Acoustic Emission (AE)
Not the audible range up to 20,000 Hertz but the range above that is what we are interested in here
The PdM technique of Acoustic Emission is an area where I have over 12 years of experience applying it to equipment in the Automotive, Food and Utilities sectors with multiple examples of cost benefits and case studies.
AE technology is transferrable to any industry that uses rotational equipment, it is compliant to very low speed rotation and it can also be used to monitor pressure systems in a similar way to Ultrasound.
Ultrasound and AE are so close I refer to them as twin sisters during my training sessions
The AE instrument I have experience of using is the Holroyd MHC Memo Pro, a simple to use data-logger in the field that can be trained to Engineers and front line operators alike. I have a Holroyd MHC to demonstrate and teach with, I also offer PdM contracting as a separate service if required.
Previously I have trained PdM Specialists, Reliability Engineers, Technicians, Maintainers and Operators how to use this AE data-logger, that includes, asset criticality, taking meaningful readings, forming an inspection route, trending software, signature triggers and Spectrum Analysis where required.
Read the Uptime Consultant articles below to find out more about this little known PdM technique.
Contact me directly at andy@uptimeconsultant.co.uk for more information about training availability or a proposal.
AE technology is transferrable to any industry that uses rotational equipment, it is compliant to very low speed rotation and it can also be used to monitor pressure systems in a similar way to Ultrasound.
Ultrasound and AE are so close I refer to them as twin sisters during my training sessions
The AE instrument I have experience of using is the Holroyd MHC Memo Pro, a simple to use data-logger in the field that can be trained to Engineers and front line operators alike. I have a Holroyd MHC to demonstrate and teach with, I also offer PdM contracting as a separate service if required.
Previously I have trained PdM Specialists, Reliability Engineers, Technicians, Maintainers and Operators how to use this AE data-logger, that includes, asset criticality, taking meaningful readings, forming an inspection route, trending software, signature triggers and Spectrum Analysis where required.
Read the Uptime Consultant articles below to find out more about this little known PdM technique.
Contact me directly at andy@uptimeconsultant.co.uk for more information about training availability or a proposal.
Acoustic Emission Articles
Acoustic Emission saves another asset
This is an experience I share when training engineers, managers and front line staff the benefits of Predictive Maintenance (PdM) as part of a holistic reliability strategy.
Like all methods it's always good to have an example from the front line that demonstrates the way joined up thinking and collaboration can return the maximum benefit. This is one that tends to come up when I am explaining not just the mitigation of future failures but the wider benefits this approach can deliver.
If you have read any of my previous articles you will know that I believe the first line of defence in any industrial plant are our plant operators. These are the people that spend the most time living with the assets whilst they are producing the revenue driving products that pay all of our wages.
My first piece of advice to anyone starting a predictive program is make sure to include these staff and identify those individuals that notice things more than others, as a PdM practitioner make sure you chat with them regularly about how the process is running and anything unusual happening with the product, sometimes this information can highlight hidden issues that no technology can identify readily.
It was on one of my daily chats with an operator that he mentioned and unusual 'surge' sound that had come from an extruder barrel, that lead us to examine the ingredients throughputs and ruled out one process variable. Sometimes ingredients powder can bridge causing issues with the mix into the barrel, I could still consider this but put it to the back of the queue in my mind.
The asset in question was a twin screw extruder that produced a snack food half product; that is a pellet that is cooked later to produce the final savoury snack. The criticality was high on this particular asset as it and the sister extruder were the only two that produced this product in the whole of Europe. Any disruption to this equipment would reduce availability in the supply chain causing a chain reaction of ramifications not only in plant but across the entire customer base. Some of these customers would readily reallocate shelf space in their supermarkets if a product was crossed off due to any disruption or shortage of supply, it would be a nightmare to grab that space back if this happened.
If you have read any of my previous articles you will know that I believe the first line of defence in any industrial plant are our plant operators. These are the people that spend the most time living with the assets whilst they are producing the revenue driving products that pay all of our wages.
My first piece of advice to anyone starting a predictive program is make sure to include these staff and identify those individuals that notice things more than others, as a PdM practitioner make sure you chat with them regularly about how the process is running and anything unusual happening with the product, sometimes this information can highlight hidden issues that no technology can identify readily.
It was on one of my daily chats with an operator that he mentioned and unusual 'surge' sound that had come from an extruder barrel, that lead us to examine the ingredients throughputs and ruled out one process variable. Sometimes ingredients powder can bridge causing issues with the mix into the barrel, I could still consider this but put it to the back of the queue in my mind.
