Detection Of Fault Through Vibration Condition Monitoring

Published Date: 02 Nov 2017

INTRODUCTION

Background

In olden days, the maintenance on the machine is conducted only when the machine is broken or not functioning. The worker should stop their work because the machine had stopped and require those mechanics come and fix the problem. This is so call break down maintenance. It is costly because usually when a machine breaks down, the machine is already in a serve condition. The maintenance cost is very high and production need to be postponed and also leads to loss in rate of production. Shortly thereafter, people plan to maintain and refurbish the machine before it breaks down. This is known as periodic maintenance. This kind of maintenance is usually done based on the schedule. The purpose is to make sure the machine is in good condition before the next maintenance. When the machine is serviced regularly, it will not break down so often and the cost of maintenance will be reduced. Although the periodic maintenance is better than breakdown maintenance, but the approach is not so effective in terms of maintenance cost. People tend to come out with another method that is condition monitoring. Condition monitoring had made good approach in maintenance because the condition of the machine is continuously monitored. Thus, the maintenance is carried out based on the condition of the machine which tends to reduce the unnecessarily opening of equipment and also unnecessarily maintenance.

There are various kinds of condition monitoring such as thermograph monitoring, infrared monitoring, ultrasound monitoring, visual monitoring, and vibration monitoring. However, vibration monitoring is the most effective among the condition monitoring method. Most of the defects faced by the rotating machinery give rise to a distinct vibration pattern so call as vibration signature or vibration spectrum. Each kind of defects has their specific vibration signature. Therefore, vibration monitoring is able to record and identify the vibration signature which makes the technique so useful for industries nowadays to monitor the rotating machinery. In short, vibration monitoring has the ability to tell how serve the condition of the machinery in a quantitative way instead of qualitative way. Quantitative means the situation is described in a numerical way however the qualitative is describing or an illusion which used to describe how serious is the defects in the machine. Therefore, qualitative way is more convincing to show people the actual figure compare to the quantitative way.

Through vibration analysis, the vibration signature or the vibration spectrum can be shown and identified. Through the vibration analysis, people can know the condition of the machine is in what stage and by comparing the signature with the benchmark, people is able to predict when the break down will occur and schedule the maintenance program before the machine is serious broken down.

Condition monitoring had made a good contribution to identify and diagnose various type of deterioration in plant machinery, so that pro-active maintenance can be performed and reduces the maintenance cost. Vibration is found everywhere in machine. There is no perfect machinery is without any vibration. Every rotating machinery will vibrates because it arises by the machinery faults such as bearings fault, unbalance, looseness, foundation settlement problem and misalignment in coupling. Vibration is undesirable because it reduces the efficiency and performance of the machine.

People tend to reduce the vibration through maintenance and therefore reduce those machinery faults to the minimum effect. Most of the tragedy happens are causes by vibration. When the subject is vibrating with its natural frequency, it is vibrate in highest amplitude. This situation is known as resonance. Thus, extreme vibration can result in equipment damage and catastrophic failures.

Vibration condition monitoring system can also be utilized profitably in industrial applications. Through the monitoring, further damages can be avoided and the cost of maintenance can be reduced. In Malaysia Liquefied Natural Gas (MLNG) Bintulu plant which is one of the largest natural gas producers. It is expected to export 21.85 million ton of liquefied gas per annum and the production is increasing every year. Petronas is making a profit about RM55.6 Billion per annum, while revenue increased by 26.9% to RM222.8 billion. Once there is any machine breakdown or accident, the plant will be barred from operation and the loss in profit margin is a huge amount. The plant will also not able to catch up the production rate due to the shutdown of the whole plant. Due to the delay and the production is not able to finish in agreed due date, the company will also be sue by client. So, vibration condition monitoring is the best parameters to precisely monitor the machine and able to predict the fault that is about to occur.

Heavy industries such as Petronas Company must comply with the statutory requirements of the Department of Safety and Health (DOSH) and the Department of Environment (DOE). Once the requirement is not achieved, the plant cannot be operated until the entire requirement is achieved. In order to give good image to the worldwide and also to attract more investor, the more safety man hour is reached, the better the company. Once accident happens, the safety man hour throughout the whole year will become zero. Besides, the serious breakdown of machine requires expensive maintenance cost. If more serious, the machine may not be able to repair again. Thus, the company will suffer large cost losses and also image of the company has spoiled. Other investors may not build confidence on that particular company anymore.

On other hand, vibration can be used to diagnose the equipment problems and provide guidance for people to schedule their maintenance activity. Through the analysis of vibration spectrum, people can predict the failure which is going to happen and instantly make maintenance to prevent it from damage. There is certain acceptable tolerance that machine can vibrate but when the vibration exceeds the allowable limit after a period, the machine be fatigue instantly.

Aims and Objectives

Aim of this project is to study the influence of Rotorkit machinery fault on vibration spectrum through vibration condition monitoring system developed using the application of National Instrument LabView software.

The objective of this project:

To determine the influence of unbalance on frequency spectrum.

To determine the influence of shaft coupling misalignment on frequency spectrum.

To determine the influence of foundation looseness fault on frequency spectrum.

To determine the influence of ball bearing failure on the frequency spectrum.

LITERATURE REVIEW

A literature review is used to study the present status of research about the condition monitoring. Condition monitoring is getting more important day by day because people uses condition monitoring to increase the reliability of the machine, decrease the delay in production, reduce machine breakdown and maintenance cost. There is quite common that vibration is set as a parameter for condition monitoring. Diagnosis of faults such as unbalance, shaft misalignment, bearing defects, foundation settlement problem, and also foundation looseness is possible by comparing the signals of the machine with the normal acceptable standard condition. The signals can also be detected while the machine is still running. This is call online monitoring method which does not require the machine to shut down for monitoring process. This tends to reduce the machine down time. some of the recent research are discussed in the coming paragraph.

