Podcast: Mastering advanced vibration techniques—From modal analysis to motion amplification
Key takeaways
- Formal training alone isn't enough—intentional, hands-on experience accelerates workforce readiness.
- Bridging generational gaps requires clear career paths and trust in leadership follow-through.
- AI tools should support, not replace, critical thinking and field expertise on the plant floor.
- Modal analysis and motion amplification are powerful tools to diagnose and fix elusive resonance issues in aging equipment.
Glen Powell is a senior engineering analyst with Hydro, Inc.’s Global Engineering division. He has nine years of experience in centrifugal pumps, vibration analysis, machinery reliability, and 3D simulation and design. Glen recently spoke with Thomas Wilk, chief editor of Plant Services, about the evolving challenges of workforce development in the maintenance and reliability space. The conversation also touches on the role of emerging technologies like AI in reliability workflows and dives deep into vibration analysis, including its limitations and advanced applications for diagnosing issues like resonance.
Below is an edited excerpt from the podcast:
PS: So Glen, tell us a little bit about what it means to be a senior engineering analyst with Hydro. What are some of your day-to-days? What are some of the perks of the job?
GP: My role has evolved at Hydro out of several different experiences. When I first joined the company, I worked at the 40th Street plant where we do repairs on pumps for the industry, and I was just the shop engineer doing repair drawings and things like that.
I transitioned from that to working at Hydro's test lab in the same facility, so I worked there for a few years. I went from the repair side to the test side, and then I went into the field engineering side—working with our customers out in the field doing reliability and maintenance troubleshooting.
From that, I’ve transitioned into the role that we made up for me now, which is the senior engineering analyst. It’s tying a lot of that together, including some training, including problem solving and troubleshooting directly with customers, working with our service centers to troubleshoot pumps. It’s really bringing knowledge and experience to lots of different situations. We carved this role out working with the other engineers in the company to try to make use of everything that I’ve done so far in my career.
PS: You must be really well positioned to see the breadth of industry and see the challenges that various verticals are facing—and also various generations of plant workers, both the generation retiring and the generation coming in now.
GP: Yeah, absolutely. When I started out in my career, I came into the pump business with no experience at all in pumps. I had an engineering background, but I didn’t really know pumps at all. I came into it very humble: “I don’t know anything about pumps—everything that I know, I’m going to learn from you, the people that I work with.”
And as I’ve gone through my career, I’ve seen a lot of older engineers retire. I’ve seen a lot of new engineers get hired—maintenance people, reliability people—and seen the struggles that they’ve had getting started and getting up to speed. That’s not only when I was working out in the field, but also the engineers that we’ve hired for reliability services and the test lab.
It’s really challenging to get up to speed, and that’s one of the things that I think is a critical thing for plants and managers to get their arms around, is trying to understand how to onboard and train people, and educate people on how to do their jobs.
PS: Is there something effective that you’ve seen on the training side that really works to ramp people up quickly? Part of the challenge is taking advantage of people’s experience who’ve been there a long time, while also making sure that that experience gets transferred effectively to new workers, augmented by whatever certification courses might be supplemental.
GP: When you look at somebody new, you have to think of them as requiring a couple of different things. There’s no way for you to get over the initial hump without some formal training. You have to get somebody inculcated in what is required to do the work and to do the job.
But I think a lot of people leave it there, and they don’t have a strong idea of how to do the next step—which to me is the experience step. They say that, “Well, the way I learned was I got thrown into the deep end and I figured things out.” But what that really leaves behind is that the culture has changed a lot. There was a lot of very direct mentoring that was going on, there was a lot of knowledge transfer that was informal, and especially with a lot of people retiring, I feel like that informal knowledge transfer and experience has really broken down.
For me, the key is to do more formal knowledge transfer and more intentional experience. For new engineers, we like to rotate them around to different programs—they work with the inspectors, they work with the mechanics, they work with the machinists—and try to give them a little bit of seasoning in a very intentional way.