The asset in question was a twin screw extruder that produced a snack food half product; that is a pellet that is cooked later to produce the final savoury snack. The criticality was high on this particular asset as it and the sister extruder were the only two that produced this product in the whole of Europe. Any disruption to this equipment would reduce availability in the supply chain causing a chain reaction of ramifications not only in plant but across the entire customer base. Some of these customers would readily reallocate shelf space in their supermarkets if a product was crossed off due to any disruption or shortage of supply, it would be a nightmare to grab that space back if this happened.
The end consumer would also be unhappy if they couldn't get their favourite snack food as usual!
These extruders were health checked every week during production cycles with Acoustic Emission (AE) monitoring using Holroyd MHC Memo Pro handheld and I sampled the lubricant every month through an in-line take off port, I used CAT Finning Fluid Lab to analyse and track the condition of the 40 Litre fill system. Various points on the assets were identified with Sensor Nodes for repeatability of reading as shown below, you can also see a section of the oil feed pipework that feeds the input and output bearings of the transfer gearbox.
I was now able to go and interrogate the history of the oil samples and the trends of the AE outputs over the past few months. There had been no warnings previously and the review revealed no anomalies, this is where some might leave it but I trusted what the operator had heard and knew the asset history.
This is another key thing I teach, 'Know Your Asset' the more you know the better, I knew this extruder was 18 months older than its sister and that it had been in service 11 years. I also knew it had never had a main or transfer gearbox failure or refurbishment, so here we were well into the asset life cycle and understood it had been well looked after.
I decided to take an early Oil Sample (still had two weeks to go) and send it to CAT Finning Fluid Lab. I also moved to the next level with the AE inspection and took some Spectra samples from the main gearbox and the transfer gearbox; the transfer was a 5:1 step down transmission that coupled the drive motor and twin screw main extruder gearbox.
Oil analysis pack is shown below:
The Finning sample would take 3-4 days as it is a postal system, but I rang the Lab and asked them to expedite the sample which they did and I had an on-line result on the database within 48 hours, this showed a slightly elevated Fe reading but nothing too alarming. I scheduled to take more samples this time at weekly intervals.
The Memo Pro Spectra result was interesting as it showed a peak that coincided in the software indicating a possible anomaly at inner race frequency (see below). With the Holroyd MHC you also have the option of using headphones to 'listen' to points of interest, it is something I rarely used unless I was lubricating equipment in operation, you can get an idea of correct level of greasing using this method similar to some Ultrasound equipment advertised. This audio inspection indicated an intermittent click at a sub rotational speed.
Another key aspect I share when training PdM is shout about it!
In this situation I had evidence of this asset requiring remedial action in the near future, emails and one to one chats were held with the operations manager, product specialist, maintenance and engineering managers, maintenance planner and our refurbishment partner.
Everyone had the heads up so we could put a plan in place.
My role was to run the Preventative and Lubrication program, but for me it should always include planning the way forward by considering the whole operation or if some re-engineering may be involved. Another key benefit of PdM is driving improvement and engineering changes backed by evidence and not hearsay.
By having easy access to the assets original manual that was held in our library I knew the original specification and could identify the SKF Bearing type without having to guess the values to input into the Spectra software. I also advised our refurbisher to purchase two of these SKF Explorer bearings to replace the input and output at the same time, they also able visited the site and we worked out a plan to turn this transmission around in a 12 hour period during the usual 48 hour planned sanitation and maintenance window.
I also contacted SKF who were our bearing partner to work on the RCA with me when I had the used parts in hand, our rebuilder photographed the parts as they were disassembled for more evidence.
A couple of weeks later (due to process load) the transmission was removed by our technicians, shipped, repaired and re-installed with no problems. The main gearbox was flushed as both units used the same lubricant and I checked filter patches for any other debris.
The bearings were returned so that I could clean, dis-assemble and inspect, I photographed the spalling damage that had occurred on the inner race so when I discussed it further with SKF we had more information, see below.
By having easy access to the assets original manual that was held in our library I knew the original specification and could identify the SKF Bearing type without having to guess the values to input into the Spectra software. I also advised our refurbisher to purchase two of these SKF Explorer bearings to replace the input and output at the same time, they also able visited the site and we worked out a plan to turn this transmission around in a 12 hour period during the usual 48 hour planned sanitation and maintenance window.
I also contacted SKF who were our bearing partner to work on the RCA with me when I had the used parts in hand, our rebuilder photographed the parts as they were disassembled for more evidence.
