rotor 1/ rev vibration and airspeed

[Mike G]

This is a question from a friend who has developed a propeller balancer and is trying to convert it to a rotor balancer for autogyros.
He knows that when autogyro airspeed increases, the vibration at two times the rotor frequency increases due to the difference in drag between the 12 o'clock - 6 o'clock position and the 3 o'clock - 9 o'clock position.
His question is:
is there an increase in vibration at rotor frequency (rotor rpm) with increase an in airspeed?

I think I know the answer but want to check with the gurus if I'm missing something.
Mike G

[Oskar]

Mike,

For the benefit of everyone I’ll just sketch the background to where this is coming from.

I am busy testing one of Mark Burton’s PB3 balancers on a MTO Sport. The PB3 was originally designed for prop balancing, but with the correct software algorithm there is no reason why it shouldn’t work really well for rotor balancing as well. The hardware implementation is superb, it uses an Android phone to display the results very nicely. Check out more at http://http://www.smartavionics.com/pb3/pb3.html

I originally used a DSS Microbalancer to balance the blades. The results were:
As from the factory: 0.35 ips, could not let go the stick, compass turned continuously.
At 0.2 ips I could let go the stick (still lots of shake), compass still turned.
At 0.1 ips the compass stopped turning.
I ended up with 0.07 ips which gives comfortable flying and a compass that works.

The original PB3 software version (designed for prop balancing) gave me a vibration of 0.7 ips. A few iterations and many emails later we are at the polar chart below. Points 1 to 4 were measured at about 60mph, 5 to 7 at increasing airspeed up to 90mph.

Now getting back to your question:
All the numbers above refer to 1 per rev.
Rotor drag due to airspeed is a 2 per rev phenomenon, drag is lowest when the blades are fore/aft and highest when the blades are left/right. This happens twice per revolution and thus initiates a 2 per rev vibration.

For the sleuths:
With the PB3 I measured vibration fore/aft, with the DSS microbalancer I measured vibration left/right. The reason was because it was easiest to mount the sensors this way. In theory there will be a difference in the 2 per rev results, but the 1 per rev results should be roughly the same. Maybe, maybe not?

Oskar

[Mike G]

Oskar
I am also playing with Mark's PB3 and am currently testing it as a prop balancer in parallel with the balance equipment I got from work before trying to use it on the rotor. Also I'm totally useless with this damned touch phone.
I'm afraid rain stopped play here in Normandy so I haven't got any further with it.
I put your/Mark's question on the French forum and they came back with the answer that the 1/rev vibration that increases with airspeed is the vertical one due to tracking. If you've got the latest PB3 accelerometer it can plot vibrations in two directions. Can you set it up to read fore/aft and up/down?
Mike G

[Oskar]

Mike,

Hadn't even got to vertical, still trying to sort out the horizontal...

Autogyro specifies that the vertical vibration should be measured at the front seat. Mine was around 0.07 ips from the factory and I cannot remember how much it changed with airspeed. Being so low it didn't really matter.

Also I don't have the 2 axis accelerometer.

Oskar

[Mike G]

Oskar
I wonder why they want to measure it at the front seat, it's the rotor that vibrates. Can you show a photo of your installation? If not PM me.
Mike

[ckurz7000]

Oskar
I wonder why they want to measure it at the front seat, it's the rotor that vibrates. Can you show a photo of your installation? If not PM me.
Mike

Mike, you want to measure lateral vibration as close to the rotor as possible and vertical vibration as far away from the CG as possible. The reason being that the vertical vibration is amplified by the virtual moment arm, the length of which is the distance from the CG to your accelerometer.

-- Chris.

[Mike G]

Chris
I'm well aware that the vibration should be amplifed by any moment arm, my question is why would you want to amplify the vertical vibration when you don't amplify the horizontal?

Also I would have thought that putting the accelerometer at the front seat introduces the possibility of picking up all sorts of other vibration scources (engine, pod aerodynamic etc.) and resonances that are nothing to do with the rotor.

