 In the last class, we talked about the introduction to this course. In this class, we will be talking about the principles of maintenance. The question is, you know, what are the techniques of maintenance available to us and what kind of techniques and tools need to be known before we can apply maintenance to different plants and which kind of maintenance is beneficial, where and how and so on. The question is, you know, why do we need to have a plant maintenance? Obviously, you know, we would not like to do overdo and maintenance. You know, of course, maintenance also involves cost. I would not like to maintain a, re-maintain and maintain and so on because it is for the sake of maintenance. The machine may be up and running and doing good, but then I unnecessarily do a maintenance. But sometimes, you know, as you all know, for strategic reasons, sometimes we have to maintain it even if the things may be okay. You know, I would not like my trigger to jam in front of an enemy. I would not like my missile not to fire when I want it to fire. These are cases which we cannot take for granted, but of course, which we do always, you know, the OEM manufacturer of your car would have told, you know, every 5000 kilometer, go to the service station to get your oil changed. Sometimes we do it. Sometimes we do not do it. Nothing goes wrong with it. Okay. Sometimes we would have gone after 15,000 kilometer. Only thing that the service manager there would have, will tell you, sir, you know, you should have come earlier and blah, blah, blah. But still nothing goes wrong with the machine. Okay. So, the idea behind maintenance is that we would not like to do a lot of maintenance. We will like to do a maintenance only when it is necessary. Because if a perfectly maintained machine is available, it will do its function it was designed for. Okay. I would not like to do have any loss of production caused by failures. Right. Suppose imagine if my maintenance, if my machine was never maintained throughout the year and I have a deadline of March 31st where I have to meet the annual production targets and suddenly in the month of March, my machine goes down and I require serious maintenance. You can imagine this is going to affect my balance sheet, but because for this financial year ending March 31st, I could not reach my achieved target because my machine had to undergo a maintenance. So, these are plans, these are things which we should never do because, you know, everything leads to loss of finance. And I mean, why am I doing maintenance? I need to have high revenues, high productivity at low cost. Okay. I would not like to lose orders when the market demand is high because my machines cannot produce. Okay. A well maintained plan maintenance or a plant engineer or plant manager would have planned for its machines maintenance during the lean market time. For example, you can imagine Diwali firecracker manufacturer cannot have his machines down months before Diwali. Okay. It is going to hit the market. It is going to affect the market. So, people always do a plan. I will just tell you an example of this auto companies. Particularly auto companies in our country, sometimes in the month of July, they do a planned maintenance schedule. Okay. And not toward the end of December or not toward the end of, obviously not toward the end of the financial year. So, everybody knowing on the strategy of the requirement of their products, they find out a time when they will do a shutdown plan maintenance. But that does not mean that you will not do your day to day maintenance. You will not do your weekly maintenance and so on. Okay. Sometimes it so happens, you know, if I know that only in the month of July, I am going to overall my machine. I did not store my spares throughout the year. I did not store it from August to the subsequent July unnecessarily because I know from August to the next July, I will be not requiring the spares. Okay. Because storing spares is inventory cost. I have to store them. I have to keep them. I have to keep a stock. Keep people do a bookkeeping on them. Maybe if I require the spares in the month of July, I may at most have them by the month of June. So, my inventory reduces. Okay. These are, I mean, this may sound very, you know, trivial, but, you know, if you think of a large plant, sometimes the inventory is as high as the certain critical spares. But at the same time, it does not mean, you know, if you are going on the high seas and you have a bearing for your propeller and you do not have a spare bearing, you will be in deep trouble. Okay. So depending on the situation you are in, you are the best judge, what kind of critical spares to carry, what kind of spares not to carry, what kind of spares when to carry and so on. Okay. And then of course, you know, once we have maintenance properly, you know, if I have a good maintenance schedule in my plant, I will always have a process which is very efficient, which will be producing quality as to having the best surface finish. For example, if I am, I am a machine shop where I produced fine machine surfaces. Suppose my, I have taken no care on the tool wear conditions, I may be having the over piece surface very rough. I would not like to have a rough machine surface because my, say, turning machine cannot run at high speeds, so I cannot have a smooth finish. Okay. There has been intermittent speed fluctuations in the machine, so these marks are going to stay up in your machining process. So even a simple lathe machine has to be taken care of otherwise, if you are doing a machining operation on that lathe, its quality is going to, the machine's quality is going to reflect on the workpiece. Okay. We take it for granted that, you know, you have a nice machine, somebody say, this machine has a bad gearbox. Somebody may say, well, what does gearbox