Airspeed triangulation?

Lawson Laslo

Pre-takeoff checklist
Joined
Jan 18, 2019
Messages
403
Location
Sundance airport Oklahoma
Display Name

Display name:
N2005H
Hello
I am wanting to test my airspeed indicator and have heard some time ago that there is a method where you fly in 3 directions recording gs and direction and you then plug it into formula and it gives you a airspeed. But I cannot find the formula? Any Ideas?
Thank you
 
Ask NASA -
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20110015011.pdf

Or more perhaps digestible -
https://www.kitplanes.com/gps-assisted-airspeed-calibration/
"An extension of this idea is to fly three legs at the same altitude and airspeed. The three-leg method eliminates the need to precisely align the flight path with the wind. A simple computational procedure (see, for example Advisory Circular 23-8B, FAA, 2003) allows one to compute true airspeed, wind speed, and wind direction.
... four or more legs ...
six legs ...
seven legs of..."
 
Last edited:
Basically high school trig. And I already forgot it.
 
I never ever took trig, is that even better then learning it and immediately forgetting it?

Probably. I mean, I took upper level calculus and differential equations too, and if you asked me to do a problem today, I'd need more than an IPC to rectify that lost scan. Nowadays software does everything for you anyway.

In the famous words of a former ME professor, 'This is the equation you show the client, and this is how you actually do it.'
 
There are several places with spreadsheets you can download to do the math, but the math isn't that difficult if you've had a trig class. When work is busy I average ~3-4 airspeed cals a week (we use them to estimate winds aloft), if you'd like some help I'd be happy to contribute.

Nauga,
whose source error is corrected
 
If you're as lucky as me and you have an advanced panel with something like my G3X and GTN650, is there any reason to think that the KIAS, KTAS and GS are not at least accurate to within 1 knot of speed ?
 
IAS and 'accurate' are contradictions ;) IAS by definition requires a correction to get CAS.
The TAS estimate depends on the accuracy of the wind estimation - I don't know what algorithm Garmin uses but a high(er) precision method is usually a few knots away from the G3X and G500 I'm familiar with. It's 'close enough' but I wouldn't bet TAS is always accurate to within a knot.

Nauga,
who misses tower flybys
 
I did a lot of research into this some years ago, reading some of these papers and articles, and collecting meticulous data. I finally decided that for my purposes, flying a 180 hp piston single, the best use of my time was to just fly a level 360 turn, note max and min ground speed, and calculate the mid-point. If you need more accuracy than what you can read off of your GPS, or read on your analog airspeed indicator, then those more involved methods may be worth your while.
 
flying a 180 hp piston single, the best use of my time was to just fly a level 360 turn, note max and min ground speed, and calculate the mid-point.
The high-precision winds aloft estimation I mentioned in post 11 does this but uses logged data from the turn to extract many points around the turn.

You can do the same thing with any number (>=3) of level legs but the math is more complicated than the three-leg spreadsheet method.

Nauga,
the least square
 
I just read the wind direction and TAS from Aspen. That’s in my wheelhouse for math.


Sent from my iPhone using Tapatalk
 
IAS and 'accurate' are contradictions ;) IAS by definition requires a correction to get CAS.
The TAS estimate depends on the accuracy of the wind estimation - I don't know what algorithm Garmin uses but a high(er) precision method is usually a few knots away from the G3X and G500 I'm familiar with. It's 'close enough' but I wouldn't bet TAS is always accurate to within a knot.

Nauga,
who misses tower flybys
I’ve flown airplanes where the pilot’s side TAS was several knots faster or slower than the copilot’s side With the same IAS. Fortunately, there was a wind shift somewhere between the two seats as well, so it didn’t rip the airplane apart. ;)
 
Just average the 3 GS to get TAS, then convert to IAS, no trig needed. Pick a day with light winds.
 
Secant cosine
Tangent sine
3.14159
 
funny to me, this thread is....when I was a kid I was discouraged when dreaming of becoming a pilot. "Pilots need to be good at math." my mom would tell me. Maybe she was trying to encourage my participation in math class.
but I think she & I both knew that math was not my thing..... as she was laboring through the multiplication table flash cards with me.

Now many years later I'm convinced that some brains are wired for math, and some are not. Mine is not.
Still, I went on to earn a BSME with a fairly decent GPA...AND I became an instrument rated pilot!
 
Just average the 3 GS to get TAS, then convert to IAS, no trig needed. Pick a day with light winds.
That only works on a day with no wind, since it doesn't take into account sideways drift.

The Gray method has the advantage that it doesn't require any particular heading on any of the 3 legs as long as they're reasonably different, all you need is a GPS and the ability to hold a straight heading and airspeed.
 
