Angle of Attack (and sideslip) Sensor?

[Trikoupis Nikos]

Hello to everyone!!

I am new to this forum and I would like to apologize if this topic has already been covered.

I am working on an UAV project this period.
Having searched a lot in the net I was not able to find a low cost AOA sensor. Only AOA which cost $$$$ for BIG airplanes.

Does anyone know of a AOA as a whole system or a pitot tube appropriate for this purpose. For example, reading articles I came across with a miniature 5-hole angle of attack and sideslip sensor. I could not find a place to buy it from.

Thank you in advance for your help!
Nikolaos Trikoupis

[space_case]

Making your own 5 hole probe will certainly be a challenge. A simpler approach would be to make pivoting 'weather vanes'. Add some small magnets and Hall effect sensor and you have yourself some sensors. Whatever you go with, I'm hoping that you have a wind tunnel for calibration.

[space_case]

Well I take back the comment about needing a wind tunnel with regards to calibrating the vane concept. Some careful measurements of the circuit at various deflections should get you close.

[blimpyway]

How do you feel about a flex sensor ? http://www.imagesco.com/catalog/flex/FlexSensors.html

One can imagine an assembly....

[space_case]

I'm sure you could get something to work using the flex sensor.

There are two problems I see with this (at least in the configuration I am picturing):

1. The angle of deflection would not only be a function of AOA but of airspeed as well - making calibration a pain, not to mention the fact you would need an airspeed sensor (if you were not already planning on having one).

2. There would be a potential for limit cycle oscillations.

[Trikoupis Nikos]

How do you feel about a flex sensor ? http://www.imagesco.com/catalog/flex/FlexSensors.html

One can imagine an assembly....

I was thinking about a flex sensor. Has anyone used this for an AOA sensor and talk about it? Could engine vibration cause a oscillation at the flex mass??? Will the result be valid? Of course the same problem may occur with the pitot but I believe the magnitude of the problem to be smaller. This is why I was looking for a ready-working AOA.

P.S. Ooooh space_case check the time :) :D. You where a bit faster :p

[space_case]

Will the result be valid?

Look at what I posted above and think about what causes the deflection. What forces are present?

[space_case]

P.S. Ooooh space_case check the time :) :D. You where a bit faster :p

OK. gotcha. Hall effect sensors are the way to go. That or very low friction pots.

[Trikoupis Nikos]

Look at what I posted above and think about what causes the deflection. What forces are present?

1) Air pressure force = f(airspeed,AOA, ... ?)
2) Inertial acceleration = g(mass,acceleration,engine vibration to airplane body,...)

Something more. How flexible are these sensors? Could only one sensor be suitable for both low and high speeds? (e.g. slow landing , hard turn?)

P.S.Again!!! What exactly do you have in mind with hall effect sensors? What are friction pots?

[space_case]

1) Air pressure force = f(airspeed,AOA, ... ?)
2) Inertial acceleration = g(mass,acceleration,engine vibration to airplane body,...)

Something more. How flexible are these sensors? Could only one sensor be suitable for both low and high speeds? (e.g. slow landing , hard turn?)


This is what I was getting at. You have the force of this 'flex sensor' device to contend with. There may be some more specs on it somewhere but all I see directly is '5 grams of force to flex 90 deg' and the range of resistance. Perhaps the design is such that it takes little force to deflect it at small angles and the resolution in that area is such that it would work fine for a reasonable range of AOA.

If you design and manufacture your 'vane' correctly, pivot angle should be independent of airspeed. The best approach would be to measure this angle as unobtrusively as possible (ie - not imposing any other forces on it).

P.S.Again!!! What exactly do you have in mind with hall effect sensors? What are friction pots?

By pot I mean potentiometer.

Also, if you set it at 0 deflection = 0 AOA - how will you differentiate between +/- AOAs?

[blimpyway]

I dont know of any use of a flex sensor for such device. I ordered a pair for a different project. They are like thin (about 0.5mm or 1/50 in), 4in long plastic sheets. They-re quite light.

I see three possible problems with them - vibration, flutter and inertia.
With first two problems mabe computed mediation over many readings might help.
But to compensate acceleration.... you need to know what the real acceleration is.
And there-s different wind speeds. problem.. need to measure and compensate it too ....yuck.

An interesting rotation sensor is inside most computer mouses. Those with a rolling ball have two tiny wheels with many gaps crossing between an IR emiter/sensor pair
PS2 mouse especially are interesting since the protocol is serial, 5V based and a microcontroller can read rotation values directly.. One 360 degrees rotation are read as approx 50 units of movement.. That means it has 7.2 degrees accuracy.. Is it too bad?

The sensor might be used to make around it a balanced, low friction, low inertia pivoting wind vane..