A couple of weeks later (due to process load) the transmission was removed by our technicians, shipped, repaired and re-installed with no problems. The main gearbox was flushed as both units used the same lubricant and I checked filter patches for any other debris.
The bearings were returned so that I could clean, dis-assemble and inspect, I photographed the spalling damage that had occurred on the inner race so when I discussed it further with SKF we had more information, see below.
In this case the asset was 'saved' and we lost no availability, but what else did this Predictive Maintenance intervention deliver?
The last point is most pertinent, as I said earlier this was the 'older sibling' of the twin extruder assets, the other was 18 months younger. My attention was now drawn to this and two years later I picked up a similar anomaly on the output bearing that lead to an intervention that avoided another unplanned event.
The conclusion working with SKF was that the bearings had come to the end of their life (>10yrs) due to an age related mode of failure. Considering the load, environment and stresses they were under I believe they did remarkably well.
Concluding I will refer back to the training I now deliver that is based on this and some other front line experiences.
Always listen to your front line operators, use multiple technologies if available, known your assets, keep your manuals in good order, talk to your preferred rebuilder, make sure the planners for maintenance and operations are in the picture, think holistically about the impacts on internal and external customers, make sure operations understand that you deliver their availability, consider running RCA even when there are no losses, consider if anything requires redesign, record and share everything, shout about the risks and the wins when they happen, the list goes on and the benefits are multiple.
My experience with this particular incident is that Operations and Engineering grew closer in their respect for each other working as a team with a common goal.
Hope you have enjoyed this PdM/AE and collaborative insight, if so please consider sharing it, feel free to post a comment with your thoughts.
- Avoided unplanned disruption to the process
- Technician labour was used as planned not reactive
- Planned work is safer than working reactively
- Collateral damage to the main gearbox was avoided
- Resulted in zero product cross offs across the supply chain
- Involved the front line operators in the PdM activity
- Demonstrated the value of Predictive Maintenance and Lubrication Program
- Involved Engineering and Operations working towards a shared goal
- Increased my personal knowledge of bearing life cycles working with SKF
- Instigated an RCA even though we had no losses, this gave us more information
- Built confidence in the Engineering function from the Operations team
- Used a cross functional team to avoid a future catastrophic event
- Improved our knowledge of the asset for future reference
The last point is most pertinent, as I said earlier this was the 'older sibling' of the twin extruder assets, the other was 18 months younger. My attention was now drawn to this and two years later I picked up a similar anomaly on the output bearing that lead to an intervention that avoided another unplanned event.
The conclusion working with SKF was that the bearings had come to the end of their life (>10yrs) due to an age related mode of failure. Considering the load, environment and stresses they were under I believe they did remarkably well.
Concluding I will refer back to the training I now deliver that is based on this and some other front line experiences.
Always listen to your front line operators, use multiple technologies if available, known your assets, keep your manuals in good order, talk to your preferred rebuilder, make sure the planners for maintenance and operations are in the picture, think holistically about the impacts on internal and external customers, make sure operations understand that you deliver their availability, consider running RCA even when there are no losses, consider if anything requires redesign, record and share everything, shout about the risks and the wins when they happen, the list goes on and the benefits are multiple.
My experience with this particular incident is that Operations and Engineering grew closer in their respect for each other working as a team with a common goal.
Hope you have enjoyed this PdM/AE and collaborative insight, if so please consider sharing it, feel free to post a comment with your thoughts.
If you would like to book a day or two day training course introducing Predictive Maintenance, Lubrication, RCM or Reliability Strategies please contact me directly at: andy@uptimeconsultant.co.uk
Acoustic Emission Matters
When it comes to getting the best out of Predictive Maintenance techniques it can sometimes pay dividends to think and operate outside of the boxWhen it comes to Predictive Maintenance tools and techniques you often have to think laterally and come up with new approaches, you may even learn something new.
Recently I had the opportunity to do just that with a major automotive client.
It was the usual story of "we know we have an issue but we need some help to pin it down".
They have some really advanced condition based maintenance techniques in place where they can monitor assets via a local network, trending motor current and vibration readings to give an indication of any breached parameters.
This is great until you have a large asset with a known flagged issue, but which component is failing?
With no individual sensors on motors, gearboxes or bearings an overall vibration was being captured from the whole body of the machine.
By applying Acoustic Emission (AE) I thought I may be able to add some more information.
The main issue was access or the lack of it, this asset sat in what I would refer to as a 'field of robots' all diligently going about their allocated tasks.
