High 2 & a half x's on a diesel gen set

[steve383]

I have a V12 Waukesha diesel driving a (60HZ) 1050KW Leroy Somer generator at 1200 RPM. The generator has one bearing on the outboard and none on the inboard. The inboard is directly coupled to the crankshaft/coupling. The vibration over all levels are around .3 IPS at each point on the engine and generator unloaded. As the generator is loaded you can watch the 2.5x’s (50HZ) on the outboard end of the generator, vertical and horizontal, grow from .05 IPS to .8 IPS. The 1 and 2x’s stay around .1 IPS. The inboard of the generator stay around .33 IPS with 2.5x’s at .06 IPS. The over all levels on the diesel grow from .3 IPS to .35 IPS. The 2.5x’s on the engine stays at around .05 IPS.
The unit skid is made of thick I beams on spring isolators. The skid has been reinforced with X braces with no change in levels. The skid was set on wood blocks and the high 2.5x’s was still present on the outboard of the generator.
Can anyone tell me what could be the cause of the High 2.5x’s (50HZ) on the outboard end of the generator when loaded?
Thanks for your help!

 

[MachineryWatch]

Steve,

Has it always been this way, or is this a recent development?

Skip Hartman

http://www.machinerywatch.com

 

[steve383]

This unit has been in and out of service for a little over a year. I was just called in to look at it last Friday. What I do know is that it has had high vibration from start up, over all vibration wise. They did not take spectral data from the start.
The vibration caused an electrical fault in the bus bars that are in the box mounted to the generator. Big boom! (200 hours) They installed a new/differant generator, same model. It also has high over alls. They hired a differant company to find the issue. They found the high 50HZ and suggested X bracing the skid.
I can only assume that the issue has been the 50HZ from start up.

 

[MachineryWatch]

Steve,

Are there other similar units running without this problem, or is this only unit like this in the facility?

It seems unlikely that the vibration would be a conventional resonance issue since you describe both horizontal vibration and vertical vibration levels becoming equally excessive. Usually, a resonance results in highly directional vibration. Have you explored the possibility of a torsional resonance?

Skip Hartman

http://www.machinerywatch.com

 

[steve383]

There are two other larger units. They are a differant model and manufacturer. They turn at 900 RPM and do not show a high 2.5x's.
I did bump test on everything, while it was not running, related to the unit and did not find 50HZ.
The company in charge of the gen sets had a strong port a power. I had him place it under the skid on the gen. end. I had him apply pressure so I could monitor the 50HZ. As he raised the whole gen set (from one end) the 50HZ did drop from .8 to .5 IPS.
I have not explored torsional resonance. Please explain it and how to test it. I use CSI 2120 equipment.

 

[GregLocock]

2.5 order is typically excited by cylinder to cylinder IMEP variation. This usually gets worse at low loads.

I think TVs are a good suggestion, you'll need a toothed wheel and a non contacting pickup and a TV analyser (FM demodulator) or a laser vibrometer if you have access to a rotating shaft.

50 Hz seems a little low for a crankshaft resonance - how long is the crank, and how is it coupled to the gen shaft? You say 'directly' is there any compliance in there? How long is the gen shaft?



Cheers

Greg Locock

 

[MachineryWatch]

Steve,

Torsional vibration is sometimes more of an issue on reciprocating machines because the impulse forces acting on the crankshaft can impart a twisting force to the rotating assembly. You can do an impact test to determine torsional natural frequencies like you do for other natural frequencies.

Find a place on the generator rotor where you can attach an accelerometer in the TANGENTIAL direction. If you have a 2 channel 2120 you can attach 2 accelerometers 180 degrees apart and compare the results of the impact. The impact should impart a twisting motion to the rotating assembly. If you use 2 accelerometers frequencies that do not appear in both sets of spectra should be considered to be from non-torsional sources.

In order to measure torsional vibration in the rotating assembly while it is running, you have a couple options. Set up an encoder signal of some kind from the shaft. This maybe from the ring gear on the flywheel, or optically triggering on the generator fan blades. Set up your 2120 to do a high resolution spectral measurement on that encoder signal. The area of the spectrum you want to examine is the pulse frequency of the encoder (number of teeth on the ring gear, or number of fan blades, times the shaft RPM). Torsional resonances will modulate this encoder fundamental frequency, producing sidebands around it spaced at the torsional natural frequency that is being excited. In your case you would be looking for sideband spacing at 50 Hz.

This method is the inexpensive way. There are other ways that are probably more accurate but will require quite an investment in additional instrumentation.

Skip Hartman

http://www.machinerywatch.com

 

[steve383]

Greg, please explain what IMEP is and why 2.5 orders is typically excited by cylinder to cylinder IMEP variation.
The crank is roughly 11’. The armature shaft is roughly 6’. It looked to be directly bolted to the fly wheel. There is no inboard bearing on the generator. The crank supports the gen shaft on the inboard side. The outboard bearing is about 1'in from the end of the gen shaft. It does have an exciter on the end of the shaft that is 11" in diameter and 6" wide.

Skip, This is a political situation where some one needs to pay for damages, repairs, etc. They hired me to collect vibration data only. I am not sure if they will call me back to do further testing. I am the type that needs to know what the issue is so it can be resolved.
They would be open to any ideas of what the cause is. They do not want to pay for my further investigation or testing at this point due to that it is the manufacturer’s problem. (Political)

Thanks guys!

 

[GregLocock]

IMEP - indicated mean effective pressure (I think).

Consistent cylinder to cylinder variation in IMEP in a 4 stroke produces 1/2, 1.5, 2.5.... order bending forces and torsional stresses in the crank and on the block. Depending on your running speed one of these harmonics will excite a resonance and bingo.

Cyl to cyl variation tends to increase at low loads for a whole bunch of reasons, at least on SI engines. It is a BIG refinement issue for us.

An 11 foot crank direct coupled to a 6 foot shaft will almost certainly have resonances down around 50 Hz, that is a driveline 17 foot long. Papers by T Priede of ISVR will have the observed natural frequencies of crankshafts plotted against crank length, I think.

The engine block/frame/generator system will also have resonances in that frequency range.


Cheers

Greg Locock