R.K BISWAS, [1] Scientist and head, condition monitoring group, CMERI, DURGAPUR presented a paper on "Vibration based condition monitoring of rotating machine" mentions that condition monitoring means the collection, comparison and storage of measurements that defines the machine condition. People will mostly recognize the fault of the machine but the problem is the fault is realized sooner or later. Therefore, condition monitoring is to diagnose the fault that is going to happen so that there are ample time to have maintenance before it is too late and minimize the disruption to operation and production. In this case, vibration is the best operating parameter to judge the dynamic condition of the machine. Vibration condition monitoring is a screening process which the vibration spectrum is compared to pre-established norms in order to recognize the fault.

Machine will break down without any warning. Sometimes a small tiny problem which people always ignored will leads to serve damage. The signs of impending breakdown are present before the catastrophic failure. Vibration signals is able for early indication of machine impending faults including those impending faults of machine like bearing defects, unbalance, shaft misalignment, looseness and more.

David Clifton, [2] St. Cross College, December, 2005 had a research on "Condition Monitoring of Gas-Turbine Engines". This research is mainly about condition monitoring approaches for modern gas-turbine aircraft engines. It also outlines the plans for novel research for the machine learning techniques in condition monitoring of aircraft systems. Therefore, the framework of condition monitoring on aircraft engines is done through the using of signatures from the engine vibration across various engine speeds in order to measure the engine health. Besides, inter and intra-engine monitoring are shown in which the engine normality model is build by using the vibration data from other engines of the same class or the test engine.

A Ramachandra, S B Kandagal, [3] research on "Prediction of Defects in Antifriction Bearing using Vibration Signal Analysis". In this paper, it mentions that using vibration analysis for condition monitoring of antifriction bearings in rotating machinery is a very well established method. It is because it reduces the down time and improving the efficiency. The machine can be monitored online without shutting down the machine in order to get the vibration signature. Besides, the researchers want to prevent the machine from further damage caused by bearing failure, the researcher developed bearing condition monitoring techniques such as temperature monitoring, wear debris analysis, oil analysis, vibration analysis and acoustic emission analysis to identify existence of flaws in the running bearings. From the result, they conclude that vibration analysis is the most commonly accepted technique due to its quantitative analysis and also ease of application.

Sadettin Orban, Nizami Akturk, Veli C, elik, [4] worked on "Vibration monitoring for defect diagnosis of rolling element bearings as a predictive maintenance tool: Comprehensive case studies". Vibration monitoring is a quantitative analysis because it offers information about anomalies which rises in the rotating machinery. The research presented the vibration monitoring and analysis case studies in the machineries that were running in real operation conditions. Failures are to be detected in early stage in the machineries on the early stage through spectral analysis. The results had shown that the predictive maintenance technique can be done through vibration monitoring to predict the impending fault.

Cornelius, [5] Scheffer, researched on "Pump Condition Monitoring through Vibration Analysis". Vibration is a powerful parameter for condition monitoring of machinery especially applied to rotating machinery such as pump and turbine. Recently, different kinds of vibration-based techniques had been developed and improve the cost-effective monitoring of pump operation and also the tackle of defectives problem. This paper mainly discuss about different kinds of vibration-based condition monitoring techniques for pumps. At the same time, those techniques are also able to improve the pump efficiency. It also states that there are specific factor to be considered when taking the vibration signature measurements for example where is the strategic place to take readings, which type of probe is to be used, what frequency range is suitable and what settings on analyzer should be.

Seema Nagrani, [6] Prof S S Pathan, Prof, I H Bhoraniya had a research on "Misalignment fault diagnosis in rotating machinery through the signal processing technique – Signature analysis". This paper mainly explained how the researchers using signal analysis for fault detection of misalignment. The fundamental of signature analysis and signal processing for identifying the spectrums for misalignment fault is discussed. The time domain analysis mainly focuses on statistical characteristic of vibration signal such as peak level, standard deviation, skewness and crest factor. However frequency domain uses Fourier to transform the time domain into frequency domain. In the paper, it is intended to apply fault to the machine in order to identify the spectrum of the defects. Thus, the spectrum can be produce through Fast Fourier Transform (FFT). The FFT spectrum will then be analyzed and compare with the normal operation conditions.

M. Kotb Ali, [7] M. F. H. Youssef, M. A. Hamaad, Alaa A. El-Butch worked on "Fault diagnosis by vibration analysis at different loading and speed conditions". In the paper, vibration is measured and analyzed because of the possibility of detecting severity of the defect, and hence predict the machine’s failure. Faults had been simulated intentionally to the dynamometer testing machine and the corresponding Fast Fourier Transform (FFT) spectrum has been recorded. Different load and speed is applied to the machine respectively. As a result, the vibration amplitude is affected by the variation of the load and the speed. The measuring position is important for a good maintenance vibration monitoring.

Stewart, [8] in much the same way as Smith [3] and Taylor [0] has examined how vibration condition monitoring assist in machinery fault diagnosis. Basically Smith research on the general kinds of faults and show a qualitatively result on how the faults can be recognized from their vibration characteristics. On other hand, Stewart and Taylor purpose about information on how actual measured data should be processed and analysis in order to perform the next step which is machinery fault diagnosis. Based fast Fourier transform analyzer is good in spectrum interpretation. They have constant bandwidth and in zoom or extended lines of resolution, they have high resolution in any frequency range. Fast Fourier Transform (FFT) analyzers also provide diagnostic tools such as synchronous time averaging, cepstrum analysis, peakness analysis, and use of Hilbert transform for the demodulation of amplitude and phase.

Sekhar, [9] Prabhu worked on "Effects of coupling misalignment on vibrations of rotating machinery" mention about how vibration induced by coupling misalignment on rotating machinery. Author also discussed that shaft misalignment can be major cause of vibration. Generally, 2X vibration response is the feature of bearing misalignment. In the research, a finite element model of rotor coupling bearing system was developed and effect of misalignment was introduced through a coupling coordinate system. The model agrees with empirical results, that 1X response is not nearly as significantly affected as the 2X response. By using this model, it is able to predict the vibration response caused by misalignment at various harmonics.