I couldn't get near the drive or a bearing housing easily as I only had wired handheld instruments with me, so I opted for the "out of the box" approach, it may work and I believe if we don't try it then we will never know.
It was the usual story of "we know we have an issue but we need some help to pin it down".
They have some really advanced condition based maintenance techniques in place where they can monitor assets via a local network, trending motor current and vibration readings to give an indication of any breached parameters.
This is great until you have a large asset with a known flagged issue, but which component is failing?
With no individual sensors on motors, gearboxes or bearings an overall vibration was being captured from the whole body of the machine.
By applying Acoustic Emission (AE) I thought I may be able to add some more information.
The main issue was access or the lack of it, this asset sat in what I would refer to as a 'field of robots' all diligently going about their allocated tasks.
I couldn't get near the drive or a bearing housing easily as I only had wired handheld instruments with me, so I opted for the "out of the box" approach, it may work and I believe if we don't try it then we will never know.
This was actually the first time I had used an AE sensor on a whole machine!
My normal sensor application would be to get as close to the component of interest, but this was impossible without stopping to fit sensors to remote locations. By taking a few one shot readings I localised the issue to one side of the asset, I then captured a couple of spectra readings at different parts of the cycle.
I had no idea if these one second snapshots would provide any value, but they may confirm or dispel some suspicions.
I had no idea if these one second snapshots would provide any value, but they may confirm or dispel some suspicions.
The Spectrum above is something I had never seen before because I hadn't used a wide approach like this previously, remember also that this is a one second snapshot, it definitely displayed some cycles that may be down to output; sort of interesting.
The second reading looked really interesting with lozenge shaped traces throughout the waveform, could this be the fault I suspected being transmitted through the structure?
AE relies on quite weak signals within components, but here I was having to think completely differently about how I was applying the AE sensor and what it might or might not be picking up.
I personally suspected chain whip on the heavy drive, I decided to zoom in ten times to take a closer look and revealed the spectra below.
AE relies on quite weak signals within components, but here I was having to think completely differently about how I was applying the AE sensor and what it might or might not be picking up.
I personally suspected chain whip on the heavy drive, I decided to zoom in ten times to take a closer look and revealed the spectra below.
It didn't tell me much more other than there was a lot going on in this structure but at quite a low dB, these readings overall were not exceeding 1.2dB. I took a comparative reading from the other side of the asset to see what that would give me.
Same scale as the first two spectra shown and it's looking very flat, this seems to confirm that the first side of the asset tested has an issue, then I thought what if I zoom in times ten on this trace? See below:
Wow!
That repeated frequency spacing is so regular, is this the chain 'noise' transmitting through the carriage?
If it is then this looks in great condition in comparison to side one.
It was time for some comparative testing with a similar technology, Ultrasound.
These spectra were only revealed later when the captured signals were downloaded into the FFT software off site, so the comparative ultrasound may give me some on the spot outputs to consider.
I used an ultrasound device from IRISS called the Sonus XT, this operates at a much lower frequency fixed at 40kHz.
That repeated frequency spacing is so regular, is this the chain 'noise' transmitting through the carriage?
If it is then this looks in great condition in comparison to side one.
It was time for some comparative testing with a similar technology, Ultrasound.
These spectra were only revealed later when the captured signals were downloaded into the FFT software off site, so the comparative ultrasound may give me some on the spot outputs to consider.
I used an ultrasound device from IRISS called the Sonus XT, this operates at a much lower frequency fixed at 40kHz.
Above the Sonus XT from IRISS also has contact probe capability
This should provide some complementary outputs that may confirm or discount my suspicions.
Tracking four passes on the non suspect side I saw dB readings ramp from zero decibels at rest to a maximum of 14dB then returning to zero in a linear way.
The suspect side however ramped sharply from zero to peak at 45dB on four separate occasions during the cycle, these appeared as two pairs during this cycle. The drive chains passed over a duplex chain drive on the main drive shaft then within approx 90cms passed over another idler of the same configuration in the horizontal plane.
From the two pairs of events it pointed to something untoward happening as the drive passed over these two chainwheels, my thoughts were either seized chain links or loose worn links but could be something elsewhere in the drivetrain.
We were nearer a resolution by using a couple of complementary techniques albeit in a reactive way, but that is sometimes the way we have to go. At least the client can now make some more informed decisions about a shut down piece of work and a list of items required to bring the asset back to a standard condition.
This piece of work was during the practical session that takes place at the conclusion of Uptime Consultants one day course 'Introduction to Predictive and Condition Based Maintenance'.