But you really just can’t stop with classroom training. And I say this as a person that does classroom training—that’s one of the things that I do in my role, is to take people into a classroom setting and teach, especially about vibration analysis. But there’s always a divide, right? There’s people who do vibration analysis and people who want to know about vibration analysis. And, there’s some people who don’t want to be there. You’re speaking to different people and they require different things. The people who are going to be hands-on, they need to get hands-on with the equipment, with the different technology they’re going to use, or it’s going to be really slow getting them to success.
I’m at the point in my career where there’s young engineers that I’m responsible for and I’m scratching my head trying to figure out, how do we get the wheels turning quickly? Experience and knowledge transfer I think is the key to it.
PS: You mentioned the term “intentionality.” That’s part of what it means to enhance reliability in an intelligent way. What I keep hearing with the newest generation of workers is that it’s really important that plants and teams do what they say they’re going to do, in terms of mapping out what a career path looks like—mapping what these skills programs might be. Because that’s one thing where anybody—folks especially from Gen Z and some Millennials who are joining a new plant—they really do want to have trust in people in leadership positions, that they’re going to do what they say they’re going to do, and fulfill that career map laid out for at least the first 3–4–5 years, right?
GP: Yeah. I think a lot of that comes from just the track that people are taking to careers. Maintenance and reliability people increasingly are college-educated people, whereas the older generation are mainly people who came up and worked in the trades and worked directly at the plant and didn’t necessarily have a college education. They were more familiar with apprenticeship, they were journeymen, they learned in a different way.
I see that a lot with the younger generation. Their track to learning has always been a classroom track, right? They’ve learned in that way, and so they think that’s the best way to learn. And so you have to both indulge that some, and give them that sit-down, classroom-style education. But you also have to say, “OK, we need to get away from that and into the practical, We need to get to the machines, taking them apart, putting them back together, understanding the equipment you’re going to use,” and going through the actions you need to do.
If you can’t get people to that step, then they’re going to get lost behind the desk. They’re going to be behind the computer, behind the desk, where they really need to be on the shop floor, in the plant, in order to be successful doing this type of work.
PS: Given what you just said about being comfortable working with people who are used to a classroom setting, how can plants embrace technology itself to enhance reliability intelligently? Are we looking at more classroom-focused or more self-paced learning? Are we looking at courses delivered by OEMs in certain technologies, like ultrasound or infrared, where people can scale up that way?
GP: Yeah, it's tough. I don’t know that I really have the answers there. I have my own personal ideas about it. Especially with the newest generation of interns that we’re seeing—a lot of people have this in their head that they're going to be able to use AI tools to enhance learning and enhance teaching and training. But I think it's the type of thing where we're going to have to be very careful with how we deploy this type of thing.
On one hand, if you sit somebody down for a classroom lecture and then give them a recording of that classroom lecture later, it can be very valuable, they can go back and review. I especially think this is important for people whose first language isn't English. If you’re teaching a class in English, that’s actually one thing that’s key. I work with a lot of people who speak Spanish as their first language, and they have good English skills but it’s not their native language, so the ability to go back and review is really good.
I also see the youngest interns wanting to think less and offload more, like to ChatGPT, and not engaging deeply with the material. Saying, “Oh, I’ll just have it summarize for me.”
That means all the knowledge lives in the summary—none of it lives in your head. Trying to get people to not over-rely on those shortcut technologies is tough. That’s where experience comes in. If you put people in front of it and say, “What are you going to do?” rather than, “What do you know about this?” where you can use those kinds of tools—“what are the actions you’re going to take?— that gets you a lot closer to productivity. Focusing on the actions more than just the learning in the classroom learning environment is going to be important in the future.