Fault diagnosis on mass unbalance tends to move from depending on art of human in interpretation of changes in parameters to computerized detection and analysis. Hocine Bendjama, [10] Salah Bouhouche, Mohamed Seghir Boucherit researched on "Application of wavelet transform for fault diagnosis in rotating machinery". They had purposed an applictation of wavelet transform for fault diagnosis in rotating machinery. This article mentions that wavelet transform is important for signal processing in non-stationary vibration measurements obtained from accelerometer sensors. It is evaluated through the experiment data in case of mass unbalance. Wavelet transform is a time-frequency analysis. Because of the time frequency domain, it decomposes a signal in both time and frequency in wavelet form. Wavelet transform can also divided into Continuous and Discrete Wavelet Transform. So, in this article, the researcher used Continuous and Discrete Wavelet Transform to test on real measurement signals collected from a vibration system containing mass unbalance and gear fault. They are able to diagnose the faults and improve conditions monitoring as well.

I. Ahmed, [11] M. Ahmed, K. Imaran, M. Shuja. Khan, T. Akram, M. Jawad, purposed on the paper "spectral analysis of misalignment in machines using sideband components of broken rotor bar, shorted turns and eccentricity". A similar research about spectral analysis of misalignment in machines is done using sideband components of broken rotor bar, shorted turns and eccentricity. This article is mainly about inspection the misaligned motors using diagnostic medium such as current, flux and instantaneous power spectrum. They found out that the side bands for a healthy and misaligned motor are different in amplitudes at different load condition especially in flux and instantaneous power spectra. Variations in the side bands are visible as compared to current spectrum and it is sufficient for detection of misalignment in a motor. However, the most efficient way for detection of misalignment comes to the use of eccentricity fault frequencies from flux signal.

Vaggeeram Hariharan, [12] PSS. Srinivasan, studied on "Vibration analysis of parallel misaligned shaft with ball bearing system". An experiment is conducted by help of vibration analysis to diagnose parallel misaligned shaft with ball bearing system. The study were conducted on a rotor dynamic test apparatus which is able them to predict vibration spectra for shaft misalignment. During the study, vibration accelerations were measured and obtained experimental and numerical frequency spectra. Experimental and numerical frequency spectra were analyzed and both vibration spectra show that misalignment can be characterized primarily by 2X shaft speed. In some circumstances, misalignment does not show up in vibration spectrum. So, it needs to be amplified and a high acceleration level at 2X shafts running speed is pronounced in frequency spectrum.

METHODOLOGY

Maintenance Strategies

Maintenance is meant for improve the efficiency of machinery and also improve mechanical performance before its breakdown. Any system must undergo optimum maintenance that can be performed by an organizational setup. It is also closely related to the industrial profitability. Maintenance can be classified into 3 categories that are breakdown maintenance, preventive maintenance and predictive maintenance.

Breakdown Maintenance

Breakdown maintenance provides the replacement of defective part when the machine is totally incapable of further operation. It is easy to implement and it doesn’t requires initial costs on training personnel and other upfront costs.

Disadvantages of breakdown maintenance are:

Failures are unpredictable.

Repairing cost is expensive since the machine is allowed to run till incapable. If the condition is serious, the industry can run away from the total replacement.

Loss in production and also the maintenance is timely. When the machine breaks down, the production line is forced to shut down for maintenance. During the shutdown, production is stopped and need to wait for people to come and repair.

Reduces the life span of the machine.

Preventive Maintenance

Preventive maintenance is a scheduled maintenance where the maintenance will be performed after some specific time of operation hour without considering the machine conditions.

Advantages of preventive maintenance:

Less damage on the machine due to often maintenance, less catastrophic failure will occur.

Downtime of the machine reduces up to 80%.

Less expenses of overpay due to less machine downtime.

Machine life expectancy increased.

Maintenance cost reduces due to less serve damage encounter by the machine itself.

Employee’s safety improved.

Disadvantages of preventive maintenance:

Too often maintenance of the machine requires often dismantling of the machine too. Therefore, each time of the dismantling will causes damages to the machine parts.

It is a time consuming maintenance since periodically maintenance requires the machine to shut often for maintenance

Imprecise scheduling ends up lead to unnecessary inspections plus even a healthy machine after maintenance will cause performance drop due to carelessness during dismantling

The effective scheduling of maintenance is very difficult. If the maintenance is schedule according to pass record, it may not be accurate and ultimately may lead to break down maintenance.

C:\Users\Joshua\Desktop\LR\bath tub curve.jpg

Figure 3.: Reflection from Smooth Surface

Preventive maintenance alone cannot eliminate breakdown because the equipment failure rate changes with the passing of time. From Figure 3.1, when the equipment is still new, it may have a high failure rate because of the design carelessness and manufacturing defect. However when the equipment is operated for several period of time, the failure rate will increase since the equipment reaches its lifespan or facing fatigue problem. Therefore, the preventive maintenance is difficult to be scheduled.

Predictive Maintenance

In the predictive maintenance, the condition of the machine is determined first then just the prediction can be done. In other words, prediction of the maintenance is according to the condition of the machine. The ultimate goal of predictive maintenance is to perform the maintenance in the most cost-effective way and just before the equipment lost its capability.

Condition monitoring can be done online or while the machine is still operation. Thereby, it reduces the distortion of normal system operations. Predictive maintenance is a tremendous cost saving and high reliability method hence the failure is being predicted and maintenance is performed before the failure of the machine.

Advantages of predictive maintenance:

When the failure can be predicted, maintenance can be scheduled at any time which is convenience before the failure. Thereby, shutdown can be done at convenient time.

When the maintenance is scheduled, it will ease the management department to manage their work force or prepared for mobilizing men, tools and ordering for necessary spare parts for the scheduled maintenance.

Costly trial and error maintenance can be avoided when the failure is predicted.

Machine is allowed to run continuously without unnecessary shutdown unless the condition of the machine is bad.

Disadvantage of predictive maintenance:

Skilled labor or professional is required to supervise the condition of machine.

