 This is another lecture on non-destructive evaluation or testing of structures or machine components for fault detection and in this module we will be discussing on two techniques today in this class. One is on acoustic emission other being the eddy current testing, first I will discuss about acoustic emission. So, as the slide says you know all materials essentially if they are stressed they will give back some energy and this energy if it can be captured and analyzed it gives us clue as to what is the defect in this material. So, if I look at this definition or what this acoustic emission is, so if any body or elastic body is subjected to any kind of a stress it is going to release an acoustical energy and this energy is called as an acoustic emission event as if the material or the machine is crying because of a stress or load on it. And the peculiarity or the characteristics of this energy which is known as the acoustic emission is that it is an body is stressed whether you can be pulling it with some force. So, this gives out some sort of an energy in all directions and if this energy can be sensed by a transducer this gives a clue as to what is the defect or the condition. If there is no defect the amount of energy will be different if there is a defect this amount of energy will be defect. There are certain characteristics of this acoustic emission signal first of all it is in the megahertz range and then this is either in the form of continuous wave form or as in burst of certain energy. Now, the acoustic emission condition or the strategy is to have this transducer we should be able to detect this kind of stress waves and this is what we are going to look into this in this course and how it can be measured analyzed and so on. So, if I was to look into the applications of acoustic emission it is an non-destructive testing of heavily mechanically stressed components or complete structures of fiber and first plastics or composites as used in the aerospace industry. So, this can be used both for materials which are metallic and non-metallic. Acoustic emission can be used in materials research to understand the material properties and breakdown of the failure mechanisms of the damage behaviors and materials. Acoustic emission particularly in non-destructive testing or condition based maintenance is towards inspection quality assurance example monitoring of welding and wood drying processes, series inspection of ceramic components, scratch test and few more. This is used to detect leaks in large reservoirs or large containers it is used for seismic research and so on. This is one photograph of an acoustic emission testing been done on 25 meter length pressure vessel of diameter 4 meter this can be instrumented with pressure sensors or acoustic emission sensors and this signals can be analyzed to find out the source of the leaks. Like this example or this diagram shows I can be having many sensors around the location and each of the sensors will give us an acoustic emission signal. So, by the method of triangulation the source of defect can be identified ok, but as I was mentioning you these signals could be their time domain they are very high frequency time domain they could be continuous signals A signals or they could be in burst burst or continuous signals. Now, there is a little historical development in this acoustic emission systems because they are of the order of few megahertz ok. Now, to do any signal analysis if you will recall the class on computer aided data acquisition to do signal analysis on such high frequency signal. We always have to if you have to do DSP on the signal we have to acquire them through an AD process and then we can this is the acoustic emission signal we have to an AD process then once we have the digital data these are the digital data we can do some sort of an analysis on the computer. However, when DSP was not developed in the early days people used analog signal analysis processors and in particular because they are high frequency signals all we could do was at most find out the RMS level or the peak level or may be the time duration etcetera or the number of peaks in a given time. So, these were very gross parameters which were used in and I will show the details of these parameters this is how they were being done. So, basically it was a E transducer and then some sort of a signal conditional and then we have the display signal conditional and A E transducer ok, but this kind of instrumentation was good enough to detect no defect and defect and it gave us a qualitative feel of the defects in a component or machine and this is how a typical transient acoustic emission signal looks like mind you this is in microseconds. So, you can understand it is a very very high frequency signal and this has to be captured in a device you know this is an analog device ok from on which this has been captured like a digital storage oscilloscope ok and from this time domain signal certain features could be found out like number of peaks beyond a certain threshold, number of peaks occurring in a time period. So, all this could be done through a simple analog processor and as opposed to the transient signal which was mostly a burst this is what is known as the continuous signal there is some sort of an energy signal either this is in some millivolt we get in a particular time waveform of the continuous A E signal. There is another related effect in acoustic emission which is known as the Kaiser effect. Now, this is because of the fact that it was named after Dr. Kaiser which says that as an irreversible phenomena in the behaviors of material called the Kaiser effect and this says that this is the effect of the that act indicates that irreversible processes during plastic deformation in the strain which can be shown by using a visual lateral crystal and because of the Kaiser effect whenever the material has stressed beyond the earlier limit then only we will have a meaningful acoustic emission signal coming out of the system. For example, just to illustrate the Kaiser effect suppose I have a machine it has been stressed and there is a defect. So, I am getting some signal some acoustic emission signal. Now, in the next instance if it is stressed less than this whatever is the acoustic emission sensor signal it will not be realized till I come to a new level of