PS: Some of the applications we’re hearing about at Plant Services align with what you’re talking about. Like, say, having an AI chatbot developed, and then deployed in a CMMS system that helps people in the field enter more accurate, more thorough data into the work order itself, instead of writing “motor broken” or “lube bearing,” the chatbot might ask, “Any other details? What’s the nameplate on the asset or the asset number?” It prompts positive action where if they’re using the chatbot, conceivably it’s a way to help them get more complete information so the next time someone has to work on that asset, they’ll have this more thorough library to draw on.
GP: Yeah, and I think that’s coming in the future too. The way people misuse AI now is by putting it in the driver’s seat and act as the AI’s secretary, feeding it little bits of info and expecting it to do the thinking. But I think the effective way to use it is for you to be the talented, thoughtful, resourceful person, and for it to be your useful secretary. It does the bookkeeping, it gets you references, it gleans insights from the data that are going to be time-consuming for you to get. You have to be the brains of the operation. If it’s a tool you use, you’ll be successful. But if you try to make it the brains of the operation, you’re going to end up in deep water and out of your depth very quickly.
PS: In one case I heard, just the inclusion of an “off” switch was critical in building adoption of the tool. Say you’re in the field and just don’t want to use AI that day, or don’t think you need it, it makes a difference to have even that level of control, “on” of “off”. It works for you, you don’t work for it.
GP: Yeah, and sometimes you just want the minimum hurdle to getting the job done. There’s lots of things you need to do where the best tool you can have is just a pencil and paper. Sure, you could bring your laptop out and type notes into an email—but if just have a little scratch paper, that’s fine too. It’s the same concept as if you take this huge, over-arching tool that has so much capability and try to apply it to a really small situation, it’s not going to be useful, so I agree with that completely. Having the “off” switch and saying, “it’s not that kind of task, this is a paper and pencil type of task” and we don’t need any further assistance.
PS: If we could, let’s focus on one of your areas of specialty—vibration analysis. It’s one of the most common predictive technologies in the Plant Services surveys we conduct every other year. Vibration always surfaces to be the primary or secondary predictive tool that survey respondents tell us they use. It’s got obvious applications for rotating assets like pumps. What are some of the limits of vibration analysis, and what are some ways people can use it as a secret weapon or an untapped opportunity?
GP: Yeah, it’s really important, because it is one of the most useful tools for reliability analysis. It’s an incredible tool for it, but it’s not a silver bullet for everything. There’s lots of situations where it will point you in the right direction, but it’s not going to give you a solution.
I’ll give you two examples. One is when you see vibrations at the running speed of the machine—so-called 1X vibrations. If something’s tripping your alarms and it’s in that frequency range, there are a lot of different sources for that vibration. You can’t just leave it at that top-level analysis. If you find that vibration, you need to be able to dig down and say, “is that due to imbalance? Misalignment? Or more likely, is it due to something like soft foot?”
Most likely the guys that are rebuilding your pump know how to balance it, and most likely the millwrights who are installing it know how to align it. Something that's a little bit less obvious like a soft foot or a cocked bearing, if something was installed and balanced and nice but something is out of place like that, vibration can lead you there. But that's the thing that the vibration analysis is really good for. If you're taking phase data, if you're doing a survey that goes into more detail on what's happening on what place on the machine, you're going to be able to suss a lot of that out. But you just have to know to take it to that second level of analysis. Your basic route-based checks are only going to reveal that there's a problem, not necessarily what the problem is.
Another one that oftentimes comes up is wanting to apply vibration analysis to too many situations. I recently had a talk with a customer who has an issue with cavitation and we have some sensors installed on that pump that are taking good vibration data. They're getting data on that machine every 5 minutes. So we're trending that data and seeing it, but their main concern is on the cavitation. And they're saying, “how are we going to see this in our vibration data?” And the real answer is they're not. Cavitation is not something that's really best to diagnose with vibration analysis.