Predictive maintenance Program:

Figure 3.: Program Flow of Predictive Maintenance

The Figure 3.2 above shows the flow of the program in predictive maintenance. First of all is the detection of failure. The machines are selected and identify the bearing locations and as well as weather the shaft is the motor drive end or motor non drive end types. Then the spot for taking vibration measurement is determined. Data is collected every fortnight or monthly. The source of vibrations is identified through trending and interpreting the vibration measurement.

In the analysis part, after the source of vibration is identified, it is then analyzed to pin point the causes for the vibrations to occur. In other words, the defect in the machine is determined through the analysis of the vibration signature. When the defect is successfully determined, then the process will proceed to correction part. At this part, skilled personnel will open up the machine and inspect the machine at convenient time. Thus, a necessary maintenance is done to eliminate the defects. Lastly, the confirmation part is to reassess the condition of the machine to make sure the defect is totally eliminated.

Condition Monitoring

Condition monitoring is the monitoring for the machines condition and its rate of performance. Condition monitoring can be precisely done by selecting a suitable parameter for the diagnosis of the defects. The parameters are recorded at intervals either on routine or continuous basis. After that, the data is collected and analyzed to give a warning on bad condition of the machine. Condition monitoring involves regular inspection using sensor and simple aids of sophisticated instruments. It is preferably done during the machine is still operating.

Condition monitoring can be a test for quality assurance as well as a procedure in order to the production to continue the operation. It can protect the company assets from damages by monitoring their condition and also to confirm the parts are installed and maintained properly. Condition monitoring aim is to detect the defects which lead to catastrophic breakdowns and loss of service. Thus, the defects can be tackled through a proper maintenance and fine tuning of machine which will increase the production and operating efficiency and also reduce the spare parts inventory. In condition monitoring, prediction of probabilistic deterioration can be highly eliminated thus prolong the lifespan of the machine by reduce the effect of failure.

Condition Monitoring Techniques

Visual Monitoring

Visual monitoring is using human visual to detect and inspect the failure on the machine body. It is applicable only for the defects such as surface cracks, oxide films, weld defects, and presence of potential sources such as sharp notches or misalignment.

Contaminant/debris Monitoring

Debris analysis is to analyze the foreign particles present in the sample. Sample such as the gear box are analyzed to find out the amount of foreign particles present in the sample. When there is present of foreign particles same as gear material, then there is probably having a gear wear. On other hand, oil debris analysis is different from debris monitoring because oil analysis is taking the machine oil to see whether the oil is in good quality after a period of operation.

Performance and Behaviour Monitoring

Performance behavior monitoring involves the checking of the performance of the machine see whether it behaves normally or abnormally. For example, the bearing’s temperature is measured to see whether it exceed the abnormal operating temperature.

Corrosion Monitoring

Corrosion monitoring is applied to the material which is aggressive on the plane of the machine to observe the rates of the internal corrosion or the degradation from the inside out on the body of machine. All this steps are to understand the rate of corrosion and actual corrosion process so that proper controlling of corrosion can be done.

Thermograph

Thermograph also known as thermal imaging which is using a thermal cameras to detect the radiation in the infrared range and produces a images of the radiation called thermograms. It allows one to differentiate the variations in temperature. When a machine is operating in abnormal temperature, the heat that produced from that particular component shows that there is a problem in that component. Therefore, maintenance can focus on that component.

Sound Monitoring

Sound monitoring is usually dong by the human operators through their sensitivity to the changes of sound during the operation of the machine. Once the sound is different from the normal machine, human operators with their experience will able to know what there is a defects such as looseness of component or the unwanted sound produced from the improper fastening. Recently there are also equipment for sound detection such as the micro phones with piezoelectric moving coils and condensers.

Shock Pulse Monitoring

Shock pulse monitoring is a unique technique that detects the pressure wave produces by the instantaneous mechanical impact.

Vibration Monitoring

Vibration monitoring is a best way to monitor the condition of the machine because it is able to detect the failure when it is still at the early stage. It is measuring the amplitude and the frequency of the vibration hence, the defects can be interpreted from the measurements or the vibration signature itself. Therefore, from the vibration signature, maintenance actions can be scheduled accordingly. Vibration monitoring can be widely used in the reciprocating machine and also the rotating machinery as every machine will generate their own vibration signature.

Raw signal obtained directly from the machine contains a lot of background noise which is impossible to analyze the precise information by simply measuring the overall signal. Therefore, suitable filter is needed to filter off the unwanted noise and also the environmentally contaminated signals so that a clear signals can be obtain and ease the interpretation job.

Vibration monitoring is an accurate method to describe the movement of the machine or the structure. Every defect such as unbalance and misalignment has their own signature or movement. Therefore it can be differentiate easily through the real time analysis of the vibration monitoring. Transducer is act as a sensor to generate vibration signal during the vibration monitoring. It is attached to the body of the machine to generate signal plus the data acquisition system to obtain the data generated from the transducer for further processing. Transducers can have many types such as electrodynamics, capacitive, electromagnetic, and piezoelectric transducer. Among all of the types, piezoelectric is most commonly used due to its accuracy.

Vibration monitoring is considerable importance among the monitoring techniques because:

All machines vibrate either smaller or greater extent due to the imperfectness in manufacturing or the mechanical defects that causes the vibration.

When the defects or inaccuracies of the system increase, the vibration will increase.

Each kind of defect has their own specific way of vibration signature.

It provides a quantitative data instead of qualitative data through the real-time analysis.

It can detect wide range of dynamic defects and also during the early stage of the existence.

Machine can be monitored continuously without shut down the operation.

Basic of Vibration

Vibration means the cyclic or oscillating motion of a machine or a component of it from the position of rest. In other words, vibration is an oscillating action of the machine structure from its equilibrium point. It may be random movement or in a periodic manner. Vibration is not totally undesirable but it is useful for some of the application. For example, the tuning fork that produces a tone during its vibration, a musical instrument such as the string of the guitar vibrates when it is strummed, loudspeaker diaphragm vibrates to produce sound and also mobile phone vibration to give people alertness.