got to do with your tool? You have a new tool, so you should be having nice machine surfaces, but people do not understand that if you have a bad gearbox, there will be backlash and there will be a speed fluctuation and that is going to affect the cutting process. They are all interrelated. As engineers, I am sure you will appreciate it much better that all these dynamics of the machine are interrelated and they will eventually reflect in the final product which you are producing from your plant. And of course, you know, I would like to have the plant once up run till its maximum useful life. And of course, if all my machines are running at their best OEM specs, I will have the production capacity increased. Okay. I would have, I have seen many plants, they never run to full capacity because they are machines, they would not like to overload them. They are not sure about the design capacity of the machine. They would have bought a new machine. You would say, you know, it can, it can be loaded up to, you know, maybe, you know, 100 horsepower, but they will not load the machine till 80 horsepower. Okay. They would say, I would not like to reach the high levels because it may damage my machine. So, these are uncertainties. Okay. They will say that the machine has not been maintained, so I would not like to load the machine and create excessive wear and tear to the machine. Okay. So, again, if this maintenance is planned, like I was saying the planned downtime, planned inventory, it is going to reduce the maintenance cost. Reduction maintenance cost means increasing the overall profit and the guys in the finance will always be happy if the cost for manufacturing component comes down because that brings about the overall production increase and revenue to the company. And that is what everybody strives for. I mean, why are we doing maintenance? Because I would like to have high productivity at low production cost or low maintenance cost. So, this is a very familiar machine availability curve and sometimes this is known as the bathtub curve. If you look at this curve, there are two axis here. One is the x axis is the time and then we have the failure rate. There are three zones given by this blue, green and the wear out zone. When a machine is initially commissioned, the failure rates are pretty high. Okay. This could be because of wrong OEM specs, wrong integration, installation defects, not trying to not being able to understand the machine, not able to understand the performance of the machine, operators are not properly trained. So, these are issues by which we have always, you all must have faced maybe a nice example I always give to all my classes is once you have bought a new computer or a laptop. Okay. I mean, laptops are pretty robust and pretty user friendly nowadays. You know, I am talking about, you know, 10 years ago when people used to buy PCs, you know, first few weeks there used to be tingle with the operating system. This audio driver is not working, that hard disk is not being properly read. I am sure all of you would have experienced this. Okay. But once you are comfortable with your laptop, with the operating system, with the kind of softers, it undergoes a smooth life. Okay. And this initial defects or down times be it your laptop, be it your equipment is what is known as the infant mortality zone. The reasons could be many, not trained staff, not trained engineers for installation, installation specifications and the requirements are not matching and so on. And then once these teething trouble are taken care of an equipment, foundation problems could be another, excuse me, foundation troubles, problems could be another. We have this green period and which is known as the useful period is machine. And again, despite our best maintenance efforts, at some point in the time, you will see that the failures happening quite often in the machine. And a point will come, you will find that it is more expensive to maintain a machine than to buy a new machine. Okay. Or the cost because of this frequent shutdowns or breakdowns in the machine is not wise, not financially beneficial for you. And that is where we call a shot and then we discard the machine. Okay. So the machine's availability is given by this function uptime is the time given for the useful period and the denominator uptime plus downtime is nothing but the total time of the machine. Okay. So we as maintenance engineers have to strive for having high availability of the machine. Okay. There are three very important types of machine maintenance and one is the very first one is the preventive maintenance or what is known as the periodic maintenance and followed by the predictive maintenance or the condition-based maintenance and the last one is actually no maintenance which is known as breakdown maintenance or the reactive maintenance. And in this course, we are actually going to focus on the second one that is the condition-based maintenance. In fact, I told you in the first class that we are gearing towards all this instrumentation, signal acquisition, data acquisition, analysis all because of this CBM. I need to know the machine's condition by monitoring its vibration, by monitoring its temperature, by monitoring its pressure, etc. So we need the flow conditions. So we will be focusing on condition-based maintenance in this course but I thought I should tell you about what this preventive maintenance or periodic maintenance is. Periodic maintenance means that I will regularly follow a schedule of maintenance. There will be a time interval fixed no matter if the machine is bad or good. I do not care. I will just go on and religiously maintain, replace the spares and then my machine will be always up and running. Particularly in the defense, they do this kind of periodic maintenance because of strategic reasons but that