I've got several dozen medium-sized AC motors (5-100HP), and a few big DC (100-500HP) motors as well. I've also got a bijillion pumps of different sorts (centrifugal, wet-ring vacuum, dry vacuum, etc.) all run by these motors. I need to get into predictive maintenance, and I'm looking for guidance as to which technology to move towards for a beginning. I want my first foray into PDM to be successful, so that I can pry more money out of the corporate wallet for more PDM systems in the future. I might need to start small, so I'm thinking of bang-for-the-buck right now, and if the technology can be applied to the pumps then that's even better.
I figured I'd ask this question in this forum because I'm hoping it will be the least partisan (vib vs. IR, etc.).
I suppose the cheapest useful system would be a decent megohmmeter, but that won't help me with the pumps... I'm willing to be convinced otherwise!
Any thoughts for a PDM wannabe with a small wallet but big dreams?

The mature program will have a mix of technologies. No doubt some people here will want to go into great detail about all the multitude of factors you need to consider in developing a strategy (type of equipment, criticality, availability of spares, etc).

Personally I think if you are going to start with just one, I'd start with vibration if you don't have it yet. Rolling bearings will always deterioriate. I think vib is the best window into the whole machine train and a multitude of problems that can be addressed early. But recognize that vib takes some investment in time to get your feet on the ground.

Lube oil would be second and actually the easiest. Assuming you already have time-based oil changes on some equipment, grab a sample during oil changes. Will help to judge the bearing condition and perhaps evaluate the frequency (not that periodic oil changes are always the right strategy). There is not much extra effort in drawing the sample if you do it that way. There are some oil labs that will manage your sample results in an internet-accessible database (like Herguth), including generating alerts and alarms and making trends available.

On the electrical side, I think you can pick a lot of low-hanging fruit surveying switchgear and panels during your first few thermography surveys of electrical equipment, but the pickings will probably dry up after your few sweeps. Also upfront you have a lot of safety issues to sort through

quote:

Personally I think if you are going to start with just one, I'd start with vibration if you don't have it yet. Rolling bearings will always deterioriate. I think vib is the best window into the whole machine train and a multitude of problems that can be addressed early. But recognize that vib takes some investment in time to get your feet on the ground.

The second question, then, would be how much technology is required to get useful results? I've seen vib meters that range from a $100 touch-probe-thingie to some very very pricey bits of kit, indeed.

quote:

Lube oil would be second and actually the easiest. Assuming you have time-based oil changes, grab a sample during oil changes.

An excellent idea!

quote:

On the electrical side, I think you can pick a lot of low-hanging fruit surveying switchgear and panels during your first few thermography surveys of electrical equipment, but the pickings will probably dry up after your few sweeps. Also upfront you have a lot of safety issues to sort through

The safety issues are a very real one, here. I would likely have to get a pair of high-tension electricians in to do the readings, so I might as well just hire an IR team and not bother paying the $10K or more for the IR camera, myself.

Mike

Actually, you will want to be careful before just purchasing or selecting technologies and throwing them at equipment. That is a common error. For instance, selecting vibration or infrared for the sake of using it will result in an expensive program with little return on investment.

This is one of the reasons for the development of such programs such as RCM. Even without going to that extent, you may want to consider deciding what types of problems you have had with the equipment in the past. This approach also relates to the selection of a specific technology or test equipment.

Please find the first paper on selecting technology. The focus of the paper is motor diagnostics, but the same concepts apply.

Howard

ConsiderationsCBM.pdf (91 Kb, 36 downloads)

The selection of a technology or even an outsourced company must be taken seriously. Especially if you select an outsource company - Part of what you must consider is that you will want all of the raw data, or access to it, especially if the contract ends.

The attached paper is entitled "The Multi-Technology Approach to Motor Diagnostics." It outlines some of the capabilities of most technologies.

More later.

Sincerely,
Howard

The_Multi_technology_approach_to_motor_diagnostics_(3).pdf (164 Kb, 31 downloads)

Hello Mike,
you have a nice problem. Just a few thoughts:
1. What do you know about your past failures?
2. What is causing your daily maintenance issues? From my experience of working with 4 plastic extrusion businesses, it is infrequently motors.
3. If lubrication, and more likely hydraulic power pack problems, is an issue, an ultrasound system from down the road in Cobourg will undoubtedly result in huge benefits. Plus you will be able to get rid of the huge number of compressed air and vacuum leaks I suspect you have - which will recover the cost of that small investment in a matter of weeks.
4. A trip to Las Vegas in a few weeks time to visit PdM2007 will also save you a lot of time and money in finding the right partners (and fellow sufferers) to help you.

Best Regards,
Tom Murphy

quote:

Originally posted by Tom Murphy:
Hello Mike,
you have a nice problem. Just a few thoughts:
1. What do you know about your past failures?