stress which is higher than the previous signal and then only I am getting a high signal. So, this is essentially known as the Kaiser effect. So, if I was to look into the analog A E process chain. So, the machine or member is stressed and because of this there will be release of the elastic energy this wave will propagate and then this acoustic emission signal will be detected and then you can record them with number of sensors and then put them in a personal computer to analyze the signals. So, test object in application of load which produces the mechanical tensions, source mechanisms are basically they releasing the acoustical energy or the elastic energy wave propagation from the source of the defect to the sensors. Sensors basically convert the mechanical wave into an electrical signal. So, acquisition the converting the A E signal into electronic data set and you can either display it or store the data. So, in the modern methods of acoustic emission sensing there has been an A to D converter which had explained for mentioned few slides ago. This A to D converter are nowadays available. Say for example, if I have to sample an A E signal which is around 2 megahertz at least because of the Shannon's sampling theorem my sampling frequency f s should be greater than at least twice f max. So, at least greater than 4 megahertz in this case. In early days getting A D systems with such high sampling rates were a problem. So, acoustic emission was one way the digital signal processing for acoustic emissions was not very popular, but off late A to D systems with high sampling rates are available and they have been used in acoustic emission signal analysis. So, this is typically like I mentioned to you on the other class regarding ultrasonic transducer where I had a piezoelectric element which was made to resonate. Similarly, we have an acoustic emission sensor where in piezoelectric crystal is there and then this will be giving it will be sensitive to this surface the waves which are coming through the structure to the surface and then we will get a signal corresponding to the stress waves. And again because these are all capacitance and we are very careful regarding the noise and that is one of the important issues practical issues of AE is the noise or poor signal to noise ratio. Off late acoustic emission has mostly been used for a quite quick qualitative estimate estimate and location of the signal source of defect rather than unlike vibration where in we have cases like the signal I can be very sensitive I mean other I can say 1.2 meters per second square or 1.209 meters per second square they are two different quantities and this can be sensed by systems which is not true in the case of acoustic emission because of this capacitance and then we have to be very careful about this cables not be band strained and then this special cables which reduces the triboelectric noise are to be used in acoustic emission systems. And again this AE preamplifier has a low input noise to distinguish smaller sensor signals from electronic noise. It has a large dynamic range to process high amplitudes without saturation, large range of operating temperature for applications in neighborhood of low temperature vessels as well as above the transition temperature from brittle to ductile material. Usually a supply voltage of 28 volt for the preamplifier is given through a signal cable and then for the calibration pulse and all that will be required. Filter sets depending on the acoustic emission particularly for leakage and corrosion test the filters are within anywhere from 20 to 100 kilohertz. For metallic component testings we have filters from 100 to 300 kilohertz and above 300 kilohertz for small distance between the sensors. Certain terms which are used in acoustic emission I thought I should be mentioning it to you here in the class. So acoustic emission is a transient electric sorry elastic wave produced by the release of elastic energy during certain processes which could be mechanical stress. Event is this burst or if you see source of the origin signal is the signal which you get from the AE transistor. Signals could be a transient or a continuous burst. AE activities occurrence of an AE signal, AE intensity is the strength, AE channel is the number of such measurement sensors which we have and usually we require a multi channel system in machine structure so that by the method of triangulation we can find out the source of the acoustic emission. So some in the time domain we can set lot of triggering events for example this is the negative threshold with the positive threshold. So any number the number of peaks beyond a positive threshold can be used to quantify some of the AE signals like above a particular threshold and the first time the crossing has happened before the bureau arrival time. So all these kind of signals can be processed then the maximum amplitude number of overruns signal saturation so there are lot of features which have to be measured and monitored in such acoustic emission signals. So arrival time is the absolute time of first thresholding crossing peak amplitude, rise time, time between the first threshold and peak amplitude signal the recent time interval between first and last threshold crossing. Number of threshold crossings are the AE counts in one direction. So AE counts were earlier used as a method of detecting whether an acoustic event has happened or not. AE count increase in AE count represents a defect in the structure. So this AE can be calibrated by what is known as a pencil lead break test and this is one example where in you see how the principle of location is used. Sensors 1, 2, 3, 4 around a defect and you can see if the signal from the source defect is one depending on the time delay for the signal to come to one of these sensor locations people can have a 3 dimensional software I mean graphic software wherein they can triangulate around 4 spheres or maybe they use 3 spheres to triangulate wherever the 3 spheres meet will be if I can draw them in a 2 dimensional plane suppose one event is here you can draw a circle around it, circle around it and circle around it. So wherever these 3 spheres meet that will be the location of the defect from each of these sensors. So this kind of triangulation algorithms are there and this is just one such locations from