does not mean that they do not do condition-based maintenance. Condition-based maintenance is a technique wherein we always, we underline the word, we always know the condition of our machine because certain extra capital investment has to be done, has been done on the machine by instrumenting it with additional instrumentation in terms of transducers and sensors and then we have an analysis unit so that every time the machine is running, I know its present condition. I also have its past history in my database. So by certain algorithms, I can say whether my machine is going to behave the present way or its condition is going to be okay or it is going to be worse. I can find out from the database of my data store in my database from the previous conditions of the machine. So periodic, sorry predictive maintenance or condition-based maintenance always depends on the present condition, sometimes on the past condition of the machine but the machine's condition need to be known to do maintenance. Another thing I will tell you, in fact I was telling you earlier in our meeting that like a doctor is no good to a dead man or a dead person. Similarly, a maintenance engineer cannot maintain a machine which is already dead which is not running. I can at most repair it but I cannot maintain it. A machine has to be running so that once it runs, it is going to give us the signals. If a machine is not running, there will be no vibration, there will be no noise, there will be no heat generation. So I cannot possibly do CBM on that machine. At most, I may look into the manuals, look into the design and do a repair of it but to implement CBM, a machine has to run. So that it gives the required signals and I have the instrumentation to capture those required signals. So I am stressing on the word signals, instrumentation in the case of condition based maintenance because of the fact that to implement condition based maintenance, I need to have extra investment done on the even a machine could be as new as I have just bought a new machine today. To do implement CBM on it, I have to do an extra investment right today itself from day one so that I have the investment to take care of the instrumentation, investment to take care of the analysis unit. I have skilled manpower to analyze the signals acquired from the machine. As opposed to the last one, the breakdown maintenance or the reactive maintenance, this is actually no maintenance. I do not do any maintenance on this machine. I can afford to lose the machine if the machine goes down. Like I was telling you, in the case of a steel plant, I mean I have a blast furnace, I cannot afford to lose the blast furnace but of course I can afford to lose couple of water coolers in my canteen. So I mean if the water cooler in a canteen in the steel plant goes out, I can always buy a new water cooler. Imagine the criticality of that equipment but of course if my blast furnace is damaged and down, I cannot afford to lose that. So these are issues which we will understand in the lecture on Femica that is failure modes affect and criticality analysis. From the failure modes which will affect or function, I can know what kind of maintenance effort has to be put in which machine. So I will be focusing this class or on predictive maintenance because in predictive maintenance you will see in an example which I will show you subsequently that the maintenance cost will come down with time because of the fact every instance I know the machine's condition. I can always know whether my machine is going to undergo a failure, is going to have an abnormal behavior or not and so on. So because of that I will have few machine failures. I will have less repair downtime because of the fact that I know the machine's present condition. I can always for example I will with time say I have a parameter which tells the machine's condition. For example if time you know this could be an intervals of you know one month, two month, three month etcetera. So this is in time in months. I have measured so here this is the parameter second month and slowly I see this is happening. So if I was trying to do a trending of the machine's condition toward the fourth to fifth month at the end of the fifth month as a maintenance engineer I would be getting worried. Well now something is behaving abnormal with the machine and why this parameter coming down. So I know well by six or seventh month I need to have a major repair or study into the problem in this machine so that I may be having certain critical spares already available to me by the sixth or the seventh month. So that from the month one I did not store this critical spares. If I know it takes you know only one month to get this critical spares I can plan by the end of fifth month to get this critical spares. That is one way of looking at it. And then I can be prepared that you know my machine is going to have a downtime. So I can for example I will give you an example here. Suppose you know you have a festival of Dashera wherein we give holidays to the workers. Imagine at the peak of Dashera my machine which is to be producing and there is a demand that it has to take a serious breakdown. So I have to pull out my workers or engineers from their Dashera vacations or holidays. I have to pay the workers overtime because I have to pull them out from a schedule holiday. I have to have the spares ready. So this all had to cost but had I planned it earlier and because I know the machine's condition well you know at the end of fourth month I know I will have a high production demand toward the end of seventh month. This is the time wherein I can take a repair. So these are the cases wherein I can do the reduce my inventory and then of course you know if I do a proper downtime I can reduce the machine's availability and increase production. And if I know this