Do Autogyro give you any instructions about how to correct a vertical vibration measured at the seat?
Mike G

[ckurz7000]

Mike,

picking up the lateral vibrations as close to the rotor as possible already maximizes the (vertical) distance from the CG. You would be placing the lateral accelerometer pointing left-right up the mast near the rotor and the vertical one forward in the cabin near the pilot's seat. With this placement you minimize crosstalk between the two while maximizing sensitivity. The best placement of accelerometers for vibration measurements is also a thing of trial and error. I can't really justify it more except by pointing to generally accepted procedures and the statements of a few experienced mechanics.

Pick-up of stray vibrations is usually not such a great concern, unless they have a 1/rev component, because the analyzer has a narrow bandpass filter centered around the rotor frequency. Engine vibrations are much higher than the rotor rpm.

If, in your machine, the standard placement doesn't work, you have to experiment and find a better one. There are no guarantees.

Do Autogyro give you any instructions about how to correct a vertical vibration measured at the seat?

To correct vertical vibrations, you have to change the angle of incidence of the blades. With an Autogyro rotorhead you shim one side of the hub so that it sits at a very slight angle. This reduces the pitch on one blade while increasing it for the same amount on the opposite side.

-- Chris.

[Mike G]

Chris
I see the point you're making but will install mine at the rotor head anyway.

Magni don't give any instructions except to increase the friction of the roll and pitch bolts. All that seems to do is reduce stick shake and increase cabin shake.

Any chance of a photo or sketch or drawing showing the Autogyro hub with shims, I have difficulty imagining your set up.
Mike G

[ckurz7000]

Mike, unfortunately no Photos for the shim. But it's really very easy. The goal is to tilt the leading edge of one blade down while the other one is tilted up. You do this by adding a piece of thin aluminum sheet metal under the teeter tower. I don't know how this would best be done in a Magni rotor head but there must be a similar solution.

-- Chris.

[Oskar]

Mike,

You can find the correction phase plots from Autogyro here.
http://www.auto-gyro.com/chameleon/outbox/public/27/Polardiagramm-Vibrex.pdf

These are for the sensors mounted as specified, on the top of the mast for lateral and under the front seat for vertical. They have already done the hard work, no need to reinvent the wheel.

Oskar

[ckurz7000]

Oskar, the way how to mount the sensors is one thing. The other is how you define your zero degree point. It seems to me that for this template to be correct, zero degrees is at the 3:00 o'clock position. Can you comment on this?

-- Chris.

[Oskar]

Chris,

You're right, the way the charts are printed 0 degrees is with blade 1 pointing to the right when sitting in the gyro. I just turn the piece of paper 90 degrees and then 0 degrees has the blades lining up with the gyro. Just feels more comfortable that way :rolleyes:.

Of course you also need to know what you're doing when deciding where to mount the position sensor.

[ckurz7000]

...Of course you also need to know what you're doing when deciding where to mount the position sensor.

I just am pointing this out because a mistake in this respect has cost me a day of data taking once before. That was one of those times when I later thanked evolution for making it physically challenging to kick my behind.

-- Chris.

[Mike G]

Chris and Oskar
Can you explain how you use these charts? I'm used to using polar charts when I balance machinery so I understand how they work you don't need to waste your time explaining the principle .
I understand that the top one tells you to move the rotor right or left (12 o'clock/6 o'clock) along the teeter bolt or adding weights to blade 1 or 2 (3 o'clock/9 o'clock). And the lower plot is for tracking (so vertical vibration at seat) but I would like to see the instructions that go with it especially on which blade do you put the reflective tape (I assume blade 1) and where exactly do you fit and in what direction do you point the accelerometer for the simple radial balnce mode (ie the top chart). Also do they give you any guidance regarding how much weight or shim to add for a given amount of unbalance? Also what are the units IPS?
Mike G

[Mike G]

Chris
Attached is a sketch of what I understand you are telling me about adjusting the pitch for the MT03. Have I got it right?
Mike G