a software which are used. So AE transducer by having multiple AE transducers and doing a triangulation the AE source can be determined. Though AE has been very qualitatively popular for knowing whether a defect has occurred and not occurred because as I was telling right in the beginning the material cries under stress and this cry signals are high frequency signals which has to be captured by piezoelectric crystals processed and through a software we can use them to quantify the AE event or even triangulate to find out the source of the AE event. However, there are certain limitations of acoustic emission one is defects which are neither growing nor moving do not produce AE and thus cannot be detected. According to the Kaiser effect only those defects are detected without exceeding the highest preceding load which are already active at the actual load level and are entering the end engineering the component anyway. So only by increasing the load above the previous maximum load level defects can be found which do not grow at standard load. Evaluation criteria do not exist in form of commonly accessible data like the rating of AE results is set firmly to the knowledge and experience of the service provider and AE's testing is very sensitive to noise threshold and noise cannot be stopped and so we have to be very careful about the certain signal to noise ratio and otherwise AE testing is not very deterministic if I may say so unlike vibration monitoring motor current monitoring or even ultrasonic or even the next topic which we are going to discuss is eddy current testing. But AE gives us a quick survey whether a defect has occurred because of stress or not and then we can sense it. So now I will move on to another related another topic in NDT and that is what is known as the eddy current testing. So eddy current of course the first and foremost thing which separates eddy current from ultrasonic testing which we have discussed earlier is eddy current testing is only possible with metals unlike ultrasonic testing which is both for metals and non-metals. So these are some of the applications of eddy current I can measure the metal thickness I can use to sort alloys, find out the heat treatment condition, damage because of heat treatment, the plate thickness or the cladding thickness, insulation thickness, non-metal plate coating thickness, certain vibrations particularly in eddy current proximity probes are used to find out the displacements in turbines or shafts, diameter of tubing or bar stock, cracks, seams, porosity corrosion, erosion, segregation, inclusion all these can be found out in metals by eddy current testing. So let us try to understand how and what is this eddy current testing. So the principle behind this eddy current testing is given in this slide. For example if I have a coil through which I pass usually a high frequency AC signal. So because of this signal being alternating there will be a magnetic field associated in this coil. Now what happens if I bring this coil because of its magnetic field around a conducting material, series of circular eddy currents will be formed. Now this eddy current will also have a magnetic field. Now what happens if there is a defect or change in thickness in this conducting material, the magnetic field produced by this eddy currents will be different and obviously because of Lenz's law this will oppose. So if some way we can detect this difference in the magnetic field because of this eddy current which could be different if there is an foreign inclusion, if there is a defect, if there is a change in thickness this gives us an indirect method to measure the thickness, the displacement of the coil etcetera from the conducting material and so on. So how is this eddy current used for finding out the cracks in a material. For example this is a test coil usually the primary is around 400 hertz or 400 sorry 40 kilohertz very very high frequency signal and then we will have a magnetic field because of this eddy current if there is a crack this magnetic field is going to change and then voltage corresponding voltage would change. Now there are certain relationships as to how close I need to bring this coil to the surface. So there is an relationship between the depth of penetration and given by the magnetic permeability and the frequency of excitation. So all these expressions are there, the eddy current strength is optimum to a certain depth. So eddy current density is greatest at surface and reduces exponentially with depth. So it is good to measure the surface thicknesses or change at the surface. So that is why lot of this coating thicknesses insulation thicknesses we can find out or surface defects or cracks on nice surfaces can be very easily measured through eddy current probes. So there is an effect of frequency if the frequency is high the primary coil in the test coil is frequency of excitation is high the penetration depth is pretty low and if the frequency is low the penetration depth is high and then you can see the proportionate intensity and depth. These are some things which have to keep in mind from this expression here where delta is the depth of penetration. Similarly the conductivity of the material does play a role if the poor conductor be eddy current intensity is different and then with the in a good conductor and so on. Effect of magnetic permeability you can see low permeability and high permeability you can see how the depth and the intensity vary in eddy currents. These are all very qualitative expressions but these equations are also there for different kinds of eddy currents. So what are the parameters by which this eddy current has to be can be influenced one is high frequency high conductivity high permeability one case another case we have low frequency low conductivity low permeability. So and to decide on the right probe for the eddy current the choice of these few parameters have to made the frequency of the primary signal frequency of the primary AC signal which is used to create the magnetic field then the conductivity of the metal and of course the magnetic permeability. So based on this there are different types of probes available if you go to any manufacturer there will be different types of probe depending on how thick material thickness you are going to measure depending on the type of material on