machine is going to have an abnormal behavior I would not like my operator to be very close to the machine. I have had a case wherein we have a plant in very close to Kharagpur, Haldia of a large FMCG manufacturer. One night the plant engineer gets a phone call that the gearbox in fact blew apart and this is a plant which they produce detergent. You know detergent production is very similar to cement manufacturing you know they have a mixing they have a rotary kiln and that rotates at a very low rpm and then it is heated at one end. So if you have a motor running at 1500 rpm and this kiln is rotating at 2 to 3 rpm you can imagine the amount of reduction being brought about in this gearbox. And this gearbox was very critical to that detergent plant and this gearbox because of reasons we will find out later on in fact that is the case study I will discuss towards the end of this course that this gearbox blew off. Imagine if the operator was standing next to it unfortunately it was in the late night shift and nobody was there. Imagine if an operator was standing next to it it had been dangerous. So if we know the machine's condition we know when it is safe to go closer to the machine while it is operating and so on. So all these are the benefits of predictive maintenance and in this course as I was telling we will be focusing more towards the condition based maintenance. Now I always give this example to all my classes on the economic analysis as to I am doing a comparison on the predictive maintenance and the preventive maintenance. The question is you know which one is good and which one is not good in the long run. These are few budget heads in the first one and if you can see this is the capital maintenance equipment cost. Day one I would require an addition though both my machines are new I am just doing a comparison. Both my machines are new. Day one I would require an additional expenditure towards the signal acquisition, the transducer, the analysis unit. I need to have that in place but for a preventive maintenance or a periodic maintenance I require no investment. Now let us look at the case of the preventive maintenance assuming there are four shutdowns a year. By for shutdowns I mean four scheduled maintenance in a year and the total cost of this maintenance which includes the machine downtime, the repair, labour cost per shutdown for repair is 50,000 rupees. For four such repairs in the preventive maintenance cost at the end of first year I would have spent 2 lakh of rupees right. At the end of 10 years if I was religiously doing this preventive maintenance I would have spent 20 lakhs. At the end of 20 years I would have spent 40 lakhs. And then if you look at the because I am doing predictive maintenance I am not doing maintenance on a regular schedule. I am only doing a maintenance as and when the signal indicators let me know that this machine requires a maintenance and usually this is a figure that at the end of 20 years if there are only five shutdowns because of maintenance since we are doing predictive maintenance the number of shutdowns will be less and this happens to be five and this is not a data which I have cooked up. This is a data which is there from the server report given by industries in many journals and this is a compilation of such. In the span of 20 years there are only five shutdowns. So, 5 lakh is my initial investment and then I would have spent additional 2 and half lakhs. So, the end of 20 years I would have spent only 7.5 lakhs if I was doing predictive maintenance whereas I would have spent 40 lakhs in the case of preventive maintenance. We do not realize that but this is the statistics obtained from industry who are doing these as benchmark studies and have reported them in journals. But always we have a problem with the engineers, maintenance engineers asking the management for more investment right day one because the management would obviously say well we have just bought a new machine why do you require an additional expenditure of 5 lakhs for maintenance because from day one I need to do a predictive maintenance so that I have the right equipment and infrastructure to acquire the signal and know the machine's condition. So, that is why this 5 lakh has to be spent there in the day one but in the long run we spend about we save about 22 lakhs in 20 years if you do such an accounting and in fact nowadays I am talking telling you the scenario of the Indian industry. In fact there are many plans I will just give you the examples in the last two decades. About two decades ago people hardly knew about vibrations people I mean vibration was always there even in engineering curriculum we are hardly being taught machinery vibration as a full-fledged course and in the industry they used to have some meteors you know used to just give you just one number and there are certain ISO standards as to certain horsepower this is to this has to be the overall vibration level. And there is to be a just a few people in a maintenance cell workers and technicians they used to they used to go to the machine measure its vibration and forget about it. If there was a breakdown people would blame each other you have not recorded the vibration they used to go back to their old log books you know log books tone pages flying here and there and this was actually the scenario of 20 years ago when people used to call me to the industries to fix their problems I used to ask them do you have any vibration data they would show me old log books you know written sometimes by pen sometimes by pencil sometimes by they do not know by what in what units they have recorded the data. Sometimes they do not know I have seen they have not put the transducer in the wrong place they do not know if there was a proper battery in the meter this is some display they write it down there was no