VIBRATION- SYNCHRONOUS ASYNCHRONOUS & SUB-SYNCHRONOUS SPECTRUM ANALYSIS

SYNCHRONOUS, ASYNCHRONOUS & SUB-SYNCHRONOUS SPECTRUM ANALYSIS

The raw vibration data consists of a wide range of frequencies & it’s analysis in raw form is very complex job. To analyse the data, we have to take the help of technology. FFT Analysis is a very good tool for analysing rotating assets. Raw vibration signal can be analysed in three ways-

·         Synchronous Vibration Analysis

·         Asynchronous Vibration Analysis

·         Sub-Synchronous Vibration Analysis  

Synchronous: The synchronous time averaging technique is used to detect the source of vibration at the 1X frequency. 1X frequency is the running speed of the machine. The vibration probe collects the data & the speed probe give the speed information mounted on the reference shaft. This running speed is tracked & the time averaged 1X data is monitored & analysed. This is very useful in the machine trains with variable speeds or if many other machines are running at close proximity to machine under measurement. Rotor related problems like imbalance, misalignment; rotor related looseness & rotor rub can be identified by using this technology. Other peaks are discarded.

Asynchronous Vibration Analysis: Asynchronous time averaging technique is used to detect the vibration components that are not related to rotating speed & are above 1X speed i.e. above the running speed. These faults are bearing fault, electrical noise, cavitation, etc. Resonance also produces asynchronous frequencies.

Sub-Synchronous Vibration Analysis: Sub-synchronous time averaging technique is used to detect problems like severe bearing looseness, rotor rub, vibration from nearby machines, belt vibrations & other low frequency components. These are below 1X running frequency.   

THERMOGRAPHY (THERMAL IMAGING) AN INTRODUCTION

In this section we will discuss about thermal imaging, another powerful tool in machine diagnostics & predictive maintenance arena. Thermography is an art & science to detect & measure radiation using electro-optical device & correlating to surface temperature of the subject. The history of thermal imaging can be understood by looking at the certain chronological developments in past. In 1672 Newton passed a beam of light though a prism & observed a coloured strip of light containing the colours purple, blue, green, yellow, orange and red. Later on Sir Frederick William Herschel in 1800 made an important discovery by moving Newton’s experiment one step ahead. He placed glass thermometers on different spectral colour patterns & observed a rising trend in temperature from blue to red colour spectrum. He also observed that the temperature beyond the visible colour spectrum was even higher. He concluded from this experiment that an invisible form of energy must be at work in that range, and that the sun emits invisible radiation beyond the visible light range. He called this radiation as ultra-red, now known as infrared radiation.

The laws of radiation that revolutionized the field of thermal imaging-

·      Kirchhoff’s Radiation Law: This states that every type of matter continuously radiates energy. The radiant energy is visible or invisible, depends on the temperature. Later on Kirchhoff coined the term Black Body which means that a body always emits precisely as much heat as it absorbs.

·       Hertz oscillator & Stefan-Boltzmann-law: In 1865 James C. Maxwell first predicted the assumption that light consists of electromagnetic waves which was later confirmed experimentally by Heinrich Rudolf Hertz. Josef Stefan and Ludwig Boltzmann refined Gustav Kirchhoff's theory in 1884 and established the findings of their research in the Stefan-Boltzmann law which states that the total thermal energy emitted by a black body depends on that body's intrinsic temperature.

·       Planck’s radiation law: In 1900 Max Planck did a breakthrough discovery which laid foundation of the present day understanding of electromagnetic radiation & gave birth to quantum physics, & still regarded as physical basis for thermography. It defines the intensity distribution of the electromagnetic energy emitted by a black body as a function of temperature, wavelength and frequency.

In principle thermograph is device to detect temperature pattern in the infrared wavelength of the subject. It started with the Herchel’s experiment & later on after discovery of seeback effect, which led to invention of the thermomultiplier, an early version of a thermocouple by Leopoldo Nobili. Samuel Langley used bolometer to detect body heat from a cow from a distance of 304 m. Sir William Herschel’s son Sir John Herschel, used a device called an evaporograph and produced the first infrared image in 1840. From photo-conducting detectors in early phase, the thermal imaging technology has advanced a lot till date. Unlike earlier cumbersome cooled detectors, now a days much sophisticated, uncooled simple detector technology is available at a reasonable price in the market. Till 1960’s the thermography technology was basically used in military applications but after that, non-military applications like medical, industrial & building maintenance stated taking prominance. Today thermography has emerged as a proven predictive maintenance technology in the industrial domain.     