[ckurz7000]

Chris and Oskar
Can you explain how you use these charts? I'm used to using polar charts when I balance machinery so I understand how they work you don't need to waste your time explaining the principle .
I understand that the top one tells you to move the rotor right or left (12 o'clock/6 o'clock) along the teeter bolt or adding weights to blade 1 or 2 (3 o'clock/9 o'clock). And the lower plot is for tracking (so vertical vibration at seat) but I would like to see the instructions that go with it especially on which blade do you put the reflective tape (I assume blade 1) and where exactly do you fit and in what direction do you point the accelerometer for the simple radial balnce mode (ie the top chart). Also do they give you any guidance regarding how much weight or shim to add for a given amount of unbalance? Also what are the units IPS?
Mike G

Mike,

Also what are the units IPS?

IPS is short for inches per second and measures the product of amplitude and frequency of vibration. Imagine a perfect disk spinning around its center with a mass stuck 1" off its axis, which is spinning at 60 rps (=360 rpm). The amplitude is 2" and the frequency is 60 per second. This corresponds to 120 IPS.

Also do they give you any guidance regarding how much weight or shim to add for a given amount of unbalance?

No, you have to experiment. Essentially you do a datarun with the bare rotor and plot it on chart. Then you add one unit (e.g., a small washer) of tip weight to one blade and do another datarun, and plot it on the same chart. From the shift between point #1 and #2 you can then deduce the effect of the unit tip weight, find the correct weight and do datarun #3.

As for shimmin, I have seen thin aluminum sheet metal of about 0.1 mm thickness. I have also seen several of them stacked.

where exactly do you fit and in what direction do you point the accelerometer for the simple radial balnce mode (ie the top chart).

Here is an old picture from 2006 showing the hall sensor used to pickup rpm and define the 0° position, as well as two lateral pointing accelerometers. Currently Autogyro is, as far as I know, using only a single left-right pointing accelerometer mounted at the same location. A vertical accelerometer is mounted near the pilot's seat.

88110

And regarding your last post with the sketch of the shim position: this is not what I meant. You don't shim the cheek plate. The location to put the shim would be on, e.g., the right half of the lower dotted line in your diagram. This would be the aluminum block with the teeter bearing, which is mounted to the blade holder by 4 bolts. You place the shim underneath two of them.

-- Chris.

[Mike G]

Chris

I wasn't asking "what are IPS" but "are the units IPS", sorry for ambiguity.

Thanks I understand better now the shimming see attached.

Does the hall sensor pick up the bolt head in the photo (see attached)?

Mike G

[Mike G]

I'm still thinking about this. I know it's obvious and I'm very slow.:bored:

If, say, blade 1 has a greater pitch (hence lift) than blade 2.

Between what angles (using the clock format assuming 12 o'clock for the forward direction) will blade 1 give the extra upward force that would be detected as an unbalance by the vertically orientated accelerometer and at what angle (time) would this force be at its maximum?

Mike G

[ckurz7000]

Chris

I wasn't asking "what are IPS" but "are the units IPS", sorry for ambiguity.

Thanks I understand better now the shimming see attached.

Does the hall sensor pick up the bolt head in the photo (see attached)?

Mike G

Mike, I don't get the question. IPS = inch per second ARE the units. You can also measure them in centimeter per second, if you feel so inclined.

The shimming is still not right. The shims go above the top aluminum bar. You have them above the bottom one. The sit between the top bar of the blade holder and the aluminum block that holds the teeter bearing.

And yes, the hall sensor picks up on the head of a magnetic bolt inserted into one of the lightening holes of the big prerotator gear.

-- Chris.

[Jincamty]

This is a great thread! Thanks Guys.
Cam.

[Mike G]

Chris
Are the units on the polar chart IPS? Sorry I wasn't clear.
Mike

[ckurz7000]

Chris
Are the units on the polar chart IPS? Sorry I wasn't clear.
Mike

Yes. -- Chris.