In most of the case, vibration is undesirable because it waste the energy by reducing the efficiency, damages the component of the machine, leads to fatigue, or it creates sound pollution. In industries, vibration of engines, rotating machineries or turbine is typically unwanted. Those unwanted vibration usually causes by machinery faults such as unbalance, misalignment, mechanical looseness and more. Serve vibration will leads to damages and catastrophic failure of the machine. Therefore, people are implementing vibration condition monitoring to minimize the unwanted vibrations.

What Causes Vibrations?

Unwanted force induced and acts on the machine that causes vibration. These forces come from:

Force generated from unbalance rotation. Forces change in direction with time.

Force change in amplitude with time when an induction motor experience an unequal air gap between the motor armature and the stator.

Force may induce from the friction between moving and stationary machine components which similar as the violin strings vibrate when the rosined bow is slide on it.

Force comes from impacts between two machine parts such as gear tooth contacts or the rolling elements of the bearing bypass the flaw on the bearing raceways.

Force are also generated randomly as the turbulent flow of fluid that passes through the machine components such as fans, blowers and pumps, propellers and also turbines.

Characteristics of Vibration

When there is a vibration, it is actually a force involved that determine the characteristics of the vibration. The factors that affect the force are:

The exciting force from the mechanical fault such as unbalance, misalignment, and etc.

The mass of vibrating component, denoted as M.

The stiffness of the vibrating system, denoted by K.

The damping factor of the vibrating system, denoted by C.

By putting all the 4 component together, the exciting force will causes vibration while the stiffness, mass and damping forces will oppose the exciting force and thus reduce the vibration.

Vibration also can be defined in various forms:

Frequency

Displacement/amplitude

Velocity

Acceleration

Phase

Vibration Frequency

Lets define period of vibration first. It literally means by how much time is required to complete a full cycle of vibration. For example, a rotating completes one full cycle of vibration in 1/60th of a second. Therefore the period of vibration is equal to 1/60th of a second. Although the period of vibration is a very simplified and meaningful characteristic that describe the cycle completed by a machine, however vibration frequency is even more meaningful compare to the period of vibration.

Vibration frequency is a measure of the number of cycles that complete in a specified period of time. In other words, vibration frequency is known as cycle per second (CPS). Frequency is related to period of vibration. Their relationship is

Frequency = 1/Period of vibration

In short, frequency is the inverse of the period of vibration. For example, a period of vibration is 1/60th of a second then the frequency will be 60 cycles per second (60CPS). Besides from express the unit as CPM, frequency can also be expressed in unit of Hertz (Hz). Actually CPS and Hertz is the same thing. 60CPS is also equal to 60Hz or put into words will be 60 cycle in one second.

Significance of Vibration Frequency

There are lots of mechanical faults that cause a machine to vibrate. In order to pinpoint the specific cause of the mechanical faults that causes vibration, detailed analysis of the vibration should be done. This is where the vibration frequency comes into handy. Vibration frequency is an analyzing tool because every fault will have their specific vibration signature.

The excitation forces that cause vibration are generally produced through the rotating motion of the machine’s component throughout the operation of the machine. These forces will change in direction and amplitude according to the rotational speed (RPM) of the machine components such as shaft or rotor. Therefore, most of the vibration problems will have frequencies that directly related to the rotational speeds.

Assume that a machine is rotating in 2000 RPM with a fan operating at 1500 RPM and the vibration signature shows that the vibration exist at the frequency of 1X of fan RPM. This indicates that the fan is the source of vibration. In other words, the vibration force or pulse will happen at the same frequency as the fan frequency (1500 RPM or 25Hz). by knowing this information will help to narrow down the possible cause of vibration.

Predominant Frequency – Predominant frequency is the frequency of vibration which has the highest amplitude or magnitude.

Synchronous Frequency – Synchronous frequency is the vibration that occurs at 1 x RPM.

Sub synchronous Frequency – Sub synchronous frequency is the vibration that occurs at less than 1x RPM. A vibration occurs at ½ x RPM is known as sub synchronous frequency.

Fundamental Frequency – Fundamental frequency is the first frequency normally associated with a cause or a mechanical fault. For example, a coupling misalignment may generated 1x, 2x, or sometimes 3x RPM. Therefore, the fundamental frequency is the 1x RPM.

Harmonic Frequency – Harmonic means an exact or a whole number of multiple of fundamental frequency. For simplicity, a vibration that happen at frequency of two times the fundamental frequency also known as the second harmonic. For example, misalignment has a vibration at the 2x RPM which also can be said as the second harmonic of the running speed (1x RPM).

Sub harmonic Frequency – Sub harmonic is the exact sub multiples of frequency (1/2 x, 1/3 x, 1/4 x, 1/5x RPM). A sub harmonic frequency can be known as sub synchronous frequency but in certain case sub synchronous frequency is not equal to sub harmonic frequency. For example, 1/2 x RPM is a sub harmonic frequency and it is the frequency less than 1 x RPM so it is also known as sub synchronous frequency. On other hand, vibration with frequency of 30% of the RPM is a sub synchronous frequency but it is not a sub harmonic frequency as the 30% is not exact submultiples of frequency. \

Vibration Amplitude

Figure 3.: Vibration Amplitude

Amplitude of a vibration waveform can be described in many ways such as Peak to Peak, Peak, and Root Mean Square (RMS). For a sinusoidal waveform as shown in Figure (a), relationship between peak to peak and RMS level are Peak to Peak = 2 x Peak value, RMS = 0.707 x Peak Value, and Average = 0.637 x Peak value. The RMS and Average values intend to show the "mea" level of vibration. These conversions only applicable to singular frequency time waveforms or a pure sinusoidal waveform. However, in real life vibration of machinery, the vibration of the machine is not in pure sinusoidal waveform. The vibrations of machineries are more complex and conversion cannot be applied directly to overall vibration readings. To get the Peak or RMS value, people have to select the measurement type read out from a vibration meter or analyzer.

