Wireless Pressure Monitor on Rotating Impeller

Hi guys & girls,

I have a uni project that requires me to install pressure sensors on the impeller blades of a rotating centrifugal fan. Pressure readings need to be taken at different locations of the impeller. Given that this is a rotating part. The only method i can think of is mounting the arduino in the centre powered by battery and have it wirelessly transmit to a LCD display so that i can record the readings while the fan is operating. Is this a feasible idea? I'm new to arduino and wonder how hard it is to set up the programming. I've had a search around for some pressure sensor modules, and the ARDUINO PRESSURE SENSOR MODULE BMP280 seems reasonably small enough to fit onto the impeller. What other components will i should i gather to complete this?

Thanks everyone!!

Can't you just connect the sensors to a slip-ring on the fan shaft?

How fast does the fan rotate?

Are you trying to measure the air pressure on the fan blades?

...R

What you're asking for is a tall order. Not sure if impossible, but not going to be easy.

Impellers tend to rotate fast, so even a small sensor will cause imbalance which may cause the part to fail. Especially at the tip of the blades you may experience serious centrifugal forces - make sure that this in itself doesn't affect your sensor.

Also you need quite a large space somewhere on/in/at the blades for the rest of the components. So size of the blades is key to feasibility. Pressure sensors themselves can get really small, such as the MS5837, just a few mm in all dimensions and a few grams at most in weight. The wireless module OTOH is much larger, due to the required antenna which has a physics-dictated minimum size.

Then mostly out of curiousity: how are you planning to mount the sensor so it can measure pressure without disturbing the air flow?

are you trying to use Bernoulli's principal to determine blade design ? that would mean a lot of sensors.

and you say centrifigual fan, is that forward or backward inclined ?

if the squirrel cage is open, you can mount a plate on the end or a spider and put your electronics inside.

are you working on an HVAC fan (less than 400 RPM?) or automotive blower ?

you should be able to run tubing from your blade to the axis with your electronics in a nachelle and then output your data.

it was brought up that balance is important. if you want to sense one blade at the leading edge, a second with maximum lift and a third with some point after, your trio forming a triangle would put tubing out to create a close balance so any out of balance would be mitigated by the other two.

also, look at both industrial op-amps and a separate ADC, the on-board 10 bit is not going to work with a raw D.P. sensor.

since you piqued my interest, in a fan, the air impact and the flow over the blade will have very similar properties as an air foil in a wind tunnel. you can bracket any aspect in a wind tunnel, then verify application on a blower wheel.

is this for a school project ?

I don't think sensors like the BMP280 are designed to handle the differential pressure and turbulence that you would get in a fast airflow. Also I think the bulk of it will influence the airflow and change it from the value that would be there if the sensor was not present, if the fan is not very large.

It would behoove you to run tubes to ports on the blade to help with that. In fact, if you had a rotatable coupling, you could move the BMP280 off the fan this way. Although, the tubes could pose a similar problem of interference.

as a note, if you are using a dp transmitter, is your reference point at the blade ? or another point in the air stream. since the dp readings will be so small, the closer to the sensor you have the reference, the greater the changes will be of full span of your sensor.

a few decades ago, I was in the air flow monitoring industry, back in the day when pneumatics were king and electronics was not even close to being able to accurate enough.

aarg: I don't think sensors like the BMP280 are designed to handle the differential pressure and turbulence that you would get in a fast airflow. Also I think the bulk of it will influence the airflow and change it from the value that would be there if the sensor was not present, if the fan is not very large.

It would behoove you to run tubes to ports on the blade to help with that. In fact, if you had a rotatable coupling, you could move the BMP280 off the fan this way.

one of the goals of any measuirng device is to have no effect on the thing being measured. an electronic sensor would be best if it were completely flat on the surface. a slight disruption in the boundary layer would effect changes in pressures. ideally a laminar flow would be the result of any addition. with tubing, you can drill a very tiny hole in the surface and with that, detect pressure changes. the BME280 is an absolute pressure sensor. think of measuring the thickness of a piece of paper by using a yard stick on the floor and the paper on your desk. the total range is never used, only a very small part of the span of the device. most of the accuracy is lost as the device error is a percent of span, and you are not using 99% of the span. a dp transmitter would measure your reference pressure (static pressure) and then your desired pressure (total pressure) to get differential pressure. you can get a transmitter in a range of 0.1 inch of water (very high sensitivity) whereas the BME280 sensitivity is many times larger (less sensitive) than the full scale of the dp transmitter. the data sheet says 1 meter, but is very lax on details.

Robin2: Can't you just connect the sensors to a slip-ring on the fan shaft?

How fast does the fan rotate?

Are you trying to measure the air pressure on the fan blades?

...R

Thanks for the slip ring idea, that should help scale down the arduino components required.

Fan rotates faster than 1400Rpm

and yes trying to measure the air pressure on the fan blades

dave-in-nj: are you trying to use Bernoulli's principal to determine blade design ? that would mean a lot of sensors.

Yeah, that's the principle involved and developing a better understanding of pressure distribution across an impeller blade. At the moment only simulations can be run with CFD software, but i'd like to run an experiment across it.

and you say centrifigual fan, is that forward or backward inclined?

Backward inclined

if the squirrel cage is open, you can mount a plate on the end or a spider and put your electronics inside.

are you working on an HVAC fan (less than 400 RPM?) or automotive blower ?

Automotive blower

you should be able to run tubing from your blade to the axis with your electronics in a nachelle and then output your data.

it was brought up that balance is important. if you want to sense one blade at the leading edge, a second with maximum lift and a third with some point after, your trio forming a triangle would put tubing out to create a close balance so any out of balance would be mitigated by the other two.

What kind of tubing are you talking about? Like a pitot tube?

also, look at both industrial op-amps and a separate ADC, the on-board 10 bit is not going to work with a raw D.P. sensor.

Thanks for the tips! .Electronics is not my strong point but a couple tutorials should be able to build the foundation

newbster: Thanks for the slip ring idea, that should help scale down the arduino components required.

Is it just my age that causes me to be astonished that someone studying engineering would not be aware of slip rings?

Don't they get students to look inside machines any more?

...R

dave-in-nj: since you piqued my interest, in a fan, the air impact and the flow over the blade will have very similar properties as an air foil in a wind tunnel. you can bracket any aspect in a wind tunnel, then verify application on a blower wheel.

is this for a school project ?

Yes, this is for a university project. That sounds like a much more plausible idea without having the trouble of rotating parts. Hmm, i guess the impeller can act like a stator vane, with the wind blowing into the center, the hub cone will disperse the airflow outwards along the surface of the curved blades. Awesome! i'll get some research into this method should it prove easier to measure. Will still need to set up some pressure sensors around though. Thanks for the insight!

Robin2: Is it just my age that causes me to be astonished that someone studying engineering would not be aware of slip rings?

Don't they get students to look inside machines any more?

...R

I wish i did, but its more books and calculations and less practice at mine.