calibration of the meters and this was actually the scenario 20 years ago in our industry. But today the scenario is altogether different if you if you go to any industry they have smart screens everywhere with touch screens next to the machine they have instrumentation transducer mounted on the machine. Data will be automatically logged into the data logger engineer just goes touches the touch screen sees the data logs sees the trending chart and there are subroutines algorithms which will tell you when the machine is going to have a downtime and this is the state of the art in the in the industry nowadays okay and now you will see wireless coming a big way into the industry there are no more long cables lying here and there as I was telling you monitoring machine sitting in Kharagpur monitoring machine in Alaska is not a science fiction anymore okay it is possible because of the advantages and the advancements done in electronic instrumentation in computers in miniaturization of the transducers transducers are smart enough they are fault tolerant they can be self calibrated by techniques okay because you know these are the advantages of predictive maintenance and engineers are smart as you are nowadays okay everybody understands instrumentation everybody understands computers and electronics everybody understands computer languages everybody understands vibration okay so there has been a paradigm shift in maintenance engineering in the industry both in the academia and in the industry okay in fact nowadays the you know we are talking about predictive maintenance you know nowadays the focus has been to if there is a problem the machine they would like to know what is creating the fault in the machine and next is there are mathematical algorithms as to find out the remaining useful life of a machine and that is very important I mean obviously to an engineer the very next question is well my machine is running fine when is going when is it going to go down that is the question everybody asks your management would ask you I mean can my machine survive for next 10 years okay and how would you answer that you have no clue all you know is you have you have data till present from the past 10 years you cannot you do so there are lot of mathematical tools available models available which will let you know whether my machine is going to last survive for the next 10 years or next 15 years or when it is going to have a condition where in the machine will be totally unacceptable okay so you all can understand the benefits of predictive exam maintenance by the small example which I showed you now what are the techniques available for vibration sorry for predictive maintenance as I was telling you we had some idea about the techniques available for predictive maintenance in the last class I told you about that 70 percent of the cases of predictive maintenance is by vibration monitoring followed by wear debris and the oil analysis which is the remaining 20 percent and then next comes the motor current signature analysis thermography and the NDT techniques which is the remaining 10 percent I will just to give an example on thermography I will just tell you an example which the industry nowadays is using to save power lost during power transmission you know this high voltage transmission towers okay wherein they have a certain switch gears you know they will be operating a switch gear and suppose you know from one transmission tower you know line line goes to another grid they will be having switch gears okay and this switch gears have to make contact okay and this is high from the ground you know from the ground and then you are you are standing somewhere here obviously this is you know 25 kv or 33 kv lines or 132 kv lines okay and they are transmitting electrical power okay now imagine in this switch gear if there was a excuse me loose contact okay there will be because of loose contact there will be arcing and there will be heat generation and there will be loss in power transmission in fact this is a case study reported by Andhra Pradesh State Electricity Board I was there in one of their Fikki meetings and they were mentioning that they have done a substantial amount of power savings in the power distribution by ensuring that there is no loose contact in the switch gears while at such high voltages and how did they find out that switch gear lose contact obviously because if the high voltage power generation I cannot physically climb up the tower and check with a measurement what they did was they had Jeep mounted thermal thermal imaging camera to shoot at the switch gear and find out the hot spots okay and they did a lot of because you know the electricity board reports to the Fikki that there is a lot of power loss in the transmission itself not in the generation but in the transmission and switch gear loose switch gears was a very very large component and that they could reduce by doing thermography okay thermography we use in detecting faults we may not be able to diagnose for example a coupling between a pump and a motor if there is a misalignment I can put a vibration transducer measure the vibration do an analysis and tell well there is a misalignment but often just to know whether there is a coupling misalignment I will just shoot the camera on the coupling the coupling is hot I know something is wrong so I can identify hot couplings I can identify hot bearings and the case of the railway axle I can identify hot conductors between behind a panel for example you are talking about ship which comes to the port for few days for an overhauling I obviously cannot remove all the cladding and the panels and to see all the conductors whether anything is hot you can just do a thermal imaging scanning and find out the hot conductors and then take corrective measures okay so thermography is a very very