Vibration amplitude means that how far or how much is the displacement of the component in a machine displace from its origin position during operation of the system. It also can use to describe the severity of the condition of the machine itself. The magnitude of vibration or how rough or smooth the vibration is, can be expressed by the vibration amplitude. The higher the amplitude it vibrates, the more serious of the defects encountered in the machine and vice versa. Vibration amplitude can be expressed in displacement, velocity, acceleration and also spike energy.

Vibration Displacement

Vibration displacement means the total distance that a component of the machine vibrates from one peak or one extreme to another peak or extreme limit. On other words, it also known as peak-to-peak displacement. Peak-to-peak displacement is measured in mils. One mil is equals to 1000 of an inch (1000mil = 1 inch). If the vibration amplitude is measured to be 100 mils, it means that the component vibrates a total distance of 0.1 inches from one peak to another peak. However in metric units, the peak-to-peak displacement is measured in micrometers (µm) where one micron is equal to 1000 of a millimeter (1µm = 0.001mm).

Displacement is measurement of the total distance traveled by the mass or the vibrating component. In short, it shows how far the mass travels back and forth with vibrations. Its mathematical expression of displacement of a simple harmonic motion is x=A sin (wt). The unit of displacement is in millimeters, mm. x is the displacement in time function, A is the maximum amplitude or the distance from neutral position to upper limit, and sin (wt) describes the oscillation motion which is in sinusoidal waveform.

Displacement can use to interpret the vibration because it is closely relates to actual distance moved back and forth by the vibrating component. The terminology of measuring the displacement is to ensure the dynamic motion or displacement does not exceed the predefine value, bearing clearance and more. Displacement can also affect stresses of the vibrating component.

Vibration Velocity

Figure 3.: Vibration Velocity

Vibration velocity is measure the speed of the mass when it is moving or vibrating during its oscillation. When the mass reaches the upper and lower limits, the speed is zero. The mass comes to the extreme point and it will stop before it changes the direction of movement and begins to move in opposite way. The mass will first accelerate start from the zero velocity at either the both extreme point and reaches its maximum velocity at neutral position, then decelerates when the mass passes through the neutral position and the speed reduce to zero again when it reaches another extreme point.

Velocity is defined as the rate change of displacement with respect to time. Commonly, it is represented in millimeters per second, mm/s. Put it into mathematical equation, known that displacement is x= A Sin (wt). Thus, vibration will be v = A w Cos (wt) is the differentiation of the displacement equation. Velocity can use to interpret vibration as it is measure of cyclic changes per second. It can use to ensure the consequently affects fatigue life of the vibrating component. It is known as a good fatigue life descriptor.

Machinery that operates for a long time period will be facing fatigue failure. Fatigue can be describe by how far an object displaced and the rate (frequency) at which the object is displaced. Therefore, the distance displaced by and object is simply measured of the displacement and the frequency is the measure of how frequent a component displaced in a given period of time. If it is known that how far a thing must travel in a given period of time, then it is easy for speed or velocity to be calculated. Thus measure of the velocity is also measuring the fatigue.

Fatigue = displacement x frequency

Velocity = displacement x frequency

Fatigue = velocity

Vibration fatigue or velocity is measured in millimeters per second peak. Displacement (mm) times with frequency (s-1) equal to the velocity (mm/s). Vibration velocity is the measure of speed at which a machine component is displaced or moving as it undergoes oscillation motion.

Vibration Acceleration

Figure 3.: Vibration Acceleration

Acceleration is defined as rate change of velocity means how fast the velocity increase or decrease during mass oscillation between upper and lower limits. When acceleration is maximum, velocity is minimum where mass is at its upper or lower limit points. Otherwise, when acceleration is minimum, the velocity is at maximum where mass is at neutral position. It is similar to a driving car that when the petrol pedal is applied to increase the speed from stationary position to maximum speed, the car is accelerating. Similarly, when the car brakes and slow down to stationary, the car is deceleration.

Acceleration usually expressed in units of g where 1g = 9.81 m/s2. Acceleration expressed in mathematical equation is the differentiation of velocity equation a = - A w2 Sin (wt). People measure acceleration because it relates to inertia in vibration and is also a measure of dynamic forces. From Newton’s 2nd law, Force = Mass x Acceleration shows that acceleration is inter-dependence of force.

Vibration acceleration is a very important of vibration characteristic that can use to describe the amplitude or magnitude of vibration. Acceleration is the rate change of velocity. When a component is vibrating within its equilibrium, the velocity will be varied. For example, the velocity at the equilibrium is always the highest as well as the acceleration. When the component moves away from its equilibrium, the velocity will decrease means the component is deceleration. At the instant the component reaches the highest extreme peak, the velocity will be zero as well as the acceleration. After that, the velocity will start to increase when it leaves the highest peak and moves toward the equilibrium. Therefore, the component starts to accelerate towards the equilibrium. However when it reaches the equilibrium, the velocity as well as the acceleration is the highest. Once the component passes the equilibrium and moves towards the lower extreme peak, the velocity decrease as well as the deceleration of the component happens. Until reaches the lowest peak, the velocity will become zero at that instant as well as the acceleration. These processes are continuous towards the whole vibration process. In vibration acceleration, the unit to express the acceleration is millimeter per second square (mm/s2).

Vibration Phase

Figure 3.: (a) Masses with 0º Phase Difference (b) Mass with 90º phase Difference

Phase is the time relationship between vibrations of the frequency. It is relative shift of one vibrating parts compared to a fixed reference point on another vibration part. In short, phase is measure of vibration motion on one reference relative to another vibration motion at another location. From figure 3.6 (a) shows that two vibrating mass are in phase (0º phase difference). Figure 3.6 (b) shows two vibrating mass has 90º phases difference because the mass No.2 has quarter cycle ahead of mass No.1. On other words, Mass No.1 has 90º phase lag relative to Mass No.2.