VIBRATION ANALYSIS TECHNOLOGIES:

VIBRATION ANALYSIS TECHNOLOGIES:

Amplitude Monitoring: We all know that all machines vibrate. But the question is what the acceptable limit of this vibration is? Basically it all depends on the machine type & its design. OEM gives a different vibration limit of alarm & trip for different machines. The different vibration technologies that can be employed for root cause analysis depends on the vibration transducer involved in picking up the data.

With contact type transducers following analysis can be performed-

1.    Amplitude monitoring & analysis of asset deterioration over a period.

2.    Frequency (FFT technology) - Raw time domain signal is broken into frequency domain signal by FFT technique (Fast Fourier Transform). As different frequencies are associated with a peculiar machine characteristic, by comparing the amplitudes of a good machine spectra & test machine spectra, it is possible to pinpoint troubles very accurately. For example 1X (one times the running frequency) frequency is responsible for unbalance, 2X for misalignment etc.

With proximity (non-contact) type transducers following analysis can be performed-

1.    Amplitude monitoring & analysis.

2.    Frequency (FFT technology)

3.    Phase: Phase is the position of a rotating part at any instant with respect to a fixed point. Phase tells about the vibration direction. Phase becomes very useful when the source of the vibration is not clear. Various useful insights regarding problems like Machine Soft Foot, Cocked Bearings and Bent Shafts, Imbalance confirmation, Looseness, Bending or Twisting, Shaft misalignment etc. can be identified by phase analysis.

4.    Form Analysis (Orbit Analysis): X-Y plane time domain signal plot: For performing orbit analysis, the input signals from the two proximity sensor placed at 90⁰ with each other on the bearing & a key phase (tacho) probe is viewed by the software & the resulting lissajous pattern is analysed. Orbit plot gives visual graph of actual shaft central line movement in bearing housing. With accelerometers & velometers also orbits can be achieved and as the sensors are mounted outside the casing of the machine, these orbits are called case orbits and provide absolute shaft motion with respect to space. But the orbits taken from proximity sensors are more common & useful. The shape of the orbit tells about the nature of the machine fault.

5.    Position: Position monitoring is done with the help of non-contact type of probes & various parameters like shaft eccentricity, Axial shift, housing & rotor (shaft) expansion & valve position can be monitored (A part of TG TSI system). With the help of X-Y probes the average shaft centreline can be plotted & rotor lift can be seen. These parameters prove valuable information regarding present condition of the machine.

6.    Bode Plot: Bode plot contains two graphs- Amplitude vs machine speed & Phase vs machine speed. It gives valuable information regarding amount of run out associated with a proximity probe, balance condition & system damping.

7.    Polar Plots: In polar plots also same variables are used as in bode plot.  They also serve the same purpose, only the methid of representation is different.

ONLINE & OFFLINE VIBRATION SOLUTIONS

ONLINE & OFFLINE VIBRATION SOLUTIONS

Vibration monitoring is just like checking the pulse of the patient. Now two methods can be adopted based on the condition of the patient. One can be a periodic scanning of the pulse & other can be continuous pulse monitoring in ICU. Similarly for the machines which lie in less critical zone, periodic vibration scanning system is advisable. Machine which are critical with respect of safety, environment & directly affect the production, online vibration monitoring systems are advocated. Following is the broad criterion for implementation of different technologies-

1.    Online System: This system holds well for critical category of assets which have immediate impact on safety, environment & production for example turbines, compressors, critical pumps & blowers etc.

2.    Offline System: Less critical or semi critical assets are good targets for portable/offline technology. On these assets periodic scanning is good enough to judge the present condition.

