Basejumper.com - archive

I've been working on a project to try to measure indicated airspeed in a wingsuit based on the pressure in the leg wing and I put together a demo video:

https://youtu.be/jQ7ZT_hvThs

The basic idea is that the pressure in the leg wing is going to be pretty close to the total dynamic pressure of the flow, so you can compare it to static pressure and get a pretty good estimate of the IAS, which differs from true airspeed depending on density. I think getting within about 5% of the "real" IAS is pretty realistic. The inaccuracies will come from the angle of attack and the porosity of the fabric. I think the porosity of the fabric is going to make a really small difference and the alignment with the flow will be an issue with any total pressure measurement system.

It makes sense based on conservation of mass and momentum. Apart from the small amount of air that leaks through the fabric, the mass coming into the inlet has to equal the amount going out. The freestream keeps wanting to ram air into the inlet and the only way the inlet has enforce conservation of mass is by pushing back with a pressure gradient equal to the dynamic pressure.

I'm imagining that rather than being a stand-alone system, this could be part of one big Kalman filter with a bunch of other sensors. The basic idea is that you make a prediction, make a measurement, meet somewhere in the middle, then repeat. Where the "middle" is depends on how sure you are about your predictions and your measurements.

On the sensor side, we could attach some uncertainty to the IAS measurement, then combine it with GPS, inertial data, magnetic data, etc (all with their own uncertainties). On the model side, we could start with the most general model possible: a rigid body that can rotate and translate in all three dimensions. Then we can refine it by putting restrictions on it that match the physical system: constant gravitational force, lift only generated perpendicular to the flow and normal to the body, lift and drag are proportional to airspeed, etc.

The CFD software is OpenFOAM, which is free, and the hardware I've been using is a few Adafuit Feathers with BMP280s. I'm about to start a new job where they basically own everything I come up with, so it might become hard to publish this kind of stuff, even for free, so if anyone is interested in playing with or adopting this project, let me know.

This is a pretty interesting idea.

Using two pressure sensors (one inside the suit, one outside) and looking at this difference is a clever approach.

One challenge I see is that the internal pressure would be significantly affected by body position and angle of attack. What I'm saying is there might be flight configurations with higher airspeed, but worse pressurization because the inlets are not directly in the airflow (or vice-versa).

Some questions:

Have you tried comparing to GPS data on the most windless day you can find?

Does this need to be calibrated per-person-per-suit, since there will be difference pressurization to speed relationships?

You remind me of Geo.

My data is pretty limited so far and I haven't jumped this set-up on a windless enough day yet. I'm mostly trying to get the idea out before I start my new job to avoid intellectual property problems. If I publish it now while I'm still an unemployed bum, then other people can start experimenting with it too.

I think the suit and person will make a difference, but the pressure inside the wing is always going to be pretty close to the total dynamic pressure. Even if we aren't truly measuring the total pressure, I think we're measuring something just as important. Think about what happens when a canopy stalls. You lose pressurization and pressure inside = pressure outside. What happens when you stall a wingsuit? Probably the same thing.

Angle of attack will make a difference but something to think about is that the same is true of airplanes. Even in an airplane, the pitot tube isn't always perfectly aligned with the flow. Think about a C172 flying in level flight at 110 kts. If you want to decrease your airspeed to 90 kts but maintain level flight, you need to decrease the power and increase the pitch. Since the pitot tube is rigidly attached to the plane, its pitch changes too, but you're in level flight in both cases so the wind is coming from the same direction. It can't be perfectly aligned both times.

Wow - interesting stuff.
I just need to read more so I can understand it all.

Very cool project. Have you thought about a more traditional pitot-static tube too? The leg wing internal pressure sounds like a pretty ingeniously good solution, though it would depend on some other assumptions like the leg wing being properly inflated which probably has some noteworthy limitations. But for most cases it sounds like a great source.

Regarding AOA correction you are basically just talking about the difference between KIAS and KCAS but you probably already know that. It's not a big issue in 99% of aircraft out there. I would leave GPS and INS data out of any kalman filter, because IAS is supposed to reflect what the wing is actually experiencing pressure-wise, and anything tied to actual displacement is only going to skew the output.

This is the hardware I used for anyone who's curious. It's an Adafruit Feather with a BMP280 and a small battery. The battery is covered up by the tape. It's pretty cheap, probably less than $50 for all these parts, but you also need a soldering iron if you don't already have one. I just used a small zip tie to tie it to the mesh in my leg wing. The one measuring static pressure can just go in your pocket. I used a third one in one of the arm wings for the jump in the video.

Edit:
I have also thought about a more standard pitot-static system and I think it would be a useful tool, but I expect that you'll get pretty similar results and I think most people would like the simplicity of taking the measurements in the leg wing vs having something sticking out. I also don't have any good ideas on how to mount a standard pitot tube. It seems like somewhere near your hand would be the logical place, but then there will be a need to manually point it in the right direction somehow.