quick and easy way to detect abnormal heat changes and this abnormal heat changes could be because of misalignment could be because of loose contact could be because of high current being flowing in the conductor etc okay and this NDT technique is another technique wherein we have the techniques of ultrasonics the techniques of acoustic emission borography how do you detect crack in a shaft from the outside for example you would have done a fine casting of a cylinder block okay you know the engine they do a casting of the cylinder block and the some of the surfaces are machined and components are fit into that machined cylinder block but if there is a thin hairline crack in that casting and because everything will be under high pressure during the operation of the engine the oil or the flow gas or the combustion gas etc are going to leak out okay and then you will lose compression in the engine and this engine is going to be no good so how do you detect such hairline crack on a surface a surface you possibly cannot see through your naked eyes everything will be nice machine but they will be thin cracks so this is what is known as a dye penetrant test you can put a layer of dye and then under special camera you can see this dye will glow and you can know whether the crack is visible crack is present or not. So certain NDT techniques are used for finding out defects in the machine. Ultrasonic is very popular in the sense I will just give an example of ultrasonics say for example I have a metal piece and there is a crack here void so if I have an ultrasonic transducer this ultrasonic transducer what happens is it gives a wave okay incident wave and this is a denser material this is of another material so there is a density difference at this interface so because of this density difference there will be a reflected wave so in this transducer receiver you will see a pulse and if you rope this transducer every instance you will see a pulse intensity equally spaced but when the scenario when it comes here just right above the pocket of the void because of the density difference this is going to bounce back so you will see suddenly a very close spacing and then again this is say time some voltage level so this means because the time taken will be less because the distance travelled is less okay so this means that there it came across an obstacle which was of a different density than this common homogeneous material okay and that is the principle behind ultrasonics ultrasonic waves are as you know high frequency they are very directional because high frequency waves are directional like would have heard of a siren if there is a siren at high frequency you always turn your heads towards the siren because high frequencies sound or noise are directional same is true with ultrasonics okay so if I have an ultrasonic scanning of this I can do that in fact the ultrasonic thickness gauge is based on this principle okay if I know the speed of the ultrasonic wave in this medium I can calculate the from the time taken I can calculate the distance or the thickness of the specimen in fact ultrasonics are used another way is to find out the in lot of these tubes lot of the large tubes you know of certain cross section whenever they carry lot of chemicals you know either some things get deposited clogged because of this clogging what happens the thickness has measured here suppose originally this was a 10 mm thickness because of which time because of this deposition of the solid particles there will be variation to this 10 mm and if you rove an ultrasonic transducer all around the circumference of this tube you will know whether in the internal which is not visible I mean possibly vibration monitoring cannot measure it but if I if I can rove a transducer all around the circumference of this tube I can know what is wrong in this if there is any formation any deposits in the inner wall of the tube okay people of course use you know radiography also x-rays another way of doing it so there are many ways to do such NDT techniques like ultrasonics acoustic emission radiography etc now to implement all this CBM or condition based this parameters vibration wear debris oil temperature current process parameters these items written in green are actually the transducers required to sense these mechanical parameters and this transducers could be for vibration it is accelerometer and for this wear debris it is the spectrophotometer because as I was telling in this wear debris the wear debris fall off a parent material by knowing the wear debris is chemical composition I can know comparing the knowing the chemical composition of the parent material I can very well know where this or which component has worn out and because of which this particle has got deposited in the lubricating oil so that is done with a specialized equipment known as a spectrophotometer and spectrophotometers are a very expensive equipment they require specialized facilities to operate so obviously I cannot be doing putting a spectrophotometer next to a gearbox in a cement plant okay spectrophotometers are are in dedicated central research facilities in in IIT Kharagpur we have several of them in many departments in the wherein all we do is we bring in the oil sample we can through a spectrophotometer we can know its chemical constituents we can we can know the composition of this wear debris particles as to what percentage of iron is present what percentage of zinc is present copper is present etc and then of course always compare with the chemical composition of the parent material and see if zinc is in that component and I have a strong amount of zinc in the wear I know that parent component has worn out oil there are many ways to measure the parameters in the oil but one very handy way is to known as the particle count meter as to what is the size