ONLINE VIBRATION MONITORING: These are continuous vibration monitoring systems employed on critical rotating assets like turbines, critical compressors, blowers, pumps etc. Sensor & system selection depends on many factors like machine type, type of bearings, machine speed, machine components, machine elements etc. These types of systems act as protection safeguard against excessive vibration & also do the condition monitoring job if the analysis software & module has been included in the system. As far as turbines are concerned where bearing of interest is fluid film (journal), Bently Nevada System is one of the best systems that I have worked on. Bently system is really good in rotor dynamics due to its long experience in this field. On rolling element bearings SKF has deep understanding and does a really good diagnostic on them. CSI is also good solution in this area.

PORTABLE VIBRATION MONITORING SOLUTION: As already discussed portable or vibration scanning solution is good for less critical assets. These are the assets which can be periodically scanned or can be scanned when there is some unusual indication regarding performance of the machine. Normally amplitude monitoring & FFT analysis is done on these kind of assets. Other techniques like phase analysis & orbit analysis may also be employed where we have facility to tap the signal through BNC port. Again SKF & CSI are very good in this area. Bently Nevada Snapshot & ADRE are also useful tools. Recently some new players in the condition monitoring field like Fluke have also introduced some exciting solution in the offline vibration segment. They have introduced two models that I have used till now. One of them is FFT based vibration analyser with auto fault diagnostic & auto advisory feature. This is a good tool for the shop floor engineers / technicians who have limited knowledge of vibration. The system is easy to use & interpret the faults.

Disclaimer: The discussed solutions are based on my own experience & are my personal views.

Vibration Monitoring & Analysis- Sensor selection Criterion

Sensor selection criterion:

Sensor selection is one of the most important things in vibration monitoring / Analysis system. Basically it’s the quality of vibration data input that decides the accuracy & relevance of the measurement. No matter how advanced electronics & complex analysis algorithm is used, the sensor being the first element in the line of measurement, has high importance in determining the quality of measurement. The mantra is gold in gold out, garbage in garbage out. Various factors and operating conditions like ambient temperature, magnetic field interferences, g range, frequency range, electromagnetic compatibility etc. decide the ruggedness of sensors in the field. Other machine parameters like type of bearings (Rolling element or journal) & machine speed decide the choice of measurement units & kind of sensor to be put in the field. Once the proper sensor selection has been done, the proper installation & sensor orientation becomes very important.

DISPLACEMENT SENSORS: These are also known as proximity sensors or eddy current sensors & are used to measure relative shaft vibration, shaft position and clearance. These probes are put on sleeve or oil film bearings (Journal bearings). These sensors are best suitable to measure low frequency and low amplitude displacements. If there is not any viability of mounting proximity probes on the machine, the accelerometer with double integrator circuit is also used for the displacement measurement. Normally in turbines & large compressors we find these probes.

VELOCITY SENSORS: Unlike proximity probes, velocity sensors are contact type sensors. Earlier electromagnetic sensors were used but due to mounting constraints & change in sensitivity issue with respect to time, now much rugged & sturdy piezoelectric sensors are used. These are accelerometers which are integrated once to get the velocity output. These sensors are used for low to medium frequency measurements (approx. up to 5000Hz). Majority of the machines lie in this frequency range (low to medium RPM) for vibration monitoring & balancing operations.

ACCELEROMETERS: Accelerometer is made of piezoelectric wafers & produces emf when there is force applied to it. These are the most preferred sensors for measuring vibration & have a very wide frequency range (almost from DC to 20 KHz). They are very useful for high speed machines & rolling element bearings. These are rugged devices & can sustain hash ambient conditions like corrosive environment & extremely high temperature (e.g. gas turbines). The sensor & associated electronics (charge amplifier) can be separated for high temperature applications. They have very good signal to noise ratio.   

 One should ask the questions like application frequency, type of bearing, environmental conditions, machine type, sensor size & mounting constraints before selecting suitable sensor.

Vibration Monotoring & Analysis- An important condition monitoring tool

Different technologies can be applied for the maintenance of different category of asset. We know that (Moubray’s PF chart) vibration is one of the earliest indicators of start of machine failure. In this section we will cover the vibration monitoring & analysis technology. Again human analogy becomes relevant to the machine also. Machines are like human body. The human body has complex balance system to maintain the internal environment of the body to a predetermined level. This is known as homeostasis, for example it maintains the body temperature around 98.6° Fahrenheit. Whenever there is some problem or a breakdown in this regulatory system, this control mechanism is disturbed. This is reflected in terms of rise in temperature, shivering etc. Similarly, if there is some mechanical / electrical problem or a process disturbance, the machine also expresses the problem in form of different variables like rise in temperature, vibration, change in electrical parameters etc. Since we know that vibration is one of the earliest indicators of the machine fault, it becomes important for us to use the available technologies for early detection of machine faults.

