Sometimes, it’s nice to know how windy it is outside. Knowing the direction of the wind can be a plus, too. To that end, [Sebastian Sokołowski] built himself an unusual anemometer—a wind gauge—to feed into his smart home system.
[Sebastian’s] build is able to tell both wind speed and direction—and with no moving parts! Sort of, anyway. That makes the design altogether different from the usual cup type anemometers with wind vanes that you might be used to seeing on home weather stations. [Sebastian] wanted to go a different route—he wanted a sensor that wouldn’t be so subject to physical wear over time.
The build relies on strain gauges. Basically, [Sebastian] 3D printed a sail-like structure that will flex under the influence of the wind. With multiple strain gauges mounted on the structure, it’s possible to determine the strength of the wind making it flex and in what direction. [Sebastian] explains how this is achieved, particularly involving the way the device compensates for typical expansion and contraction due to temperature changes.
It’s a really unique way to measure wind speed and direction; we’d love to learn more about how it performs in terms of precision, accuracy, and longevity—particularly with regards to regular mechanical and ultrasonic designs. We’ll be keeping a close eye on [Sebastian’s] work going forward. Video after the break.
A pal of mine at Cornell pondered doing that exact same thing years ago. The issues with that are the strain gauges and the elastic material they need to be mounted on to measure very small amounts of strain. You may think of something like a spaceball 3D mouse, and yes, that works under the same principal, but the thing with the space mouse is that all of it’s measurements can be relative and not absolute. If the strain gauge output drifts a bit with temperature or the elastic degrades a little bit, it does not matter. It can just take the readings it is getting when it turns on and consider that zero.
Strain gauges are deceptively hard to work with beasts. It is usually just a bridge type circuit but the connection to the gauge can take up to 5 wires to eliminate most connection errors, that gets to be more important when the gauge is remote, but it is worth mentioning. They also respond to temperature effects so you often times have to use another one to compensate for that. If you take a modern load cell apart you will see that it has a lot more than one simple strain gauge.
We pondered the whole idea and came away thinking that between the material properties changing with temperature and degrading over time, and the very low signals and other assorted issues with strain gauges, that it probably would not remain accurate for long out in the real world, and a traditional anemometer with one moving part and a magnet and hall effect switch would hold up better. Not seem as high tech but hold up better. We had a weather station from the 70’s that was still going 24/7 in the 2000’s when I left. I think 30 years is an acceptable service life, and it was not dead when I left.
Without any moving parts?
How the strain gauges measure if it isn’t moving?
Did you read the part in the first line after the lede? “no moving parts! Sort of, anyway.“
I’m a bit worried how long such a 3D printed part will last outside in full UV and weather exposure. For example on my balcony, which is quite exposed to the sun, even plastics designed for outdoor usage, like watering cans or plastic plant pots, crack after like 2 to 5 years of usage. Regular 3D printer filaments often don’t have very good UV protection.
I would have coated the whole thing in a black heavy duty outdoor paint to give it additional protection.
I’ve had good luck with both ASA and HIPS, but not very many parts and not very many years yet. Even a garden hose wye connection in ABS is still holding out (and holding pressure) after 5 years, though looking a bit aged.
I seem to remember a wind gauge project on hackaday that used 2 pairs of ultrasonic transducers mounted at right angles to each other – you alternated sending pulses back and forth between pairs of transducers to determine the wind speed via Doppler shift and a bit of trig would help you compute the direction form the relative Doppler shift between the pairs of transducers.
This is the one, very interesting project indeed
https://hackaday.com/2021/07/06/open-source-ultrasonic-anemometer/
I had one of those remote stations fall apart due to UV light.
Thanks for showcasing my project here!
Pro-tip: Wii Fit balance boards have four strain gauges and a well-documented Bluetooth interface. They’re also cheap as (Chinese, knock-off) chips at your local thrift store/charity shop.
Another useful tech here is thermal anemometer sensors. A wire or thin film resistor is heated to a temp higher than the air around it by passing a current through it. The amount of current required to maintain the wire temp is proportional to air velocity.
I used a spring in the middle and capasitive measurement at 1978. I used VIC20 as a computer.
I did not continue, because there was certain problems using the system in a sailing boat.
My anemometers use similar principle, but use full bridge strain gage configuration to avoid thermal errors and precision excitation requirements (differential signal independent of bridge excitation).
I am surprised that this gives reasonable numbers at all over (for my location) a 45C range.