which you are going to measure. So all these criteria will let you decide on the selection or let you decide on the type of probe to be eddy current probe to be used for such measurements. Now certain things are also to be noted. So when the suppose of the eddy current from the top view if the cracks are parallel to the eddy currents you will not detect any eddy currents because they are not interrupting the eddy currents. If they are interrupt like in this cross dimension you can see the eddy currents disturbed and then you will get a signal for that. By the way some of these details about this instrumentation can also be found in my website www.iitnoise.com and in our lab we have these measurement facilities both for eddy current, ultrasonic and so on. So you can visit this website to get an update on the instruments which we have. Certain edge effects are also occur. Sometimes edge produces signal just like cracks and misled the inspector. So different techniques can be used to avoid edge effects. So a lot of these NDT techniques you know if you think of thermography, ultrasonic, di-penetrant test, eddy current, acoustic emission a lot of this actually depends on the experience of the operator who is using them in the field to do the measurements. So in a classroom we can very nicely tell you the theories of these measurements but once you go out to the field to do this NDT test techniques many times you will be either misled by poor signal to noise ratios or because you do not have the right kind of transducer to capture the phenomena because if you have a digital signal processing followed by a display unit you know garbage N is garbage out. So we have to very very careful while doing measurements in the field using such techniques which are tested well tested and improved but always I would recommend you know any of you who are practicing NDT techniques for CBM to always do what is known as an in situ calibration of the test equipment and for such NDT techniques there are lot of ASTM standards both for eddy current testing, acoustic emission, ultrasonic, thermography etcetera even certain ISO standards are there ok which need to be referred and sometimes I am we have this kind of standards and at this website. So you can you can have the list of the standards and then of course you can get these standards in your local library or from their websites. So this has to be followed whenever you are doing such NDT techniques testing techniques for CBM. So there is some sort of an inductive coupling between the primary coil and the secondary magnetic field because of the eddy current. So this coupling is what we sense and then we use it. There is another parameter which we have to very careful as to how close should I bring this probe to the surface and that is something defined as the fill factor ok lift off and fill factor ok. This fill factor if you know the diameter D naught and D i. So this D naught square by D i square ok is what is known as the fill factor. Usually 70 to 90 percent of the fill factor is target for reliable inspection. So if you lift off ok this this diameter and this diameter is going to change because of the lift off. So we have to be careful about that we are within 70 to 90 percent of the fill factor for reliable inspection. So what I mean to say you should not be too far away or too close ok. This is the work piece or this is the eddy probe this is the eddy probe. So this distance I meant as the lift off similarly lift off is here. So this depends on the 70 to 90 percent of the fill factor ok. So we have to keep that in mind. So these kind of experiences have come from the analysis in the field and the these are some thumb rules you have to keep in mind while you are measuring the field. And there will be certain phase lag between the surface and the depth of penetration and this can be given by this method. So certain variables comparison we can do. So this is the test variable. So change in the variable and then effect on the surface density and effect on the depth of penetration. So depends on the material conductivity relative magnetic permeability, geometry discontinuity, testing frequency, electromagnetic coupling, coil, current temperature. So these are the variations I will not go into the details. So like I want to say if material conductivity will increase it will have an increase in the surface intensity or the decrease in the depth penetration. So this kind of a table helps one understand which is what to be done so that we have good signal to noise ratio and good measurements. So this is a handy table which somebody can use and keep it in mind for a radio reference. And let me tell you this is something you all can keep in mind and then you can visit them again and again to ensure that you have a good acoustic sorry eddy current measurements. This is how the type of probes which can be used. This is like a surface probe. This is like an external probe like an external coil and this is like an internal probe. You can use eddy current also to measure the internal thickness of tubes and pipes like you can have an eddy current probe going inside a pipeline. So they can be traversing and then you can see. So these are used to measure the wall thickness from the inside. These are the methods like I am the one of the classes as mentioning how we can use an acoustic emission sorry an ultrasonic transducer. But you can also use an eddy current probe if it is a metal tube to find out its thickness this kind of been done. So this mode of operation can be either absolute or differential depending whether it is one coil or two coils you can find out whether it is sensitive, whether it is not sensitive. So this is how we can use this kind of eddy current probe configurations. So these are the views of some of the eddy current probes which you have in the laboratory. These are some of the internal probes, some of the external probes. In fact in vibration monitoring I was also mentioning about the eddy current proximity probe. For example when we have a shaft which is there on a particular turbine shaft which is on a journal. While this is rotating if there is a change in this gap because of an out of roundness this gap and here I can be putting in any of these locations you know maybe I can put an probe eddy current probe such proximity probe and I can measure