[Mike G]

Back to the subject.

If blade 1 has a greater pitch (hence lift) than blade 2.

Between what angles (using the clock format assuming 12 o'clock for the forward direction and anticlockwise rotation looking from above) will blade 1 give the extra upward force that would be detected as an unbalance by the vertically orientated accelerometer and at what angle (time) would this force be at its maximum?

Mike G

[bryancobb]

Hi Mike G,
Let me rudely butt-in here. Please don't get angry, I just feel like I can add a little to the conversation.

There is a number of things you can do to any rotor system that will create or eliminate a particular vibration. I.E. pitch or depitch a blade (incidence) : bend trim tab(s) (if present) : Add weight at any number of points : Shift chordwise CG by leading/lagging a blade or shifting the entire head with shims. These are just a few. I'm sure some ships have many more. And...keep in mind, a perfectly-tracked rotor MAY NOT neccessarily be the SMOOTHEST rotor.

Each one of these VARIABLES, makes the data-point on the polar chart move in a different direction, when you make changes. EVERY AIRCRAFT IS DIFFERENT. The only way another person's chart can be used to balance YOUR aircraft is if the two aircraft have almost identical rotating and controlling parts.

The clock position readout for where a vibration is occurring doesn't have any relationship to the nose of the aircraft. It is related to where the accelerometer is mounted and where the photocell pickup is installed, and which blade has the reflective tape on it. You can install the accelerometer pointed ANY DIRECTION and install the photocell ANYWHERE as long as the reflective tape triggers it. Just be aware, your move lines on the polar charts will only be correct when these two devices are installed as they were when the charts were developed for YOUR aircraft.


Let's say you install the accelerometer up at the rotor head, pointing left, to measure lateral vibration. Then let's say you put the photocell about 12" in front of the mast, with a piece of reflective tape on one (master) blade.

Now say you built the ship from scratch and it's the only one that exists. Say you want to determine how adding or removing tip weight to the master blade makes the data point move. You can determine the azimuth of a move line on the polar chart by taking 2readings. One at rpm as-is, then plot it, then add tip weight to the master blade and plot it. That gives you move DIRECTION in relation to the accelerometer and photocell.
You can then plot intermediate points with different amounts of weight so that you end up with A) Move-Line Azimuth .and B) Weight/Distance Correlation.
This procedure can be done with each thing you have the capability to change on the rotor.

The concept to "take to the bank" here is... Your move lines and amounts of change per unit, ARE NOT GOING TO BE LIKE SOMEONE ELSE'S unless your ship is and exact twin of theirs and you have the accelerometer and photocell installed exactly like they did.

[ckurz7000]

Thanks, Brian. That's what I tried to say a couple of posts back. To do a good balancing you need some theoretical understanding but mostly you try a "unit" displacement of each of your tuning variables and plot them on the chart. Then, as you already said, you see in which direction and how far each one takes you. And with some experience and these data you can figure out -- by linear superposition -- which combination of adjustments will get you the smoothest rotor.

-- Chris.

[Mike G]

Brian and Chris
I really appreciate your input, please bare with me for a bit longer.
I actually do understand a bit about balancing but I like to keep my questions as simple as possible and sometimes it gives the impression that I'm out of my depth (sometimes I am).

http://www.mediafire.com/?erbl3v9n0xkgbjk

is my understanding of the signals you would get on an oscilloscope if it where monitoring:

1) a sensor that signalled the passage of a passing bolt (hall sensor) or the passage of a bit of reflecting tape (as Vibrex does).
and
2) a vertically mounted accelerometer orientated such that it read positive accelration when rising and negative when falling.

and assuming that blade 1 had a greater pitch than blade 2 with the direction of rotation and flight as shown.

I have shown the plan view of the rotor in 4 positions with the readings I THINK we should get and a final sketch showing the screen after a few revolutions.

This is, of course, grossly over simplified and assumes no other scources of vibration except that cause by Blade 1 having a greater pitch than blade 2.