It gives the attendees an idea of what some of the handheld PdM tools can and cannot achieve whilst working on their real issues that they suspect or know of. Sometimes it confirms some of those suspicions and on other occasions it may allay their concerns; it always adds value and in most cases pays back the total cost of the training course.
If you are interested in or planning your predictive and condition based maintenance strategy perhaps I can help, send me an email at andy@uptimeconsultant.co.uk for more information and an informal chat.
Tracking four passes on the non suspect side I saw dB readings ramp from zero decibels at rest to a maximum of 14dB then returning to zero in a linear way.
The suspect side however ramped sharply from zero to peak at 45dB on four separate occasions during the cycle, these appeared as two pairs during this cycle. The drive chains passed over a duplex chain drive on the main drive shaft then within approx 90cms passed over another idler of the same configuration in the horizontal plane.
From the two pairs of events it pointed to something untoward happening as the drive passed over these two chainwheels, my thoughts were either seized chain links or loose worn links but could be something elsewhere in the drivetrain.
We were nearer a resolution by using a couple of complementary techniques albeit in a reactive way, but that is sometimes the way we have to go. At least the client can now make some more informed decisions about a shut down piece of work and a list of items required to bring the asset back to a standard condition.
This piece of work was during the practical session that takes place at the conclusion of Uptime Consultants one day course 'Introduction to Predictive and Condition Based Maintenance'.
It gives the attendees an idea of what some of the handheld PdM tools can and cannot achieve whilst working on their real issues that they suspect or know of. Sometimes it confirms some of those suspicions and on other occasions it may allay their concerns; it always adds value and in most cases pays back the total cost of the training course.
If you are interested in or planning your predictive and condition based maintenance strategy perhaps I can help, send me an email at andy@uptimeconsultant.co.uk for more information and an informal chat.
Above are the course folders for 'Introduction to PdM & CBM'
Acoustic Emission (AE) the Earliest Warning of Potential Failure in Predictive Maintenance?
Acoustic Emission, is that the noise we can hear?
No, this is structural borne 'noise' from the earliest inception of component wear.
So why isn't it more widely used in Industry?
No, this is structural borne 'noise' from the earliest inception of component wear.
So why isn't it more widely used in Industry?
Decided to revisit the Acoustic Emission (AE) subject after a recent request from an Engineer at WMG Warwick University.
They do some great work in the SME support team there helping small businesses with Productivity, Systems and Technology. He had a WAV. file from a machine he had never seen and therefore had little information about the data capture.
This article came out of the subsequent visit and conversation about AE, it's application, limitations and use in industry.
They do some great work in the SME support team there helping small businesses with Productivity, Systems and Technology. He had a WAV. file from a machine he had never seen and therefore had little information about the data capture.
This article came out of the subsequent visit and conversation about AE, it's application, limitations and use in industry.
Within Predictive Maintenance (PdM) is a little known technology called AE, in this article I aim to explore the basics of 'how it works', and the pragmatic, practical side of its use in Industry without getting too technical.
The slides I am using here are from Uptime Consultant's "Introduction to Condition Monitoring (CM)" these are used to train front line Operators, Engineers, Technicians and Maintenance teams; it can be covered in a half day information give session or a full day that includes hands on with instruments and time in your production area to identify solutions on the spot.
If you are reading this then you probably already know the importance of PdM and the positive impact it can have on availability, with reduced unplanned downtime and less reactive work.
Implemented correctly the PdM tools will serve you well, most will have heard of Vibration and Ultrasound but I think less will know about AE. There is a reason for this and I will cover that later.
So where does PdM sit in the grand scheme of things?
See the slide below:
The slides I am using here are from Uptime Consultant's "Introduction to Condition Monitoring (CM)" these are used to train front line Operators, Engineers, Technicians and Maintenance teams; it can be covered in a half day information give session or a full day that includes hands on with instruments and time in your production area to identify solutions on the spot.
If you are reading this then you probably already know the importance of PdM and the positive impact it can have on availability, with reduced unplanned downtime and less reactive work.
Implemented correctly the PdM tools will serve you well, most will have heard of Vibration and Ultrasound but I think less will know about AE. There is a reason for this and I will cover that later.
So where does PdM sit in the grand scheme of things?
See the slide below:
Slide shows where the PdM technologies or tools sit in the lifecycle of our assets, the trick is to identify evidence before the 'Potential' or within the P-F interval. The five disciplines listed are those covered in Uptime Consultant's training packages.