the gap if it was a sinusoidal it will be essentially a sinusoidal with rotation. So this kind of displacements can be can be measured by eddy current proximity probes and in fact some of this eddy current proximity probes are permanently mounted on turbines at an angle of 45 degree to such probes and you can use them to plot what is known as the orbits is about 90 degree. So you can have an x and a y and you can plot them the signals and then you can you can get a circle or if you can get a point you can get an ellipse. So depending on the phase relationship between x and y you can find out from the orbit plots what is the problem with the shaft which is rotating the machine. Such eddy current proximity probes are also used in turbines or vibration monitoring or displacement monitoring and though here we just explained to you few probes which are used for measuring the coating thickness and so on. And this is a typical eddy current probe with it can have a digital meter or an analog meter which can see the voltage you know this can be all calibrated to a equivalent displacement you can have multi-frequency equipment you know where multiple probes can be done. So certain probes are used or certain calibration standards are available wherein holes and grooves are machined at different percentage of wall thicknesses like this on the inside. So I can move the eddy current probe inside it and then I can see what kind of signals are there. So like I was telling you in all this ASTM standards there are what are known as standard calibration blocks available both for ultrasonic testing I have mentioned about ultrasonic testing in one of the previous classes ultrasonic testing and also for eddy current testing. So whenever you purchase such equipment or sensors for eddy current testing or ultrasonic testing I would recommend that you also purchase a standard calibration block for example then you can calibrate it in the field ok. So some of the testing procedures is balance on the sound portion and then you know because this eddy current probes are actually the signal which we get can be balanced through a woodstone bridge circuit and then you can set the change the resistances to see what are you can balance it by changing the resistances and if there is a difference because of the eddy currents you can sense it through a galvanometer or some sort of a display meter. So effect of frequency on the signal you know you can see. So these are all different calibration standards. So you can see the frequency for high frequency and low frequency this kind of effects can be done. So now you see if you move this acoustic sorry the eddy current probe inside a tube which is made artificially with certain thicknesses you will see converting the eddy current signal to thicknesses. So 0.6 mm, 1.32, 1.98, 2.6, 3.30. So if this kind of in-situ calibration is done it gives us confidence as to whether we are what we are measuring is correct and this is a standard calibration tube which is made available to buy and this can be used to test the deflector. And some of this in a this is another example where we had a tube bottom tube top tube different baffles are there in the tube and then the thickness measurements were done and this can be achieved. So eddy current to summarize is actually used for systems where there is a metal present because when only we have metal there will be an electromagnetic induction between the primary flux or primary magnetic field which will produce a secondary magnetic field and then these two magnetic fields will interact and we will get an EMF according to Lenz's law and these if the material was not metallic I would not never get an eddy current. And few things we have to keep in mind regarding the penetration depth left of good signal to noise ratio is the magnetic permeability, the depth of penetration, the frequency of the signal which has been used to drive the eddy current, the diameter of the coil and so on. This is something which one has to keep in mind. This is another example depending on the crack. So this cross section of the crack which can be very easily detected by moving the eddy current through this can be also done. But like I was telling you all this NDT techniques so far we have discussed about thermography, then we have ultrasonics, then we have what we did was acoustic emission, then we have eddy current. We have discussed these NDT techniques so far and few more which we are going to discuss are the radiography or what is known as the x-rays, then we have the dye penetrant testing and of course, the last one the easy one is the visual inspection on places hard to reach. So, we are going to discuss about this in the next class on NDT techniques that will be your lecture number 39. And just to recap almost 10 percent of CBM or condition based maintenance use such NDT techniques. And we just discuss about these four so far in three classes I think this lecture number 36, 37 and this was 38. So, this four lectures would actually correspond to 10 percent of the CBM techniques which are used worldwide. And one has to be coming back to this eddy current testing. One has to be very careful about the frequency, the diameter of the probe and the traversing speed, magnetic permeability. There are few examples wherein such cracks can also be detected by acoustic sorry by eddy current probe. This is another view which I am going to show you is this kind of surface defects on pipes. Now, this kind of surface defects on the pipes can be detected by eddy current testing techniques ok. So, just to recap I mean eddy current is convenient like ultrasonic, but only thing that we are keeping in mind the material has to be metallic. And again lot of this NDT testing techniques lot depends on the skill of an experience of the operator and user of this equipment and mistakes do happen. So, it is always recommended that in-situ field calibration is done while using certain NDT test equipment ok. And then unlike vibration which is very well established where the signals are has have very good signal to noise ratios. They are well deterministic and can be analyzed through automated software and that is the reason why vibration monitoring is used for 70 percent of the machinery for diagnosis throughout the world ok. So, this should bring to a close this lecture. Thank you all.