Vibration: Everything in the nature, either static or dynamic vibrates. So, question is what makes all these things to vibrate? The answer lies in the basic fundamentals of Physics. Atoms are the basic building blocks of the matter and they have inherent property of vibration. Vibration is the harmonic motion of a machine or machine part in either side of its neutral or stationary position & the response of a system to some internal or external excitation / force applied to it. So, the question is why different machines vibrate differently? The answer lies in the machine design & fundamental conditions like difference in mass, stiffness & damping. Even the same model & same design machine behave differently under similar conditions. God has created human beings & all are different from each other. Even two identical twins have some difference & they behave differently. Similarly no two machines can be same, even if all the design parameters are same & this follows law of nature. This distinction makes the machine diagnostic discipline not only a pure science but also an art which can be developed by the detailed study of the discipline & wealth of experience on different set of machines.

Units of Measurement: Vibration can be measured into three units-

 

1.    Displacement: The total distance travelled by the vibrating part, from one extreme limit to the other. This can be explained by simple spring mass balance where displacement is equal to the peak & bottom position travelled by the mass. The unit of displacement is microns (Pk-Pk).

2.    Velocity: Velocity is the speed at which displacement occurs. Since the seed is changing constantly, peak or RMS velocity are usually selected & the unit is mm/sec (Pk / RMS) & ips (inches per second)

3.     Acceleration: Acceleration is the rate of change of velocity.  At the extreme limit of travel of the vibrating part, acceleration is maximum or peak & most popular unit of acceleration is - g’s (peak).

 Relation between Displacement, Velocity & Acceleration

Integration of Acceleration =   Velocity, Integration of Velocity = Displacement

Machine Condition Monitoring Technologies

In the first section we had an overview of the different maintenance strategies in industrial settings. We saw the selection of maintenance strategy is being driven by the business objective of the organization. We also saw that not one strategy alone is suitable for all the assets in a plant. So, the question is why we require the support of technology for condition monitoring of the machine? I started my career around sixteen years ago with a power plant, in maintenance department. At that point of time I could find many experience technicians & engineers who were having very good common sense in judging the condition of the machine. Many operators & technicians joined the plant year’s back & were looking after the O&M of same machines till that point of time. They were so much acquainted with the machine behaviour that simply by looking at the machine, by hearing the sound & by gentle touch of different components, they come to know whether everything was OK or there was some problem. ‘LLF’- (Look, Listen & Feel) was part of their daily routine. At that point of time we were very new to the industry, so, we used to be amazed with their ability of judgement by employing a method which looked so crude to us. But the reality was that they were accurate in their judgement up to a great extent. Simply by hearing they can detect the probable fault in the machine and a gentle touch was sufficient to tell the surface temperature & vibration.

 

So, the question was that whether it was art or a science? To me it was a combination of both. Earlier the operator or technicians used to work on same machine for a longer period of time, may be their complete working life on few selected machines. They were emotionally & spiritually connected with those machines. If we see at sub-atomic level, machine is like a child who can’t speak but expresses its feeling, when it’s in trouble or there is some problem. It expresses its pain in form of different process variables like vibration, temperature, sound & drop in efficiency. The operator or technician was like a mother for that machine, who instantly recognizes the fault in machine by judging these parameters (to us it was crude at that point of time) because of sub-atomic level connect with it.

 

Today the condition has changed completely. First factor of the loss of this emotional connect with the machine is the shortening of duration of work of the operator on the machine due to frequent job changes as now more opportunities of growth are available due to rapid industrialization & management job rotation philosophy. Second factor as per me is the technology itself. Now advanced condition monitoring technologies are available, so, people rely more on technology rather than developing that connect by themselves. So, different advanced condition monitoring technologies are enabling the O&M people to effectively maintain their machines by employing suitable technology as per management’s business objectives & asset classification. In the next section we will look into these different technologies, their benefits & shortcomings.