AE like all the PdM disciplines is the augmentation of a human sense, in this case our hearing.
Human hearing has a range of 20 Hz to 20 kHz, useful structural frequencies start at 25 kHz up to 1 MHz.
Condition Monitoring can actually start in the audible range as we move around or operate in fixed positions in a plant, our hearing can pick up differences in how machines sound if we work with them every day.
Our hearing is also a great way to start thinking about how we can deploy AE.
One thing to clear up is that AE is not Vibration (Vib) or Ultrasound, I can see where people get confused as Vib has waves and sound is transmitted by sound waves.
AE are structure borne 'sound' waves outside of our hearing range.
Ultrasound is a very close relation but has more to do with fluid and airborne equipment inspection, pneumatics, steam traps, pressure vessels, valves, arcing, discharge and tracking in electrical equipment.
Some Ultrasound equipment is now being marketed for bearing inspections, but it's not AE.
The other thing to set straight is that AE uses Transducers that are not the same as a Vib Accelerometer.
It's been known for hundreds if not thousands of years that metals 'sing' or 'cry' when under stress as witnessed by tinsmiths and metal workers. In the late 1940s with the advent of high frequency sensors and amplifiers we were able to 'listen' to materials 'crying' under load.
Josef Kaiser was the researcher who published his thesis in 1950 reporting that materials under stress emit low amplitude 'clicks', this lead to a new NDT (Non Destructive Testing) branch called Acoustic Emission. It has uses not just in metallic engineered structures but also in large concrete vessels and civil engineering applications including bridges, dams, motorway flyovers, silos and pipelines where whole structures can be monitored.
I'm more interested in process equipment in production or utilities and manufacturing in general.
The breakdown in rotating or sliding surfaces is triggered by a number of modes including lack of lubrication allowing surfaces to clash or rub, too much lubrication causing stick slip and other effects, excessive heat, ingress of water or other harmful chemicals, excessive loads, environmental issues with dirt and debris, poor quality parts or sub standard assembly.
Human hearing has a range of 20 Hz to 20 kHz, useful structural frequencies start at 25 kHz up to 1 MHz.
Condition Monitoring can actually start in the audible range as we move around or operate in fixed positions in a plant, our hearing can pick up differences in how machines sound if we work with them every day.
Our hearing is also a great way to start thinking about how we can deploy AE.
One thing to clear up is that AE is not Vibration (Vib) or Ultrasound, I can see where people get confused as Vib has waves and sound is transmitted by sound waves.
AE are structure borne 'sound' waves outside of our hearing range.
Ultrasound is a very close relation but has more to do with fluid and airborne equipment inspection, pneumatics, steam traps, pressure vessels, valves, arcing, discharge and tracking in electrical equipment.
Some Ultrasound equipment is now being marketed for bearing inspections, but it's not AE.
The other thing to set straight is that AE uses Transducers that are not the same as a Vib Accelerometer.
It's been known for hundreds if not thousands of years that metals 'sing' or 'cry' when under stress as witnessed by tinsmiths and metal workers. In the late 1940s with the advent of high frequency sensors and amplifiers we were able to 'listen' to materials 'crying' under load.
Josef Kaiser was the researcher who published his thesis in 1950 reporting that materials under stress emit low amplitude 'clicks', this lead to a new NDT (Non Destructive Testing) branch called Acoustic Emission. It has uses not just in metallic engineered structures but also in large concrete vessels and civil engineering applications including bridges, dams, motorway flyovers, silos and pipelines where whole structures can be monitored.
I'm more interested in process equipment in production or utilities and manufacturing in general.
The breakdown in rotating or sliding surfaces is triggered by a number of modes including lack of lubrication allowing surfaces to clash or rub, too much lubrication causing stick slip and other effects, excessive heat, ingress of water or other harmful chemicals, excessive loads, environmental issues with dirt and debris, poor quality parts or sub standard assembly.
AE is captured in this case by a portable data logger from Holroyd Kittiwake.
The magnetic transducer is connected via a metre long co-axial cable to the hand held instrument and the readings are captured as close to the centre line and perpendicular to the component of interest.
If you had to go and download, use complicated software and have years of experience to interpret (sound familiar) this would be a long winded process for little gain so would probably have little or no benefit.
I'm all about proving a benefit before using any PdM and this AE device gets around this issue by crunching the data in the collector, transposing it to an easy to read understandable figure that can be compared to the previous reading for immediate action.
Multiple data points and routes can be inspected and downloaded into a history file that can trend points of interest and flag warnings if required.
