Basic Vibration Analysis

All of my vibration work was very different from your problem so I don't really know what I am talking about. With that disclaimer.....

It is interesting to try to come up with a comment for your question. Using an instrument to collect data seems like a very good step. But, what vibration is "good", what is "ok" and what is "bad"? If you have 10 of the same design and age engine then with good vibration instrumentation you could probably look at data and identify the source engine by the vibration signature. They would probably all be a little different and they would all change with age.

An off topic analogy would be my experience with the RF location system LORAN. Once upon I time I used LORAN signals and at times would monitor them daily. I used to freak out the Coast Guard by calling them up, telling them I was looking at their transmission and asking for confirmation that they switched from transmitter serial number X to transmitter serial number Y during the night. I would often get quickly transferred to the station commander. The commander understood that all transmitters were the same unless you looked carefully enough to see that they were not all the same. I would get my confirmation from him and that would be the end of it. The point is, with good measurements all the engines are different and that does not necessarily mean that there is a problem.

I guess my comments are:

1) Other than documenting "before" and "after" what will you do with your vibration data? Will someone try to use the data to "prove" a status of "good" or "bad" when all they really have is data that they are not qualified to interpret?

2) How do you "I know there is abnormal vibrations present" and why is that not good enough to justify the repair? How will you use this instrument to justify a judgement of "abnormal" that is more valid than the judgement you have already made?

3) Vibration is a tricky animal. Where and how you make your measurements will probably make a big difference in the data. You will have to be very consistent. You will possibly also affect the data in a "non bearing way" by any gaskets or bushings you replace. Note that vibration from other sources and vibration standing waves could easily make this task more tricky.

On the other hand using this instrument will probably be very interesting and could be very useful. Possibly very, very useful. Just be careful putting faith in an new instrument and technology unless you understand them.

Do you have any vibration measurement procedure standards or pass/fail criteria to use?

If anyone has experience using a similar instrument for a similar task then put much more faith in what they say than what I said. I'm just saying that it is interesting but you need to be careful.

The Fluke is good gear, as good as it gets in the price range. But not for IC engines. It works best for simpler machines, with several identical bearings for example. The complexity of the combustion, the tappets, valves, timing, injectors, pistons and everything else that moves, maybe 10,00 parts means its difficult to analyse this sort of machine with an 805.

Its all down to the interpretation. With respect to all the tech heads out there, we must remember that for many many years there was no analysis available apart from the engineer's ear and hand.

If you have severe vibration then one or more of the following must be true.

Misaligned shaft / out of center boss or rotating component / loose or damaged moving part/ badly worn bearings. Any of these can indicate imminent catastrophic failure. However, if its a trivial external part, like a loose circulation pump or sheet metal housing it may be noisy but not catastrophic.

I would get someone trustworthy to carefully examine and video all around, under and over the plant. Then do the thumb test with a rod of dowel and your thumb to the ear-flap and listen to the internal workings. Then you have a document of the outside (which you can review and compare with later similar documents) and an educated guess about whether the problem is internal and therefore more serious. With some practice you can use the rod to safely hear each journal, push rod tappet etc individually and you might be amazed how the sounds vary from a good tight unit to a worn out one.

I have never found an engine that does not respond to this sort of examination.Its cheap and takes seconds.

Like everyone else says, in order to make sense of your Fluke data you need to have a good control or two, which you can use use to subtract from the suspect machine and you are left with the error data. This only works well if the entire installation is standard. For example it might work with a standard production car or a generator providing there have been no modifications. Unfortunately if your suspect unit is different externally from the control, all you can do is a Fourier analysis and guess.

I would say the trained experienced engineer's opinion is much more use that the Fluke data, intriguing though it might be. The Fluke is really only for machines with a few moving parts, like compressors, saws, fans, or pumps. Nice gear though.

There is alot of not very helpful advice here. You should really call an expert to help you weed through. My voice is just like everyone else here, you get to choose who you listen to.

The best advice I saw here was by Haroldus when he suggested misalignment, unbalance, bearings etc. but then he went off on comparing machines.

I don't believe you need to compare vibration from any other machine, that might give some indication on valves, pistons or other such parts, but not on what vibration analysis is usually used for.

Your problem could be bad compression or combustion on one (or more) cylinder, that could be seen by observing and comparing with other motors (but there are many other ways a mechanic could easily test for that.

What vibration analysis can tell you it can do just by measuring one engine, one part at a time.

I am far from a vibration expert, but what I do know is that you want to take vibration readins as close to the shaft or spinning part as possible, you want to read it on an x, y and z axis so you can see how much the shaft is moving in all three directions.

If your motor is driving a pump through a shaft and it is misaligned it can cause vibration up, down, side to side and back and forth along the shaft.

If your pump (or motor,or shaft coupling) is unbalanced it will usually show on only the x or y (perpendicular to the shaft).

If you have roller bearings with a damaged race, it will show as a number of knocks equal to the number of bearings per revolution as each bearing hits the spot. If there is a bad bearing, it will happen at a much higher frequency than the shaft speed rpm (because the small roller is spinning faster).

The point is, you need to take readings all over your machine on each spinnign part you think might be causing the vibration.

Then you need to analyze the data, and sometimes it is not any of the above (or other not mentioned problems with the motor/pump). It could be a harmonic with a support structure or pipe (but that is easily identified because it only happens at specific rpm).

Once you take the vibration data, you need to download it and examine it with software that will display the spectra of vibration. This can show in vibration acceleration, distance etc. It also shows the rpm the problem is at, and by knowing where the problem is at on the spectra you can pretty much (after referring to a book or years of experience) know what the problem is.

Drew K

Tim,

The Fluke 805 meter is a good troubleshooting tool and can answer a lot of questions.

Regardless of what you discover/find with the meter and before you disassemble the engine; it would be worth you time and effort to look closely at the harmonic balancer on the engine.

Some (most) of the harmonic balancers are of two part design with a pliable dampening material sandwiched between the two mating surfaces.

The balancer is often damaged by personnel working on the engine such as changing of drive belts, adding/removing auxilliary equipment drive pulleys, etc.

In addition, radical impulse loading of auxilliaty equipment being driven by a pulley mounted on the balancer can also cause displacement of the two components of the balancer.

The resultant is that the outside part gets rotated independent of the inside part or gets "cocked" off center and therefore radically affects engine balance.

Some engines also have counter balance weights added to the flywheel and if the flywheel is removed or replaced, the weights can be moved off their original installation location and in some cases get completely knocked off during maintenance.

Normally this vibration condition is not present when the engine is new but appears after maintenance has been performed on the machine over a period of time.

Over the years I have found harmonic balancers are more often than not the cause of engine balance issues rather than an internal component.

I have occasionally found new harmonic balancers units right out of the box that are not in balance.

One other thing I would suggest is to check the engine crankshaft end thrust clearance as some are shimmed between the flywheel and crankshaft end and the shims can be left out/altered during maintenance.

Hope this helps.