(a)

(b)

Figure 3.: (a) Masses with 180º Phase Difference (b) Shaft In Phas and Out Of Phase Condition

Figure 3.7 (a) above shows the mass vibrating with a 180º phase difference. Mass No.1 moves in opposite direction of Mass No.2. However, Figure (b) shows a shaft moving in same phase which is 0º phase difference between two ends. This is known as in-phase planar motion. On other hand, the below shaft are move in out of phase which is 180º phase difference between two ends.

Phase is measured because it can used to determine the time relationship between a force and the resulting vibration it causes. For example, force that result in mass unbalance and causes vibration and phase measurement is used to balance the unbalance force.

Vibration Data Analysis

Time Domain

Figure 3.: Time Domain Waveform

The vibration waveform basically is plot in the amplitude of vibration versus time changes over a complete cycle of vibration motion. This is in fact represented the vibration viewed from a time perspective and referred as the time domain. The simplest of time domain is a pure sine wave where time needed for a complete cycle is call periodic time, T (s). On other side, in one second the number of cycles completed is call cycle per second which is cycle/second = 1/T termed as frequency (Hz). Reality, real vibrations have a more complex waveform usually consist of a few sinusoidal waveform combine together and also difference phase. In fact, the ral waveform produced by adding sine and phase of sine waves correctly.

Frequency Domain

Figure 3.: Frequency Domain Waveform

The vibration signal waveform from a real world can be split into a series of pure sine wave. All the combination of sine waves is unique and different combination represents different condition of machine. The figure above shows a three dimensional graph of this addition of sine waves. Two of the axes are time and amplitude, which represents the time domain signal graph. Another axis is frequency which enables people to visually separate the sine waves which when add up gives back the complex waveform. Thus, the plot of amplitude against frequency is called frequency domain. Sine wave is separated in the frequency domain and appears as a vertical line in different frequency. The height represents the amplitude and its position along the frequency axis represents the pertinent frequency value (1/periodic time) of that particular sine wave. Spectrum is the frequency domain representation of the vibration signal. Each sine wave line on the spectrum plot is called frequency spectral component.

Figure 3.: Frequency Domain Analysis

Above figure 3.10 illustrates how the frequency domain analysis and also the series of pure sine wave from time domain applied to machinery. The rotor experience imbalance, a ball bearing defect, and reduction gear meshing and the fault can be analyzed from the frequency domain. Imbalance produces a sinusoidal vibration at a frequency of once per revolution while the single defect in outer race of ball bearing produces impulsive vibration each time a ball passes over the defect around four times per revolution. Two smaller sine waves around this frequency are influenced by the interaction or modulation of the bearing defect force with imbalance force. These signals are called sidebands, and often occur in machinery vibration. Other than that, the gear mesh frequency appears at running speed multiplied with number of teeth on the main shaft gear. There are also two small frequencies sidebands around the gear meshing frequency which usually caused by eccentricity in the gear. Frequency spectrum clearly shows that the machinery fault because different fault happens at different frequency in extreme amplitude.

Vibration Response to a Force

When a rotating machinery rotor experiences a heavy spot or simple imbalance, it will result in outward radial centrifugal force. The force is expressed as

F = M r (2pf)2

M is the rotor mass, f is rotational frequency (cycle/sec) and r is the distance of eccentricity of the heavy spot or eccentricity mass from the geometric centre.

Figure 3.: Radial Outward force in Reference Direction

The radial outward force referred to a reference point produces a forcing component in reference direction as sinusoidal variation. At points A and C where the imbalance force is right angle to reference direction, the resultant force in the reference direction is zero. On other hand, at points B and D the force is positive and negative maximum respectively because they are in line with the reference point.

Displacement Resulting From a Force

Figure 3.: Displacement Against Time Graph

When a rotor rotates with a heavy spot and simple imbalance weight, the rotor shaft will displace and centre of rotation moves away from geometric centre. The displacement is caused by the force which is proportional to the magnitude of the force. The two forces is not the same since the dynamics of the rotor will affect the force response. Vibration measured on a machine is the response to the vibrating forcing input ant not the force itself. Other than that, the vibration response measured on the machine casing or bearing housing is influenced by mechanical impendence which is also known as the inverse of dynamic resistance or the propagation path. The mechanical impendence will change significantly with speed.

When people usually measuring the response of the machine to vibration forces and not forces themselves. Mechanical impendence have a direct affection on the measured vibration. First, if the response is small, the vibration will be hard to analyze. Secondly, if the response changes drastically with frequency, changes in running speed can produce misleading changes in measured vibration level.

Natural Frequency

Figure 3.: Vibration Response Versus Frequency

A plot of vibration response versus frequency for machine housing shows how vibration level changes with rpm. A defect can be found from the frequency B because it produces a much larger vibration response than the same force level at frequency A. In figure above, the vibration response peaks occur at pertinent frequency values also known as natural frequencies. Natural frequencies are the structure vibrate naturally when experience an impact. Thus, resonance occurs when a vibration force occurs at natural frequency and produces a large amplitude vibration.

Three important areas that natural frequencies relate to machinery vibration analysis are first, resonances of the structure can cause changes in vibration level with rpm. Second, the dynamics of rotating shafts have big changes near the natural frequencies or critical speed. Other than that, resonances limit the operating frequency range of transducers and accelerometers.

Natural frequencies relates to mass and the stiffness. Express natural frequencies in equation would be natural frequency (wn) = (k/m)1/2. K is the stiffness and m is mass. This shows that when natural frequency goes up, the stiffness should be increasing while mass is decreasing.

