The difference is massive and explains some of the behavior that I have seen.
It looks like I have to change the AIRSPEED_RATIO to 0.7-0.8 to have closer speeds, which does not make send (should be 2/volumetric density of air (1.2/1.3) = 1.5/1.6.
My pitot is parrellel to the ground but the plane (skywalker) may have a bit of tail down attitude. Would it be due to this and I need to change the position of my pitot tubes ?
Replies
The range on the two sensors is different, but the sensitivity is the same, and is what matters. We use the same constant for both sensors.
Past changes probably came from changes to the ADC. We are using 2 more bits here than we did on the previous hardware.
That is the reason that pilots fly by calibrated versus true airspeed. When you fly at higher altitudes where the air is less dense and you maintain the same calibrated airspeed you had at lower altitude the wing is flying at the same angle of attack.... Stall speed remains at the same calibrated airspeed.....
For ArduPilot, if we were to flying significantly higher than 400' then the place that changes in density would be most interesting is in our barometric altimeter calculation. We assume a certain pressure profile with altitude based on standard atmospheric temperature lapse rates. However it is rare that the temperature profile matches this ideal.
for this reason I use a curve in my pitot reading that takes in account all the error of ports postions.
From my program:
Force of air resistance
F_a = A2 * (V + H)^2 + A1 * (V + H)
From W&W p85ff, air resistance coefficient C_a
C_a = F_a / (dynamic_pressure_of_air * frontal_area)
Where F_a is force of air resistance. At bicycling speeds:
dynamic_pressure_of_air = air_density * (V + H)^2 / 2
where V + H is the relative air velocity.
So...
F_a = C_a * air_density * frontal_area * (V + H)^2 / 2
p90 suggests that the Reynolds number is in the ballpark of
200000, so there may be some advantage to wearing rough,
rather than smooth clothing.
air_density is 1.2047 kg/m^3 at 20^C = 68^F.
p91 suggests that roughly, C_a = 0.9 for a bicycle and rider.
F_a = (0.9 * 1.2047 / 2) * frontal_area * (V + H)^2
= 0.54 * frontal_area * (V + H)^2
This doesn't directly apply. But you can see how much temp changes things here. http://bikecalculator.com/veloUS.html
I have tested with the value in the code and it worked VERY well for me in the range from 10 to 20 m/s. Much closer than your results with the modified value.
What sensor are you using?
Also know that we are using a simplified equation, which will only provide reasonable results over a relatively small range of airspeeds. I get good results between about 8 and 18 m/s. At lower speeds the equation will under-report the value and at higher speeds it will over-report the value. If your nominal speed is in a higher range then adjusting the value is a good way to compensate.
We use the simplified equation as the more correct equation requires significantly more computation time. You can find it and experiment with it if you like. Look for a term raised to the 2/7 (or maybe it was 7/2 power) inside the square root. That is the one.
Other thing is that you should check the plane balance since this could be another reason of your tail down attitude...it could be tail heavy... Your model instructions should tell where to locate de CG on the wing!
Good Luck
Just a thought, I'm pretty new to this so I can't help you on the programming aspect.