of particles available in the particle and what are what is their concentration in parts per million how many such particles are there temperature infrared infrared detectors RTDs thermistors etc can be used to detect temperatures electrical current you know you can use the simple ammeter but all effect sensor is a technique you know which has a very very high frequency response because I was telling you in the motor current signature analysis it is just not the amplitude of the current but it is the nature of the current in terms of its frequency content it is just not 50 hertz but it could be 50 hertz and also frequencies around the mechanical defect frequencies which is being driven by this of the mechanical unit which is being driven by this electrical motor so on all effect sensor is a transistor which is used to monitor electric current and then next comes the process parameters as to the finding of the flow rates of the processes particularly in chemical process plants in fertilizer plants etc people would like to monitor the flow rate with orifice or a venturi meter or a turbine flow meter there are lot of pressure gauges to measure the pressure in the chemical process and so on so this process parameters indirectly will give a clue as to what is wrong with the machines running that chemical process now you have a fairly good working idea about the components in the predictive maintenance system we of course need to have the machine we need to have the transducers we need to have the data acquisition unit we need to have the computers where the data has to be stored in a database we need to have a skill manpower and an analyst who will analyze this data and so a very generic predictive maintenance system will have the user friendly hardware and software hardware could be the transducers for measuring vibration transducers for measuring the pressure process parameters temperature transducers for measuring the temperature the transducers for measuring the current etc and on software which will tell when to acquire the data how much of data to acquire how long to acquire and so on what are the data has to be acquired simultaneously so once this software drives this whole thing this data has to be stored acquired by a computer related data acquisition it need not be just one data at a time it could be multiple sensors at all times I could collect data from all sixteen sensors around my machines at one given instance because there will be lot of statistical processes signal processes wherein we will find out relationship between two signals whether they are related or not otherwise we cannot if they are if I if I do not acquire them at the same time it is very difficult for me to tell them that they are related because once they have to be correlated they have to be acquired at the same time and then we need to have a data management system which will be responsible for storing the data and doing the trending because this if this has to be automated imagine if my plant if a steel plant have hundreds of gearboxes and pumps obviously everybody cannot be going running to each machine and doing a data logging or data taking by hand so this has to be automated this has to be through a computer driven system it may have cable it may not have cable it could be wireless but all this databases with the proper nomenclature and identity of the machine has to be taken in and then today I have a system for a steel plant the same system perhaps could be used in a cement plant the vibration processes is the same the principles are the same only thing is that I have to mount instead of a gear steel plant I have to mount it in a cement plant so system has to be very automated it has to be flexible and of course it has to be reliable I mean garbage in is garbage out if I give a junk data to my computer it is going to give a give me a junk output that should not happen okay and then the reliability it has to be accurate the transducers have to be calibrated and then it should not give any false trending and so on okay so while coming to a close I would say the predictive maintenance management is a philosophy that uses the actual operating condition of a plant equipment and systems to optimize the total plant operation okay it uses the actual plant operating condition it cannot be of a dead plant and then it is a condition given prevent it is a condition driven preventive maintenance program wherein in predictive maintenance we assume that the machine will degrade sometimes you know because of some slips unnecessary repair may be done of course but downtime can be controlled whereas in reactive maintenance because I have done no maintenance suddenly I have to have high expenses involved high spare parts high overtime rush day I have to repair this machine spares not available because machine has suddenly under taken a breakdown I have to import it so these are the advantage disadvantages of the reactive maintenance so to have a successful maintenance program I need to have these conditions they are dedicated manpower who understand the entire process and there has to be good communication between the place where I am taking the data and where I am doing the analysis I sample oil from a one machine and the technician on the other end is thinking it as machine V where I need to the data from machine A labeling and data communication between the labs is very very essential okay so some of the plants as you all know this by now that these are the common elements in the plants and machinery which will be talking about in this course and this is the generic arrangement of CBM the transducer signal conditioner and the signal analysis unit okay and you will see CBM is almost used everywhere and in the subsequent classes you will see how we use them to different machines okay thank you those of you who need to.