Instruments for Vibration Detection and Analysis

Rotorkit Design

Coupling

Adjustable screws

Bearings

Adjustable screws

AC Rotor

Adjustable screwshttps://fbcdn-sphotos-d-a.akamaihd.net/hphotos-ak-ash3/552068_10151016571373037_1219049769_n.jpg

Bearing supports

Shaft

Adjustable screwsFigure 3.: Rotorkit

Rotorkit consists of several components such as AC induction motor, shaft, bearings and an accelerometer sensor. The design of the rotorkit is in such a way that the shaft is coupled with the AC motor by a coupling. AC motor will tends to rotate the shaft which is supported by two bearings. Bearings are mounted on two columns and the two columns are mounted on an adjustable platform.

https://fbcdn-sphotos-f-a.akamaihd.net/hphotos-ak-ash4/486599_10151016584708037_454976515_n.jpg

Adjustable screws

Adjustable screwsFigure 3.: Screw for Adjusting the Foundation Platform to Create Misalignment Phenomenon

There are screws which located at the corner of the foundation platform are purposely made to adjust the height of the platform as illustrated on Figure 3.11. When the screw is tightened or loosened, it will either move the platform up or down. This is meant to create a phenomenon which illustrates the shaft and the motor are not coupled nicely on an uneven platform and causes the misalignment fault.

Vibration Transducer/Probe

The most important vibration instrument is the transducer. Transducer is a sensor to detect the vibration signal. Without the transducer, vibration signal can be measured. In order to produce signal, the transducer is attached to the machine. When the machine vibrates, the transducer will also vibrate at the same instant. The transducer will transform the mechanical vibration into electrical signal that can be processed by the associated instrument to obtain the vibration amplitude, frequency and the phase.

There are various kinds of transducer. However, the transducer that provide standard with vibration meters, analyzers and also data collectors is the vibration accelerometer. Accelerometer is a self-generating device that can produce a voltage output that is proportional to the vibration acceleration. The sensitivity of the accelerometer is means how much voltage (mV) that be produced by the accelerometer per unit of vibration acceleration (G). Therefore, the unit of sensitivity is expressed in milli volt per acceleration (mV/G). Most of the accelerometers will have the sensitivities range from 10 to 100mV/G.

http://lukask.dyndns.org/files/datasheet/Sensors/Info-Technisch_Englisch/si/Technology_Acceleration-Dateien/f_tech_accel_text-Dateien/tech_accel_fig2.gif

Figure 3.: Basic Structure of An Accelerometer

Above figure shows a diagram of an accelerometer. An accelerometer can generate a voltage is because the piezoelectric material or known as quartz. It is a non-conducting crystal and when the material is mechanically stressed or compressed, it will generate an electrical charge. The greater the force applied, the higher the voltage generated. The mass inside the accelerometer will compress or sheared against the piezoelectric disks during vibration. The size and the amount of the piezoelectric material will determine the sensitivity (mV/G) and also its usable frequency range.

The amount of the electrical signal produced from the accelerometer is relatively small plus in many times when the small electrical signal is transmitted using a very long interconnecting cable to the vibration analyzer which is very far apart and until the end the small electrical signal will be even lesser. This is where the electric amplifier incorporates with. The amplifier usually sited inside the accelerometer so that the signal is amplified and then transmitted through long cables without worrying about the dissipation of signal or interference Horn radio frequencies or high voltage electro-static interference or high voltage transformers or electrical fields around motors. The amplifiers usually can be used with interconnecting cables up to 330 meters without experience signal loss or interference.

Figure 3.: Response Curve for Piezoelectric Accelerometer

Accelerometer operates below its first natural frequency. When the frequencies are far below the resonance frequency of mass and spring, the displacement is directly proportional to acceleration of frame. The sensitivity increase when resonance is approached. Generally, sensitivity of an accelerometer which is the ration of electrical output over input acceleration is acceptably constant to approximately 1/5 to 1/3 of its natural frequency from the mean value as shown in Figure 3.15. The frequency response curve can be influenced by mass, stiffness, and degree of system damping. The resonant peak of the accelerometer can be eliminated by increasing the damping. The increasing of damping will result in phase shift in linear range. Whereas for un-damped accelerometer has low phase shift until near the natural frequency. Therefore, un-damped accelerometers with high natural frequency are usually used because the linear range is extended the most.

Fast Fourier Transform, FFT

The vibration signal obtained from the accelerometer is in time domain or time waveform but the information is not useful as it cannot be used for analysis. Therefore, the time waveform needed to transform into frequency. Therefore, FFT is used to convert the time waveform into vibration frequencies present along with their amplitudes. FFT can be done manually but it Is time consuming. With the present of digital technology, Fourier transform can be done instantly as the term Fast Fourier Transform comes in. The digital analyzer and data collectors usually are a computer hardware that is programmed to perform the FFT function.

Data Acquisition Board NI PCI-4474

https://artisan-scientific.com/itemimages/NI_PCI4474_View1.jpg

Figure 3.: Data Acquisition Hardware NI PCI-4474

The National Instrument PCI-4474 shown in Figure 3.18 is a 4 channel dynamic signal acquisition device that is able to produce high-accuracy dynamic and transient measurements. The 4 analog inputs with 24-bits resolution analog to digital conversion; that can simultaneously sampled at software-programmable rate. PCI-4474 has a sampling rate at 102.4 kS/s with ±10V and analog and digital triggering. The high resolution makes PCI-4474 board suitable for dynamic and transient applications. Phenomenal measurement can be achieved with PCI-4474 board when compared with standard 12-bit or 16-bit acquisition. For example, sampling at 52.3 kS/s with 16 k FFT, a spurious-free dynamic range greater than 130 dB can be displayed. PCI-4474 can be synchronize with other boards using the RTSI bus, so that it can work together with accelerometers, strain gauges, tachometers, and proximity probes. In this project, PCI-4474 is used with LabView Sound and Vibration Toolset to obtain accurate time and frequency measurement.

Using delta-sigma modulating Analog to Digital conversions (ADCs), a low noise and low distortion can be achieved in PCI-4744 because ADCs use 1-bit quantizer oversampled at a multiple of sampling rate, and thus produce extraordinary linearity. Further on, extraordinary flat, linear-phase, lowpass digital filters will then remove aliases and shape the quantization noise, divide the sample rate by the oversample factor, and increase the resolution to 24-bit. Using delta-sigma modulation ADCs, PCI-4474 is immune to DNL distortion associated with conventional data acquisition device