I would think flying four legs of equal distance would give the most accurate model. Fly each leg at 90 degrees to the other, you are flying a perfect square
Where as at the middle of each point and at level flight and same altitude record both Gps ground speed and indicated Airspeed.
Take all four numbers and divide it by four, for both GPS ground speed and Indicated Airspeed.
By having an extra data set, I would think that would increase the accuracy (and decrease a pilot induced math error) .
 
Winds change and vary making an in air test a little bit better than a guess. Bring it to a shop and get it calibrated. Use gps if you want to know your ground speed.
 
I would think flying four legs of equal distance would give the most accurate model. Fly each leg at 90 degrees to the other, you are flying a perfect square
Where as at the middle of each point and at level flight and same altitude record both Gps ground speed and indicated Airspeed.
Take all four numbers and divide it by four, for both GPS ground speed and Indicated Airspeed.
By having an extra data set, I would think that would increase the accuracy (and decrease a pilot induced math error) .

there are 3 unknowns: TAS, wind speed, and wind direction (or two components of the wind vector, as some of the papers prefer). That makes it a “3 equations, 3 unknowns” problem, and so 3 legs gives three data points which allows an ‘exact’ solution.

Given the lack of precision in the measurement (+/- .5 kts) (unless you can get your GPS to read out xxx.x or xxx.xx kts], lack of precision in reading the instrument the OP is trying to calibrate (ASI) (+/- .5 kts at best?), the methods referenced above yield results that may be accurate, but not of practical use in the cockpit.
 
Given the lack of precision in the measurement (+/- .5 kts) (unless you can get your GPS to read out xxx.x or xxx.xx kts], lack of precision in reading the instrument the OP is trying to calibrate (ASI) (+/- .5 kts at best?), the methods referenced above yield results that may be accurate, but not of practical use in the cockpit.
I disagree...I think being able to validate your flight planning is very practical.
 
there are 3 unknowns: TAS, wind speed, and wind direction (or two components of the wind vector, as some of the papers prefer). That makes it a “3 equations, 3 unknowns” problem, and so 3 legs gives three data points which allows an ‘exact’ solution.
It's only exact if there is no uncertainty and variability in the measurements. More samples gives a more accurate answer. You are correct that 3 is sufficient withing the precision/accuracy most of us need, but it is far from exact.

Nauga,
certainly uncertain
 
I disagree...I think being able to validate your flight planning is very practical.

Sorry, I did not mean to imply don’t validate. The way I should have said it is the extra effort to validate to a higher precision than I can measure is not practical for me, in my plane, for my mission. I grant that YMMV.
 
Sorry, I did not mean to imply don’t validate. The way I should have said it is the extra effort to validate to a higher precision than I can measure is not practical for me, in my plane, for my mission. I grant that YMMV.
What is an acceptable level of precision? +/-5 knots? 10?
 
It's only exact if there is no uncertainty and variability in the measurements. More samples gives a more accurate answer. You are correct that 3 is sufficient withing the precision/accuracy most of us need, but it is far from exact.

Nauga,
certainly uncertain

closed form or empirical solution is what I was trying to say.
 
funny to me, this thread is....when I was a kid I was discouraged when dreaming of becoming a pilot. "Pilots need to be good at math." my mom would tell me. Maybe she was trying to encourage my participation in math class.

Same for me. My third grade teacher told me I would never be a pilot with my math grades being what they are. I guess she was trying to encourage me to do better, but the message I got from her, and a few other teachers of that time period was....You will never be a pilot. So I gave up that idea for several years.
 
Remember everyone - it is only a pressure gauge/transducer that labels its dial as airspeed.
 
Slide rule! Slide rule!
Slip stick! Slip stick!

and I don’t remember the rest.

Just found it on the M.I.T. Women's Crew website (check out the beaver pictures) https://www.mitathletics.com/sports/w-crewlt/cheer

Beaver:
I'm a Beaver!

All:
You're a Beaver! We are Beavers all,
And when we get together, we do the beaver call!
e to the u, du dx, e to the x, dx;
cosine, secant, tangent, sine, 3.14159;
integral, radical, mu, dv;
slipstick, sliderule, MIT!

Go Tech!
 
closed form or empirical solution is what I was trying to say.
Agreed. Depending on the systems available I'll do it with anywhere from 3 to ~1000+ points. :cool:

With a logging EFIS and GPS (or even just a logging GPS) it's easy to get a huge amount of data from a single test sequence to smooth out the irregularities.

Nauga,
from the Hertz locker
 
You don't need a huge amount of data, just two points at the beginning and end of leg to calculate groundspeed from time and distance, no irregularities to smooth out if the points are far enough apart.
 
Back
Top