Torsional Vibration of a diesel engine

[steve383]

I have CSI 2120 vibration equipment. A customer wants to know if I can take torsional vibration data on their Walkasha deisel engine. Can any one tell me what he is looking for and can this be done with an acellerometer and tach?

 

[enginebob]

Not familiar with that equipment you have. But
The torsional vibrations are the twisting and untwisting of the crank shaft. The greatest amount of vibrations will be at the opposite end where the power is taken from. That is most always the end opposite the flywheel. So to show the overall variance from the flywheel end you would have to have a sensor on both ends. You want to measure the twisting in plus and minus degrees from the zero point. That of course is taken with the engine stopped.

 

[GregLocock]

Usually people measure TVs on engines using a toothed wheel (60-120 teeth), a non contacting pickup, and an FM demodulator, the ONLY one I have seen in use is made by A&E (I think) and looks about 40 years old. The last sounds very techie, all you actually need is a frequency to voltage converter, and then look at the variation in the voltage out. Calibration is another issue entirely.

So no, you can't use an accelerometer very easily.

Cheers

Greg Locock

 

[steve383]

Thanks for the info guys! I was hoping I could use an acellerometer, put reflective tape on the dampener, use a tach and collect a high resolution spetrum. I was hoping the torsional vibration would show up as side bands around tach sensed turning speed. Is that just wishfull thinking?

 

[GregLocock]

yup, that's wishful thinking.

first order off the tape is completely useless. In fact first order TVs are pretty much useless in general, unless you take a lot of care with eccentricity and so on.

Having said all that, do Onno Sokki make a TV analyser attachment? Wouldn't have to cost much, and they are a very cluey company.



Cheers

Greg Locock

 

[Rob45]

There is a far simpler method that gives useful results, and it's what I use frequently on Diesels in large trucks (10L to 15L displacement): use a mag pickup to count flywheel teeth, then run a single differentiation to get the angular acceleration, which is what torsional vibration is all about.
I've been doing it with an earlier version of Head Artemis software, which is a little cumbersome but gets the job done. Requires that I edit the data file, dividing the actual number of teeth by 2xpi, so the final answer is in radians/sec/sec.

Good luck;
R

 

[GregLocock]

Rob, isn't the flywheel nodal for the important crank torsional mode? Or do you still get enough amplitude to get a fix on the frequency?

One of the projects I was involved in measured the TVs at different points along the driveline, very interesting stuff.

Cheers

Greg Locock

 

[Rob45]

Greg:
(Started to say, "what are you doing on at this time of night?" but you're in Oz, aren't you?)
I'm sure you're right and the flywheel's a node for certain frequencies, but the other masses especially the crank damper and the FEAD are so large that much comes through, and I can even detect some camshaft and geartooth modes at the flywheel! Also tells me a little about clutch torsionals.
For drivelines, I use the same method, but read the transmission output shaft speed via a 12T tone ring, to look at the results of u-joint angle non-cancellation as well as shaft imbalance effects.
- Robert