Understanding Dynamic balancers

[JMarkWolf]

I have a dynamic balancer I borrow from time to time, with which I balance the rotors on my helicopter.

It is a small hand-held self-contained unit to which you attach an accelerometer and a photo tach. After a data acquisition cycle, it will report the velocity of the imbalance and the phase angle at which it oocurs, on the integral 1-line LCD. This report typically looks like this:

"0.35ips @ 160 degrees"

You then manually plot this "dot" on a polar chart. The location of this dot on the chart "points" to the corrective action, which is either to add/remove weight from one blade or the other, lead/lag one blade or the other, or change the flight path of one blade or the other.

It works well.

I have fabricated a desktop rotor simulator from a small variable speed AC motor and an 8-inch diamter platter from a disk sander, with which I can capture and examine the waveforms generated by the accelerometer and opto-sensor, at the comfort of my own desk.

The acceleration signal is a near-perfect, near-noiseless sine wave. The 1-per phase marker strobes are crisp and clean.

The balancer instrument is connected to the simulator while I simultaneously acquire the same data with my digital storage oscilloscope. The balancer generates it's solution and reports the result on the LCD.

I have subsequently examined the waveforms generated by the simulator, integrated the accel waveform to velocity, and have tried to correlate the waveforms to the "report" from the balancer.

Disregarding the velocity amplitude, the phase angle reported by the instrument bears no obvious correlation to the peaks or zero intercepts of either the accel nor velocity waveforms.

Can anyone educate me a little as to specifically what analysis this instrument performs, that is not obvious by viewing and analyzing the waveforms? Or refer me to a consultant that can help me?

Like I said, I want to understand how it does what it does.

 

[electricpete]

The most common form of balancing involves adding a trial weight of known magnitude and physical angle. Then by observing the vector change in vibration you can compute an influence coefficient which can be used to translate your original vibration into a correction weight.

Is there some calibration of this unit that has been carried out previously? If not, all I can guess it that the the influence coefficient was determined ahead of time for this rotor and this measurement point and programmed into the unit.



=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Tmoose]

"Disregarding the velocity amplitude, the phase angle reported by the instrument bears no obvious correlation to the peaks or zero intercepts of either the accel nor velocity waveforms."

That's a good thing.

I'm guessing the photo tach reads a rotating shaft surface that you have to prepare with zones of contrasting light/darkness/shiny-ness. One of the transitions is "zero" degrees.

 

[Skogsgurra]

As long as you are below the critical speed of the assembly, my experience says that the acceleration signal peaks where the actual center of gravity is and that it is proportional to the unbalance mass. Above critical speed, difficult. Also depends a lot on what plane you measure acceleration in.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[JMarkWolf]

Tmoose wrote:
"I'm guessing the photo tach reads a rotating shaft surface that you have to prepare with zones of contrasting light/darkness/shiny-ness. One of the transitions is "zero" degrees"

This is correct. A piece of strategically placed reflective tape is used to trip the photo-sensor, and this is zero degrees.

My point is that none of the resultant waveforms seem to correlate to the phase angle report, referenced to this zero pint.

ElectricPete wrote:
"Is there some calibration of this unit that has been carried out previously? If not, all I can guess it that the the influence coefficient was determined ahead of time for this rotor and this measurement point and programmed into the unit"

This instrument is used on numerous types of aircraft, with various blade configurations, and the only difference is the polar chart on which you "plot the dot". This means that the instrument itself is a relatively general purpose data acquisition unit.

 

[sms]

The phase angle is measured from the point in time that the tach pulse fires to (typically) the positive peak in the wave form. This may be a leading or lagging angle depending on the instrument.

So on your scope put up the tach signal and the vibration wave form and considering positive peak to positive peak in the wave form as 360 degrees count the space between the tach pulse and a postive peak.

If you are looking at a raw acceleration signal, and your instrument is reading in velocity, then there will be a 90 degree error due to the integration.

"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

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

 

[JMarkWolf]

Thanks sms

This is the procedure I have been following, but no joy.

Will persevere.

Kelly Heroes, by the way, is a great movie. Always got a grin at Sutherlands 1950's era beatnik character in a 1940's period piece, however.

 

[Strong]

I would make sure your test setup is as simple as possible. Set the accelerometer and tach at the same clock position. Put a fairly heavy weight at 0-degrees where the leading edge of the tach marker is located. Run the rotor at a fairly low speed so that the heavy spot is close to the high vibration. The phase angle from accelerometer should be close to 0-degrees. Most accelerometers (ICP-type) have positive phase for vibration outward from the mounting surface (ie. heavy spot is going by). Some charge-type accelerometers are 180-degrees phase since they call positive as motion inward to the mounting surface. The phase angle from velocity should be 90-degrees, but may show -90. True velocity phase lags acceleration phase by 90-degrees (or +90 opposite the direction of rotation).

For balancing phase angle you need to keep track of:
Sensor phase (0, 180, or other)
Sensor and Tach orientation
Integration (accel = 0, accel->vel = 90-lag, accel->disp = 180-lag)
Tach Sensor polatity (some optical sensors have inverted signal)
Instrumentation Correction (if any)
Rotation Direction

The absolute phase angle is nice to know to find the heavy spot and for making a good trial weight placement, but it is not needed for trial run method, since phase is relative.

Walt

 

[Tmoose]

If I confirmed that the signal lag to be something other than what I consider logical or "natural" (using something similar to W Strong's method) I'd sleep well assuming it was done electronically and intentionally, and I might contact the mfr as to their intent, but only as a confirmation of my logic.

When I've configured a field balancing instrument for a particular repetitive job, and a consistency in heavy spot emerges, I'd include some correction weight placement lag (influence coefficient) in the procedure. It would make have been a little easier to twist a knob.
(

http://buy1.snapon.com/products/diagnostics/time_mt1241P.asp)