Test measurement of torque pulse on tractor engine

[CaptainCrunch]

I've been asked to measure torsional vibration due to the torque pulse of a two cylinder engine on a tractor. The idea is to isolate the effect of torque pulse and determine the effectiveness of a torsional damper in reducing this effect. I'm visualizing the accelerometers 180 degrees apartand measuring the circumferential vibration. Does this sound right?

Has anybody done this before? I link to a paper / test method would be helpful.


Thanks in advance for help.

George

 

[GregLocock]

60-240 tooth tonewheel is the usual approach, followed by an FM demodulator, (aka TV analyser).

Your accelerometer idea will work, but you'll need a way of getting the signall off the crank to your instruments. Slip rings are noisy, wireless may be a bit big.

Interesting stuff, TVs. The bible is by Kerr-Wilson.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Strong]

George,

I use a strain gage telemetry system for measuring torque and torsional vibration.

http://www.binsfeld.com

I also use an optical or magnetic encoder with a frequency to voltage converter to measure torsional vibration as angular velocity. The engine flywheel with starter gear teeth may serve as a target for a magnetic sensor. This may not be the best location, since it is on the engine side of the torsional isolator/damper.

The choice of measurement method, sensor type and measurement location can depend a lot on access to the drive system.

Walt
w_f_strong [at] msn [dot] com

 

[GregLocock]

No, you can't easily use the flywheel as it is nodal for the mode of greatest interest.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[rob768]

You measure torque by strain gauges on a straigh shaft section. You can measure the vibration on the damper housing by using suitable laser equipment. The vibration will differ considerably after the flywheel (is there also an elastic coupling between engine and driven equipment?), so you cannot so anythin on damper load from the strain gauge measurements. You'll also need a torsional vibration calculation to interpret the vibration results with regards to damper load. Not the easiest stuff, if you never done this before.

 

[MikeyP]

We would use a high line count (depending on speed) optical encoder with either a frequency-voltage converter or a high-speed counter-timer to estimate velocity fluctuations.

M

--
Dr Michael F Platten

 

[GregLocock]

Yup, f/v converter is an fm demodulator.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[btrueblood]

Mike, Greg

I like the f/v and FM approaches, but I've done it "digitally" too when timely acce$$ to those methods was not available - just record a data stream of the optical encoder output (volts vs. time, or 1's and 0's vs. time) and post-process with some c/c++ code to find apparent position, speed, and angular acceleration vs. time. Big whopping data files, but storage is cheap. Writing the code was painful, but once done it came in handy quite often. Nowadays, with fourier analysis built-in to Excel, it would be a bit easier yet.

 

[GregLocock]

Yes, that'd work fine. I thought about doing the same thing - it would be nice not to have to calibrate the analyser, and as you say they are pretty expensive. Also you could use a notched wheel (ie one tooth missing) which would be handy.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SomptingGuy]

Inductive probe targeted at flywheel teeth is one of the simplest methods. Forget Excel, use something stronger like Matlab and use hilbert transforms to do your demodulation.

 

[btrueblood]

Dang, Greg. Yes, that was exactly what I did (had one larger notch), to get a timing index. Actually, I had the photodiode sensors from some junked encoders, and printed out my own custom tone wheel on a laser printer to acetate overhead projector sheet material.

Why do it the hard way like that? I was young, there was no money or time, and... well, it worked.

Still, I like the FM approach and the f/v approach, these would be simpler and probably yield better results. I like even more the fact that other, much brighter people like Mike and Greg have used similar techniques, for good reasons. I once tried to convince people at another job that the technique would work (and it would have) for a very high-rpm measurement where it was questionable if slip rings or wireless techniques could be made to work due to the imbalances the devices would impose on the system.

 

[GregLocock]

I once did a complete engine and driveline, from crank pulley right through to rear wheels. I wish still had that data!

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.