[Blue Sky]

How about a moving vane capacitor as a sensor?

-Dave

[deh6]

Analog Devices has new chip: capacitance-to-digital converter, AD7745,6,7. Looks interesting with regard to a capacitance-vane AOA sensor.

[Blue Sky]

Neat chip!
It can read very small amounts of capacitance so the sensor can be kept small.
And plain ol' air is a near perfect dielectric for use in it! (Air has very low conductive
losses).

[ashvani]

Hi,
Its a quite old thread but my question remains the same. Gentlemen with UAVs, how are u measuring Angle of Attack and Side slip angle for the control law of ur UAV?

I need a system for my RC plane for the same to do in-flight tests.

Thanks in advance

Ashvani...

[deh6]

I have a Hall-effect sensor hooked to a HCS12 micro on my bench at the moment. I have not flown it yet, but it is close. The measurement is a vane with a small neo magnetic on the vane shaft.

A LED/photodetector might good also, but I'm trying the Hall-effect first.

Gary Weaver has been using an accelerometer for side-slip. This has been on gliders so some attention to the vibration from a motor would be needed for powered planes.

An issue I discovered recently is worth noting--Getting the derivative of the signal. Derivatives are inherently noisy. If one takes the simple approach of taking the difference between successive samples for the derivative, all sorts of jitter get added to the servo signal. Theoretically the noise is not a problem because it is damped out in the system, but inbetween, the servo is chattering away, sucking battery power, and wearing gears out, likely creating drag, etc. My approach was to implement a differentiating fir filter where output is proportional to the derivative of the input up to a given frequency then drops off. Investigating this has been another one of those side-trips that have delayed getting to the air.

[Tom Harper]

Here's a simple AoA sensor that I designed after the ones I noticed on the noses of Boeing aircraft.

I used a feedback potentiometer out of an old servo. These are small and very free moving. I drilled a hole in a piece of spruce to exactly match the shaft so that it acts as a bearing. Then I pressed the pot to the back ot the wood strip and ran glue around the outside. If any gets inside the pot may bind.

For a vane I used another spruce strip with a balsa vane on one end and a counter weight on the other. This was not a finished piece. It was a first try and it worked fine. Next time I will finish and paint it but I think the TE should be square instead of tapered.

The travel of the unit has to be limited. I used toothpicks in my foamie.

The pot is about 5K. I designed a 555 circuit to give me the 1-2ms output of a RC rcvr. The limits (toothpicks) are set with a scope to give the 1-2 ms span.

I plugged this one into an Eagle Tree data system. Fired the SS with the rudder output of the rcvr in case the Eagle Tree was using synchronous sampling.

It's cheap, it's accurate and it works.

Tom

[ashvani]

Thanks Tom!

Whr is this vane mounted on the plane?

[Tom Harper]

Ashvani,

This model had a long nose for camera gear. The vane was mounted on the flat fuselage side forward of the wing. I don't think it is critical. It just has to be clear of any large surfaces.

[MattChave]

ok how about sensing the stagnation point on the wing with multiple reed sensors, or more expensively a pressure sensor just below and behind the leading egde of the wing, see the below link for the description:

http://www.av8n.com/how/htm/airfoils.html#sec-stall-wrng

good luck :)

[deh6]

markdrela's comment in the following thread might be useful:

http://www.rcgroups.com/forums/showthread.php?t=402653

[RajeshRaman]

I used a feedback potentiometer out of an old servo.

Tom, a newbie question for you!

I assume these feedback potentiometers are light, accurate and have a very low torque. Are they available from some hobby or electronics stores, or does one have to sacrifice a servo for this?

[Tom Harper]

Rajesh,

They may be available, I don't really know. I have a pile of wounded servos that sacrificed themselves to various causes. Usually they strip their gears so the electronics are still OK.

If you do sacrifice one be sure to keep the electronics board. You can substitute a couple of 10K resistors for the pot and get a handy switch for turning things on and off with the rcvr.

Tom

[coafin]

Why do you want to know AOA ?

For landing, or avoiding stalling or what ?

There are usually better solutions.

[t_j_walker]

Here's a simple AoA sensor that I designed after the ones I noticed on the noses of Boeing aircraft.

The problem with these is that you really only know your AOA in level flight unless you know your bank angle and acceleration. One of my professors from college has a patent (I believe that it is expired now) on a pressure port AOA system. Here is the link to the patent (http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=14&f=G&l=50&co1=AND&d=PTXT&s1=hoadley.INNM.&OS=IN/hoadley&RS=IN/hoadley) if you would like to see it. It is for full scale aircraft, but I believe you could adapt it. A couple problems I see are resolution on differential pressure transducers and putting pressure ports on the wings.

Tim