It can also be set up to capture Spectra readings that can be transposed as FFT (Fast Fourier Transform) to analyse specific frequencies associated with outer/inner race, cage, element, journal or ball spin frequency.
This analysis can pinpoint early damage like spalling or fretting inside bearing assemblies (see above).
So how does it work?
If you can imagine a defect on a single roller in a bearing, every time that imperfection contacts the inner or outer race it sends out a high frequency low amplitude 'ping'; some refer to this as Shock Pulse or Spike Energy but hopefully you get the idea.
If we can track this anomaly over time and trend its journey then we will witness the degradation from a very early state before things get bad enough for vibration to be detected readily. We then have advanced warning to plan bearing or asset replacements or overhauls in a planned event; maybe in a product change or the Planned Maintenance (PM) period.
It can also be used as a pure CBM (Condition Based Maintenance) tool to target our Lubrication program.
I have experience of removing fixed time PM Lubrication tasks that risked under or over lubrication and lubricating from the AE outputs alone.
This not only reduces the PM labour time and lubricant use but also enhances the lifespan of equipment and overall equipment efficiency.
The magnetic transducer is connected via a metre long co-axial cable to the hand held instrument and the readings are captured as close to the centre line and perpendicular to the component of interest.
If you had to go and download, use complicated software and have years of experience to interpret (sound familiar) this would be a long winded process for little gain so would probably have little or no benefit.
I'm all about proving a benefit before using any PdM and this AE device gets around this issue by crunching the data in the collector, transposing it to an easy to read understandable figure that can be compared to the previous reading for immediate action.
Multiple data points and routes can be inspected and downloaded into a history file that can trend points of interest and flag warnings if required.
It can also be set up to capture Spectra readings that can be transposed as FFT (Fast Fourier Transform) to analyse specific frequencies associated with outer/inner race, cage, element, journal or ball spin frequency.
This analysis can pinpoint early damage like spalling or fretting inside bearing assemblies (see above).
So how does it work?
If you can imagine a defect on a single roller in a bearing, every time that imperfection contacts the inner or outer race it sends out a high frequency low amplitude 'ping'; some refer to this as Shock Pulse or Spike Energy but hopefully you get the idea.
If we can track this anomaly over time and trend its journey then we will witness the degradation from a very early state before things get bad enough for vibration to be detected readily. We then have advanced warning to plan bearing or asset replacements or overhauls in a planned event; maybe in a product change or the Planned Maintenance (PM) period.
It can also be used as a pure CBM (Condition Based Maintenance) tool to target our Lubrication program.
I have experience of removing fixed time PM Lubrication tasks that risked under or over lubrication and lubricating from the AE outputs alone.
This not only reduces the PM labour time and lubricant use but also enhances the lifespan of equipment and overall equipment efficiency.
The above slide shows the AE route file with a highlighted point of inspection against the Uptime Consultant schematic
I have just covered the basics of AE and hopefully it has given you an insight.
But if it's so good and user friendly why is it not as popular or ubiquitous as Vibration?
Firstly it's niche in that not many manufacturers exist that produce analysers or software.
Vibration has hundreds of options for equipment and there's also loads of standards that support this technology. The practitioners are also quite protective and some still like to guard it as a 'black art' that needs years of qualification and experience.
In some cases that unfortunately is true; you'll need a degree in Vibration to understand some data outputs, and it probably doesn't provide the return on investment and benefits I would expect.
I think Vibration and AE should co-exist as tools in the CM toolkit as each has their strengths and weaknesses.
Strengths wise AE is largely unaffected by background noise, it is more sensitive to faults under normal running, applicable to all machines at nearly all running speeds, presents faults at the asset and has good parameters to trend from.
On the other hand it needs special transducers and signal processing, it's sensitive to turbulence and crushing forces, so it doesn't like turbines, chains or some servos that give out ultrasonics.
It's not sensitive to minor imbalance, mis-alignment or bent shafts; aspects Vibration is very good at measuring.
But if it's so good and user friendly why is it not as popular or ubiquitous as Vibration?
Firstly it's niche in that not many manufacturers exist that produce analysers or software.
Vibration has hundreds of options for equipment and there's also loads of standards that support this technology. The practitioners are also quite protective and some still like to guard it as a 'black art' that needs years of qualification and experience.
In some cases that unfortunately is true; you'll need a degree in Vibration to understand some data outputs, and it probably doesn't provide the return on investment and benefits I would expect.
