graph) because it can indicate
which frequency of this wave is how large the amplitude is.
There is a limit to the frequency band that humans can
hear, and even within that range, the sensitivity of listening differs
depending on how many Hz the sound or vibration heard is. In other words, even
if sound
or #vibration of the same size (#sound pressure, #amplitude), according to the
frequency (#low frequency? or #high frequency?), humans say 'the size is
different'.
Phone curves - Equal loudness contours
Compared to animals with large ears, such as elephants,
humans tend to be less able to hear in the low-frequency range. However, it can
be
sensitively noticed, especially around 3,900 Hz. The reason is that they
saw the advantage of the resonance effect ............
No matter how much people try to deny and avoid it, they
cannot avoid the #wave that is transmitted through any form. A person hears
because the eardrum vibrates, and sees the reflected form with the eyes because
of the transmission by the wave of light. Breathing involves repetitive
movements of the lungs, and walking is achieved by cyclic oscillating movements
of the arms and legs. In addition, it speaks through the movement of the larynx
and tongue, and it is possible to know that there is a call through sound or
vibration. Early scholars in the field of vibration focused their efforts on
the development of mathematical theories to understand natural phenomena, and
were applied to the design of machines, structures, engines, and control
systems in an engineering way, and focused on the design of motion, and safety,
productivity, and asset management in the field. In order to efficiently
proceed with vibration, vibration continues to be researched on the health and
destruction of structures and machines, and human discomfort.
The
damage of vibration and the use of vibration
Rotating machines (engines, fans, pumps, motors, etc.)
generate #imbalance vibration due to poor design or current bad condition. The
imbalance may cause a second large vibration when rotating at high speed, and
the repetition of repeated motions leads to plastic deformation or fracture.
If an accelerometer is used to measure vibration, it
must be attached securely to the vibrating surface. Be mindful of the knowledge
of contact resonance, such as magnets, bonds, screw studs, and wax. On the one
hand, consider the issue of where to attach the #vibration sensor. Also, in
order to solve this vibration problem well, there will be questions about
which direction to measure and which spatial position of the target object to
set. Before
finding evaluation criteria for evaluating vibration, these principles should
be understood first. Here, we will explain the points and directions for
vibration measurement (#condition monitoring, #facility diagnosis, #online
monitoring, #vibration frequency analysis) promised to engineers
internationally.
Setting the vibration measurement
location and direction!
The first reason for measuring the vibration of an
object (structure or machine) is for evaluation (#Assessment), and the second
reason is for diagnosis (diagnosis) to identify the cause of vibration. However,
in order to diagnose, it is necessary to check the behavior (moving, deflection
shape) of the object in advance in addition to frequency analysis. (Non drive end, NDE,
Outboard) in each direction (horizontal, vertical, axial-H, V, A) must be measured.
This is because only then can the behavior of an object be guessed.
Of course, if the location to be measured is limited or
the object is too small, and...................................
The hypothesis that “resolving the problem of vibration
is the same as solving the cause of a mechanical defect” through vibration
analysis, that the machine condition is not good when the vibration is high No
need.
When you first enter the vibration field during
equipment diagnosis, you will learn and speak 1X (“ONE X”). Everyone has
been thinking about whether to call it “1 times, 1X” until they first hear it
being called 1X. Because the book doesn't explain it so easily. Also, in order to
find the cause of 10X, it is “one X” that must be understood..
1X, 1X, 1x TS, Order, 1x RPM
‘1X’ refers to the first of the components
proportional to the turning speed of the machine and is usually expressed in
Hz. (If it is RPM, divide by 60.) So 2X is a frequency that
is twice 1X, and therefore 10X is a frequency that is an integer multiple of 10
times 1X.
For example, when expressing 90Hz vibration in a motor
rotating at 1800rpm, you can say "90Hz amplitude is large, or 5400rpm
amplitude is low". This can be conveniently referred to as "3X is greater
or lower". In addition, this expression of 'X' can be used more usefully when the number
of rotations changes...........................
Representative forces that generate vibration include
'shock excitation force' and 'centrifugal imbalance force'. Among them,
centrifugal imbalance acts as a centrifugal imbalance force (F = mrw²) and
accounts for more than 80% of the defects of the entire rotating machine. It is
also a very important cause of defects that occupy But can this centrifugal
unbalance force (1X, 1Order) be removed?
The meaning of "imbalance must exist during
rotation" in vibration analysis is the same as "unbalance can be
reduced but not eliminated". Therefore, so-called 'balancing' is necessary.
Correction of rotational unbalance force reduces unbalance force. It can be
divided into single-sided balancing and multi-dynamic (Dynamic) depending on
the plane of occurrence. It is divided into field balancing and shop(lab)
balancing.
Multi (Dynamic) balancing
If you are balancing a plane whose thickness is almost
small compared to its diameter, it is enough to attach a mass of the same
moment in the opposite direction to the axis of the point where the unbalanced
mass is located on the single plane. Even with single-sided balancing, it cannot
be corrected if the plane is over 1/4 thickness. This is because the ‘coupling
force’ occurs in the Z direction rather than the X and Y directions. Therefore,
the
amount of unbalance exists in each of the two planes, and the amount of
vibration can be reduced only by correcting the two planes at the same time.
Field balancing
Usually, the rotating body constitutes a large machine
including the rotating shaft, so it is better to perform balancing at the site
that can produce the same number of rotations as the rotating shaft. However,
when the environment is not favorable (in the case of an important facility
that needs to be operated by replacing the spare facility immediately), the
Rotor is transported to the shop, lab, and corrected and then reassembled.
At this time, the important point is that balancing should be performed
in a normal rotation state. However, balancing in the shop is performed by
measuring at a low speed and optimum rotation speed rather than the normal
operating rotation speed, so the plane is balanced, but the unbalance force of the
shaft is not taken into account. Most of the unbalance force is measured by phase and
vibration amount and the calibration result is checked, so it is vulnerable to
the second unbalance force produced by the frame, restraint facilities, and
other 'resonance' during reassembly.