Wind gust detector using MAP sensor.

[Circlotron]

Hi Guys
I'm working on an application for a wind gust detector for an automatic sliding door. The idea is, the door is normally open and facing the outside world, but if the wind starts blowing the door needs to automatically close until the wind calms down. Don't want to use an anemometer with rotating cups because this is too open to vandalism. Was thinking of using an automotive MAP sensor and monitoring the moment to moment changes in barometric pressure to see if this could detect wind gusts. Don't want to reinvent the wheel. Anyone done any work in this area before?

[Ian.M]

Hmm.  Two durable weather resistant horizontally mounted disks, the bottom a bit smaller with a lip on the top one to shed rain,  a coarse metal mesh to prevent tampering round the perimeter and in the centre, on a thin stalk to the midpoint, a glass bead thermistor as the self heating sensor.  The heat loss will be proportional to the square of the wind speed.  You need a reasonably well matched reference sensor at ambient temperature - probably best to mount it with thermal grease in a hole drilled in a block of metal attached to the bottom disk as a heatsink.  The top disk should be bright white or highly polished to reduce solar heating and ideally should be shaded from high angle direct sunlight.  The big advantage of the thermistor between two disks approach is it's robust and doesn't have any thin tubes that may attract mud dauber wasps, mason bees or other insects that favour them for nest building.

https://www.apsnet.org/publications/PlantDisease/BackIssues/Documents/1984Articles/PlantDisease68n05_413.pdf should give you some ideas. 
The easiest way of getting an adjustable clean airflow cheaply for calibration is to mount the sensor to a short mast on a board on a car roofrack then drive up and down a quiet road on a calm day.

[jt]

I like Ian.M's idea.  As a variation to consider, maybe you could go without the reference sensor and just trigger off of rate of change of the voltage across the self-heated thermistor (could be an analog differentiator circuit or a MCU based circuit) . 

Not sure if you would get false triggers from transient sunlight (ie. clouds moving overhead), but it seems like you could calibrate the rate threshold and go to some effort to configure the sensor housing to minimize these affects - maybe recess and suspend the thermistor in a thermally stable tube that lets air flow through.  By thermally stable I mean of an insulating material (plastic) or something of significant thermal mass.   

[DTJ]

What about using a cheap digital barometer IC? It may be easier than stuffing around with custom hardware.

A BMP180 or similar perhaps. If you wanted to go over the top put one on either side of the door and calculate the difference which would be proportional to the force on the door.

[calexanian]

A funnel feeding a sensitive mems sensor would work as well. Also there are plenty of cheap anemometers out there ready for use, or you can make your own with sole plastic bits, a photo eye, and some glue to put it all together.