I think Vibration and AE should co-exist as tools in the CM toolkit as each has their strengths and weaknesses.
Strengths wise AE is largely unaffected by background noise, it is more sensitive to faults under normal running, applicable to all machines at nearly all running speeds, presents faults at the asset and has good parameters to trend from.
On the other hand it needs special transducers and signal processing, it's sensitive to turbulence and crushing forces, so it doesn't like turbines, chains or some servos that give out ultrasonics.
It's not sensitive to minor imbalance, mis-alignment or bent shafts; aspects Vibration is very good at measuring.
Slide above shows the trend history from the previously highlighted route file and schematic, this is one 'point' reading over five months.
It shows Distress® in the top graph and dB in the bottom. Alarm triggers can be set once the 'signature' level is established.
It shows Distress® in the top graph and dB in the bottom. Alarm triggers can be set once the 'signature' level is established.
Hope this article has been of interest.
My previous case study on AE in industry was published in Reliable Plant a few years ago and you can read it here:
www.reliableplant.com/Read/28900/condition-monitoring-saving
If you would like more information or a demonstration of AE using my Holroyd Memo Pro then give me a call or drop me an email.
If you would like to investigate CBM or PdM as part of your Engineering Maintenance strategy then please consider my half or one day course, the rates are very reasonable starting a just a few hundred pounds for a working group of up to six employees (call for a quote).
It covers my top five techniques, Acoustic Emission, Vibration, Thermography, Ultrasound and Tribology.
From my experience the factory floor practicals usually covers the fee by confirming and explaining a suspect asset or piece of equipment in plant, saving you time and money immediately.
My previous case study on AE in industry was published in Reliable Plant a few years ago and you can read it here:
www.reliableplant.com/Read/28900/condition-monitoring-saving
If you would like more information or a demonstration of AE using my Holroyd Memo Pro then give me a call or drop me an email.
If you would like to investigate CBM or PdM as part of your Engineering Maintenance strategy then please consider my half or one day course, the rates are very reasonable starting a just a few hundred pounds for a working group of up to six employees (call for a quote).
It covers my top five techniques, Acoustic Emission, Vibration, Thermography, Ultrasound and Tribology.
From my experience the factory floor practicals usually covers the fee by confirming and explaining a suspect asset or piece of equipment in plant, saving you time and money immediately.
How does Acoustic Emission work?
The Condition Monitoring technology of Acoustic Emission is an area where I have over 10 years of experience applying it to equipment in the Food Sector, with multiple examples of cost benefit case studies
The technology is transferrable to any industry that uses rotational equipment, it can also be used as a device to monitor pneumatic systems in a similar way to Ultrasound
The instrument I have experience of using is the Holroyd MHC Memo Pro, a very simple to use instrument in the field that can be easily trained to Engineers and front line operators, I've trained Reliability Engineers, Technicians, Maintenance Engineers and Operators how to use this data-logger including, equipment Criticality, taking meaningful readings, how to approach inspection routes, database trending, signature triggers and Spectrum Analysis where required
Holroyd Kittiwake have produced a few YouTube videos if you would like to take a closer look, the best one by far is in the one below as it gives a really good understanding of this technology and the instrument
If you would like more information please give me a call or email me at andy@uptimeconsultant.co.uk
I have a Holroyd MHC Memo Pro as demonstrated in the video if you are interested in on-site Condition Monitoring or maybe a training day to find out more
www.kittiwakeholroyd.com/images/looptb4/Onsite_instruments.pdf
The technology is transferrable to any industry that uses rotational equipment, it can also be used as a device to monitor pneumatic systems in a similar way to Ultrasound
The instrument I have experience of using is the Holroyd MHC Memo Pro, a very simple to use instrument in the field that can be easily trained to Engineers and front line operators, I've trained Reliability Engineers, Technicians, Maintenance Engineers and Operators how to use this data-logger including, equipment Criticality, taking meaningful readings, how to approach inspection routes, database trending, signature triggers and Spectrum Analysis where required
Holroyd Kittiwake have produced a few YouTube videos if you would like to take a closer look, the best one by far is in the one below as it gives a really good understanding of this technology and the instrument
If you would like more information please give me a call or email me at andy@uptimeconsultant.co.uk
I have a Holroyd MHC Memo Pro as demonstrated in the video if you are interested in on-site Condition Monitoring or maybe a training day to find out more
www.kittiwakeholroyd.com/images/looptb4/Onsite_instruments.pdf
Short practical demo of the Holroyd MHC recording a dry bearing being lubricated