When you have cavitation, you're going to hear it and you're going to be able to feel it, but it's going to show up in a spectrum as a very broad based noise. So because of that, what's the noise? You know, this is noise, it's not really something that's useful for diagnostics. But at the same time, if you have cavitation, you do really have a serious problem that needs to be addressed. It's like understanding the gaps there. That's something where our noise level went up, and it could be cavitation or it could be the train went by, you know? That's one thing I see a lot, is once you once you open up this playbook and say wow, there's so much great stuff that vibration analysis can do, and then making the mistake of thinking it can do everything, which is not really true.
But then on the other hand, you know the question of what's the big opportunity with vibration analysis? I think for me, something that isn't as well understood in plant maintenance is resonance. Vibration analysis and taking good data in terms of vibration can reveal a lot of problems with equipment, especially as over time as things get older, 20-30 years into the operation of something, you can have something enter a resonance condition. If you hit the natural frequency of a machine with the exact right forcing frequency—as it's doing its job, it's creating vibrations, and if those vibrations from it doing its job line up with the natural frequencies of that machine, you get this really large vibration resulting due to resonance. That condition can be very confounding to understand diagnose and remedy.
It's something where I think that people see it as a niche application of vibration analysis. But I actually think for a lot of plants with bad actors and pumps that have given them a lot of problems over the years, that resonance is at the root of that in a lot of cases. You have a lot of machines that are trashing bearings every six months when the sister pump, which is right next to it, is running for five years. And the root cause of that isn't that they don't know how to align it. The root cause is due to the resonance and the bearing housing, something like that, and you're going to end up with these repeated failures and maintenance headaches if you're not able to correctly diagnose that. And that's something that I think people need to be using vibration analysis more, for is really understanding resonance problems when they come up.
PS: In a case like that, are there complementary technologies that would help you unpack exactly what is happening with the resonance? Would ultrasound be useful to help separate out the frequencies?
GP: Yeah, there's two things that I think come into play that are really useful. There's a method of taking vibration data called a modal analysis, where instead of simply taking data on a couple of places on the machine, you're going to take data all over the machine. And instead of taking data while it's running, you're going to excite the machine while it's off by impacting it with an instrumented hammer.
When you build up all that data all over the machine, you can understand the mode shapes of the machine. If it's hitting a resonance then you can use that analysis to understand how is it moving—is it twisting? Is it bending? Is it rocking? What's the actual result of that resonance condition? -- and that will point you toward how you fix it. If you have a discharge head that's waving in the wind in the horizontal direction, you have some idea that you’ll need to brace it up in that way. But if you put that bracing on there and the actual resonance is that it's twisting, that it's moving in torsion, it's not going to do anything. Using that kind of technology is really nice.
The other technology that I think is really interesting for dealing with those problems is motion amplification, which is a technology that is doing a technique called an operating deflection shape. It's taking data on the machine while it's running, and in the case of motion amplification, it's taking that data with a camera, and it's plotting how that machine is moving as it's experiencing those vibrations, and you can dial it in to see the motion of a particular frequency.
So in the case of like a misalignment, you got your motion amplification set up, you're watching the coupling, and you will see the misalignment happening across the coupling. So it’s saying, OK, we have a vibration at 1X, right. We have a vibration at the running speed of the machine. But what is the implication of that vibration? That vibration is telling me, oh, yeah, my whole casing is moving, or my bearing housing is twisting, or my pump coupling is moving out of phase with my motor coupling. It’s the idea that a picture is worth 1000 words.
Those two techniques—modal analysis for resonance, and operating deflection shape (or motion amplification)—are really valuable for troubleshooting. And those are two of those high level techniques that give you a lot of traction when you're trying to deal with a problem pump or a problem installation.
PS: When it comes to redesigning and rerating existing equipment, ever since COVID the trend has been to try and repair what you've got instead of automatic replacement. With current business conditions what they are, I think we're going to see more of that. There was a survey done two years ago by EASA, which indicated that more people were holding on to their assets and trying to repair them instead of replacing, especially high horsepower assets. What are your thoughts on the opportunities right now that exist for redesigning and rerating equipment, rather than replacing them as plant conditions change?
