Torsional Vibration Problem, Need help, Strobe Analysis
Torsional Vibration Problem, Need help, Strobe Analysis
(OP)
I'm engaged in a project to fit a new engine into an air-powered vehicle. The engine is very different requiring an all-new setup. A sketch can be seen of the setup, attached to this post.
Initially, there were awful torsional vibration problems at low rpms. We added the centrifugal clutch to disengage below 2,000rpm before this speed is reached. This solved our low-rpm problems and enabled us to continue testing.
Now were are trying to analyze torsional vibration problems through the remaining range of power. We have been using our stroboscope to try to spot at what other rpm ranges is the torsional vibration at is maximum -- and from there determine the vibration frequency.
We believe the fan is the cause of the vibration. It has a very high moment of inertia. Our objective is to shoot the fan with the stroboscope at a set speed and "zero" the strobe in, so the image "stops". We then double the strobe frequency and re-observe. We were going to do this at about every 250rpm from clutch engagement thru peak power and record where torsional vibration is observed.
Since the fan diameter is over 30 inches, a small torsional flex of the shaft will visibly displace the fan tip, which should be easily recognizable with a stroboscope...
It seems the problem is, that piston engines don't run exact on the rpm. It always changes a little, in about a +/- 50rpm, regardless of the throttle setting. So we can never get the strobe to "close in" exactly on the rpm. As soon as we get the "picture" to stop, it starts moving the other way... We just simply cannot get the stroboscope to "zero in" enough to determine anything... Add on top of this, a screaming engine, flaming exhaust temperatures, a roaring fan blowing air all around you, and the possibility of a catastrophic failure with the fan guard removed.... and it's pretty damn hard to determine anything.
We don't have a large budget so a stroboscope is as high-tech as we can get...
Can anyone suggest any techniques or other methods to help determine our torsional vibration problems, without expensive technology?
Initially, there were awful torsional vibration problems at low rpms. We added the centrifugal clutch to disengage below 2,000rpm before this speed is reached. This solved our low-rpm problems and enabled us to continue testing.
Now were are trying to analyze torsional vibration problems through the remaining range of power. We have been using our stroboscope to try to spot at what other rpm ranges is the torsional vibration at is maximum -- and from there determine the vibration frequency.
We believe the fan is the cause of the vibration. It has a very high moment of inertia. Our objective is to shoot the fan with the stroboscope at a set speed and "zero" the strobe in, so the image "stops". We then double the strobe frequency and re-observe. We were going to do this at about every 250rpm from clutch engagement thru peak power and record where torsional vibration is observed.
Since the fan diameter is over 30 inches, a small torsional flex of the shaft will visibly displace the fan tip, which should be easily recognizable with a stroboscope...
It seems the problem is, that piston engines don't run exact on the rpm. It always changes a little, in about a +/- 50rpm, regardless of the throttle setting. So we can never get the strobe to "close in" exactly on the rpm. As soon as we get the "picture" to stop, it starts moving the other way... We just simply cannot get the stroboscope to "zero in" enough to determine anything... Add on top of this, a screaming engine, flaming exhaust temperatures, a roaring fan blowing air all around you, and the possibility of a catastrophic failure with the fan guard removed.... and it's pretty damn hard to determine anything.
We don't have a large budget so a stroboscope is as high-tech as we can get...
Can anyone suggest any techniques or other methods to help determine our torsional vibration problems, without expensive technology?





RE: Torsional Vibration Problem, Need help, Strobe Analysis
"Wind up" the fan a bit by hand.
Let go.
It will oscillate at its natural frequency.
RE: Torsional Vibration Problem, Need help, Strobe Analysis
OR
The strobe you have may have an external trigger input, that you can trigger with the engine. A couple turns of wire around the #1 sparkplug wire (with NO connection to the ignition circuit) might be enough. Or maybe you can pick off a lower level signal from the crank position sensor or whatever.
Of course, a fixed trigger gives you a fixed view, relative to the crank. To vary the view, you need a timing delay that you can adjust, and it would be better if it's slaved to the engine speed, so the delay is fixed in crank degrees, not in time. Something like that is already built into the fancier versions of automotive timing lights, hence my first suggestion.
Failing that, your local sparky could build such a PLL delay circuit.
Mike Halloran
Pembroke Pines, FL, USA
RE: Torsional Vibration Problem, Need help, Strobe Analysis
Alternatively you could fit stripy bands at various points on the shaft and use a photoelectric probe to measure the TVs.
Also look at crankshaft speed pickups or ABS wheelspeed sensors from cars, they can be used to measure TVs if there are any handy teeth to pick up.
Finally you can measure the surface speed of any shaft (more or less) using a laser tv meter. You should be able to hire one.
Modding the strobe with a trigger box seesm the cheapest to me.
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?
RE: Torsional Vibration Problem, Need help, Strobe Analysis
What is the belt drive pulley ratio?
What are the angles the u-joints are operating at?
RE: Torsional Vibration Problem, Need help, Strobe Analysis
have you looked into flexible couplings to isolate the fan drive from the engine?
RE: Torsional Vibration Problem, Need help, Strobe Analysis
Ed Danzer
www.danzcoinc.com
www.dehyds.com
RE: Torsional Vibration Problem, Need help, Strobe Analysis