My problems tend to go in trends... some piece of machinery starts going to hell in a hand-basket and it causes me grief for two or four weeks as I wait for parts or permission to do something to it. Usually permission comes quickly, but sometimes it's a 'head office' thing. Anyway, right now my biggest problems are resin-transfer problems, but I'm in the process of rebuilding every vacuum loader we have (two dozen or so) one at a time so that problem should get sorted out before too long.
It's fair to say that my problems in general fall into two categories - pneumatic cylinders, and pumps. I have lots of equipment with hundreds of cylinders that have been running for five or more years without being looked at, and things need to get sorted out. Lots of leaky cylinders just not doing the job quite right both in the aforementioned loaders, and in packaging lines, etc. As for the pumps, I've got oodles of pumps, and our tubing lines are all wet systems with water-cooled extruders, water-cooled mold blocks in the corrugators, plus the chillers etc. Lots and lots of pumps. Add to that the fact that the whitewater system was not designed, rather it was simply 'allowed to happen' over the course of 20+ years, and you get an absolute nightmare. My boss keeps saying he'd like to just blow it all up with dynamite and start from scratch, and I can sympathize. This winter I will be totally rebuilding the whitewater system. Until then, pumps haunt my dreams. I had one fail last week, and another died today. I am the biblical Job of the pump world. The way the system is set up right now, as soon as a pump fails to start for any reason (maybe the switch failed, or somebody forgot to turn it on, or it got overloaded, whatever) the tank overflows ALL OVER THE PUMP and the motor dies. Morons designed this system. Hateful, evil morons. The kind of hateful, evil morons who knew that they would not be the ones who had to maintain it. I could use washdown-duty motors, except it takes two weeks to get them, and the bosses don't want to order one until they need it, and then they need it _now_....
So anyway, pumps and pump motors are a big problem for me, but when the system gets rebuilt I'm looking forward to pumps _not_ being a problem.

quote:

3. If lubrication, and more likely hydraulic power pack problems, is an issue,

I don't let anyone else grease anything, so lubrication isn't a problem. I consider a loaded grease gun in the hands of an operator to be a potentially lethal weapon.

quote:

an ultrasound system from down the road in Cobourg will undoubtedly result in huge benefits. Plus you will be able to get rid of the huge number of compressed air and vacuum leaks I suspect you have - which will recover the cost of that small investment in a matter of weeks.

Hold the phone and WHOA there.... ultrasound to find air and vacuum leaks? Really? That sounds like a Good Thing (tm). Maybe I'm just giddy from working 14 hours today (so far... I could still get called in tonight) but that sounds like something that I really need to research. Who in Cobourg makes 'em?

quote:

4. A trip to Las Vegas in a few weeks time to visit PdM2007 will also save you a lot of time and money in finding the right partners (and fellow sufferers) to help you.

Yeah, I got all of the mailings, but I can't even get a weekend off as it is... they're not going to pay me to take several days off to fly to Las Vegas. Maybe next year.

quote:

Originally posted by MotorDoc:
Actually, you will want to be careful before just purchasing or selecting technologies and throwing them at equipment. That is a common error. For instance, selecting vibration or infrared for the sake of using it will result in an expensive program with little return on investment.

I read that paper, and I can certainly see what you're getting at, but right now I'm just looking to make the transition from no predictive maintenance at all to some semblance of predictive maintenance as a proof of concept. There won't be tens of thousands of dollars spent and a squadron of techs, there will be me and me alone.

Lonely Mike,
If there will be you and you alone then why don't you go for the most cost-effective way: a screwdriver between your ear and the machinery surfaces? Machines never lie. Yes, for myself I am using sophisticated instrumentation. However a screwdriver I have always at hand to let the machine confirm what my instrument tells me!
Old-fashioned? Certainly not in my opinion!
Arie Mol

quote:

Originally posted by arie mol:
Lonely Mike,
If there will be you and you alone then why don't you go for the most cost-effective way: a screwdriver between your ear and the machinery surfaces?

That's what I do now. I was doing that just today, actually, to check the bearings in a 500HP DC motor, but it's hard to 'trend' that data. How do you put a value to it to say whether it has gotten louder or higher pitched since the last time you checked? Granted if it's a significant difference then it's noticeable, but just today I noticed that the drive-end bearing on that motor seems to have a bit of a whine to it, but can I say for certain that it wasn't there last time? No, I just don't have that ability.

quote:

PDM wannabe with a small wallet but big dreams

Sure, trending is the dream. But at least you are able to capture a prime suspect that your screwdriver can pay more frequent visits to in order to get an estimate of the progress of failure.

Mike:

Pardon the delay. I was a little distracted this week.