Do you agree with my sketches?

Mike G

[bryancobb]

Mike,

You say you wish to keep your question in simplest terms. Well, your understanding that you used to develop that simple question is flawed.

If a blade with more pitch, passed the sensor, and caused an upward vibration that could be picked up by the accelerometer and displayed on an O-Scope, then moving the cycling would be changing blade pitch and thus, the vibration all the time. The vibration azimuth and magnitude would be all over the chart.

A vibration, by definition, is a back-and-forth oscillation. The algorithims and accelerometers that are used to balance rotors don't even detect the forces (caused by more or less pitch) you are ferring to.

I couldn't open your Visio sketches. Print them to a freeware PDF printer and I can open the PDF file or do a screen capture and post it as a JPG.

[Mike G]

Bryan
I have understood from various threads on this forum that if you have a tracking problem (one blade pitch set greater than the other) the result is vertical 1/rev vibration "like a galloping horse". Of course I'm talking about two bladed gyros.
What I'm trying to clarify is the relationship between the magnitude of this vibration and the angular position of the blade that has the greater pitch.
I've converted my sketches to pdf and hope it makes more sense now
Thanks for being patient..
Mike G

[C. Beaty]

Out of track shake is quite simple, Mike.

Out of track: one blade has slightly more pitch than the other. It flies higher by an amount that can be determined by simple geometry. *

When the high blade is on the advancing side in forward flight, total lift is slightly greater than when the low blade is on the advancing side.

Viewed on an oscilloscope that responded to amplitude, the vibration would be a near sine wave with maximum amplitude when the high blade is at 3 o’clock. And of course, the faster you go, the greater the shake.

Out of track also means the rotor thrust vector scribes a conical path. This will stir the stick in a 1/rev circle with severe out of track.

********
*If the difference of pitch is 0.1 degree and the rotor radius is 14 feet, the blade tips will fly with a separation of 0.29 inches. (tan0.1 * 14 * 12)

That’s the reason shims made from beer cans are so useful for tracking rotors.

I don’t have the foggiest notion of how a Magni rotor is tracked but surely, there’s an eccentric bushing or similar surrounding the teeter bolt.

[bryancobb]

Mr. Beaty,

If I remember correctly, you are a highly experienced gyro pilot with an in depth understanding of gyro aerodynamics. I will now attempt to delete my previous post since it was obviously made from a position of100% ignorance.

[Mike G]

Chuck
Thank you once again for putting me straight.
To conclude:
Does that mean that with out of track blades I can expect a vertical vibration at 1/rev frequency as shown in my attachment plus a horizontal 1/rev vibration that would be at 180° to the blade with the greater pitch (ie. when the blade of gretaer pitch is in the 9 o'clock positionj, the horizontal force will be in the 3 o'clock position?

From the other thread that, I started because I thought I wasn't getting your attention, never having galloped a horse, I presume that a galloping horse is at about 5 hz?

Mike G

[C. Beaty]

Mike, the rotor thrust vector, in the case of a seesaw rotor, is perpendicular to the hub.

With one blade high and the other low, which way does the thrust vector point relative to the rotorhead?

It is easily confused with out of balance.

[Mike G]

Chuck
Thanks again.
When you say "hub" do you mean the bar that connects the two blades or the part that includes the bearings and teeter towers?

Mike G

[C. Beaty]

The hub is the thing that connects the two blades together.

[Mike G]

Chuck
Thanks

So back to the first question of this thread.
"What 1/rev vibration increases with airspeed?"

The answer seems to be that out of track blades give

1) a 1/rev vertical vibration that if plugged into a simple balance machine would look like the blade with the most pitch was heavier than the other.

and

2) a 1/rev horizontal vibration that if plugged into a simple balance machine would look like the blade with the least pitch was heavier than the other.

Both of these vibrations increase with airspeed.

Attached is a revised version of my previous attachment which had 2 errors in it.

Mike G