The idea behind the paper was to reinforce the concept of selecting the tools that you need for the types of issues that you have. For instance, if you have problems related to the motor system (distribution, controls, motor, coupling, load and process), then you would want to know the types of problems that you have seen. Then be aware of the service companies that are able to provide troubleshooting or additional services.

For example, if you are seeing bearing problems in fans, pumps and motors, you may want to consider vibration instrumentation. If you are seeing a number of winding problems, you may want to consider a motor circuit analyzer.

I would recommend that you review your records for the types of problems that you are having. I often hear people tell me that they primarily have mechanical problems only to discover that they have a high volume of electrical problems. And, for whatever reason, perception will cloud reality. When you look at the history, you might find that it reflects something different than what you are seeing.

If you are having multiple issues, there are inexpensive tools that require minimal training or experience that can give a broad overview of condition.

Howard

quote:

Originally posted by MotorDoc:

Pardon the delay. I was a little distracted this week.

Howard

Just a friendly reminder. You are behind with answers on other topics too. Something like 2 months behind!
Regards,
jank

Sorry, Jank. You are mistaken. I have not promised anything on this particular string two months ago. It was only just started on August 12.

quote:

Originally posted by MotorDoc:
Pardon the delay. I was a little distracted this week.

No problem. I appreciate the response.

quote:

For example, if you are seeing bearing problems in fans, pumps and motors, you may want to consider vibration instrumentation. If you are seeing a number of winding problems, you may want to consider a motor circuit analyzer.

The electrical problems that I see in motors tend to be catastrophic, caused by an overload or water infiltration. By the time any problem is seen all three windings are 'open' and no amount of poking at it is going to help it. Part of the problem is the previous regime having a rather elastic interpretation of allowable amperages in the heaters on the motor starters. Let's just say that I find myself routinely turning down the dial on overloads. Anyway, from reading that paper and talking to Allan Rienstra at SDT, it seems to me that while ultrasound may do fabulous things for us for leak detection, vibratory equipment would be better for us right now for detecting bearing faults, so that's where I'm leaning.
As always, that leaves me with another decision regarding the level of technology to aim for. A 'standard' vib. meter can give you a quantitative result, so it can be trended without trouble, but it's a manual system. No doubt there are guys out there who wouldn't want to have to do without their data-gathering vib. gear that does trending etc. on-board, but for the perhaps 100 bearing pairs that I'll be testing per year, that's not really necessary. I'm not even sure if I need both velocity _and_ acceleration readings.

quote:

I would recommend that you review your records for the types of problems that you are having.

Unfortunately, I only have four months worth of records, so it's a bit of a shot in the dark.

Mike

Generally, if the overloads are set higher than normal, then they may have either set them that high because of 'nuisance tripping,' or to avoid such a problem. Nuisance tripping is often the result of insulation degradation prior to the catestrophic failure.

If you are looking for just the ability to trend with vibration, and you are looking for primarily bearing, unbalance and the like, SPM puts out a basic system that focuses on shock pulse. The instrument, designed for looking at bearings, in particular, also has a contact ultrasonic capability. We evaluated it as part of a 'technology on the floor' concept we were reviewing with an automotive manufacturer. The amount of training is minimal (self training is possible).

Of course, you are limited with its range. But, from what you describe, it would be a relatively low-risk investment to start the vibration part of the program.

Howard

Ah, here it is:

http://www.spminstrument.com

I am attaching the technical details for the specific instrument we were looking at, but there are others on their site, as well.

Howard

TD249B.pdf (112 Kb, 18 downloads)

quote:

Originally posted by MotorDoc:
Ah, here it is:

http://www.spminstrument.com

I am attaching the technical details for the specific instrument we were looking at, but there are others on their site, as well.

Howard

I'm hot and cold on this thing. It looks pretty darned simple to use, but it's not a technology that I'm familiar with. Will it realiably trend? I mean, would you expect that you just get green lights and then one day they turn yellow, and it's that easy?

Hello Mike,
I am very interested in your problem and also in some of your comments.
Would you mind a discussion outside this forum?

Best Regards,
Tom Murphy
email: tom@reliabilityteam.com

quote:

Originally posted by MotorDoc:
Generally, if the overloads are set higher than normal, then they may have either set them that high because of 'nuisance tripping,' or to avoid such a problem. Nuisance tripping is often the result of insulation degradation prior to the catestrophic failure.

No doubt it was something like that. I don't like setting them higher than the rated amperage. I figure the service factor is there just in case the voltage or phase balance goes a bit wonky, not so you can be foot-loose and fancy-free with the amperage. Mind you, some of these have been set far far beyond even the SF-corrected maximum. Like, 200% of rated amperage on small motors. Some day I'm going to spend a few hours and dig into every motor drive and motor starter in the whole place and set them all. I haven't done it yet.