MVHR Commissioning Calculations - DIY Anemometer

[Barnacles]

Hi again folks,

I have had my system up and running for a few weeks now and I'm hugely impressed with the results. I initially configured the system by [semi-educated] guesswork and now that I have an anemometer to test the air flow rates it seems that I actually did far better than I could have hoped for with just a 6% difference between the total extract and total vent (in favour of the vents and therefor giving me the slight positive pressure I was looking for). ?

 

However, I now want to balance each room more accurately so I constructed an anemometer to help. I built my anemometer using a piece of 125mm pipe with a cap on one end. I cut a hole in the cap and used hot-melt glue to attach a Uni-T UT363-S to it. I added some gasket material to the other end to make a good seal as I press it against the ceiling valves. The whole thing cost around £28 and it appears to work well.

 

My question is this; when I do the calculation between air flow (in m/s) to volume should I use the 125mm diameter of the pipe or the 70mm of the anemometer's aperture? Common sense tells me that I should be using 70mm but a confirmation by someone more knowledgable would be great.

 

Happy New Year to all!

[dpmiller]

shouldn't be a cap with a hole as the 70mm bit will restrict and unbalance the flow. Yes, you can use the 70mm in your calcs but the numbers will be skewed by the restriction.

 

You be better cutting a hole in the side of a length of 125mm and putting the AFM in through the hole, central to the tube.

[Ed Davies]

1 hour ago, dpmiller said:

You be better cutting a hole in the side of a length of 125mm and putting the AFM in through the hole, central to the tube.

 

I'm sceptical. A quick scan of the specs for the UT363-S  didn't give the numbers but generally the problem with these cheap anemometers is that they don't have the resolution to make meaningful measurements at the low air flow rates typical of MVHR so they need the speed “amplified” by by concentrating it in a smaller area. As you say, that'll unbalance the flow a bit though how much is an open question. Also, if you're doing the same thing measuring all of the vents it cancels out, at least a bit.

 

1 hour ago, Barnacles said:

My question is this; when I do the calculation between air flow (in m/s) to volume should I use the 125mm diameter of the pipe or the 70mm of the anemometer's aperture? Common sense tells me that I should be using 70mm but a confirmation by someone more knowledgable would be great.

 

I agree with you; use the 70mm diameter. The diameter of the 125mm pipe is irrelevant, you'd hope to get the same reading if you used 200mm pipe, for example.

Edited January 6, 2020 by Ed Davies

[Jeremy Harris]

One reason I bought a Testo hot wire meter (the one that I donated to the tool share scheme here) is because it measures the true flow velocity past the hot wire elements, and can be fitted inside any duct to measure that flow velocity, with just a simple bit of arithmetic to convert flow velocity to volume, or mass, flow rate for the size of duct.

 

I first came across these super-sensitive hot wire air velocity measurement sensors years ago, back in the 1970's, when I first learned to fly.  Some of the gliders we had in the club used a bottle, connected to the static port via a pipe with a differential hot wire flow sensor, and this sensor was sensitive enough to work OK as a variometer (a sensitive form of vertical speed indicator, near essential when seeking out lift).  Somewhere I thing I may still have a glider variometer hot wire sensor, as I had a few spares left over from when I used to repair and calibrate instruments.  The tiny thermistors used as the "hot wires" were prone to damage from shock and vibration, and I seemed to often end up having to repair and recalibrate the things.

[Barnacles]

10 minutes ago, Ed Davies said:

 

I'm sceptical. A quick scan of the specs for the UT363-S  didn't give the numbers but generally the problem with these cheap anemometers is that they don't have the resolution to make meaningful measurements at the low air flow rates typical of MVHR so they need the speed “amplified” by by concentrating it in a smaller area. As you say, that'll unbalance the flow a bit though how much is an open question. Also, if you're doing the same thing measuring all of the vents it cancels out, at least a bit.

 

 

I agree with you; use the 70mm diameter. The diameter of the 125mm pipe is irrelevant, you'd hope to get the same reading if you used 200mm pipe, for example.

Thanks for that. Makes sense that the cheap AFMs aren't really ideal for accurate measurements of such systems. As my house is quite leaky at the moment I suppose I'm aiming more for proportionally distributing the air flow based upon the room size rather than trying to achieve a particular volume per hour. That said, if I could get somewhere close that would be great.

 

I've just looked at the spec sheet for the UT363S and it quotes a resolution of 0.01 in the range 0-99 (although it maxes out at 30m/s) with an accuracy of +-(5%+0.5m/s) which could put the readings out by as much as 30% so on a typical reading of 2m/s (which at 70mm aperture gives around 110m3 p/h) I could be getting anywhere between about 80 and 140m3 p/h. TBH that's probably close enough for my set up. As long as I'm somewhere around 100m3 p/h rather than 1000m3 p/h I'm a happy chap ?.

[Barnacles]

4 minutes ago, Jeremy Harris said:

One reason I bought a Testo hot wire meter (the one that I donated to the tool share scheme here) is because it measures the true flow velocity past the hot wire elements, and can be fitted inside any duct to measure that flow velocity, with just a simple bit of arithmetic to convert flow velocity to volume, or mass, flow rate for the size of duct.

 

I first came across these super-sensitive hot wire air velocity measurement sensors years ago, back in the 1970's, when I first learned to fly.  Some of the gliders we had in the club used a bottle, connected to the static port via a pipe with a differential hot wire flow sensor, and this sensor was sensitive enough to work OK as a variometer (a sensitive form of vertical speed indicator, near essential when seeking out lift).  Somewhere I thing I may still have a glider variometer hot wire sensor, as I had a few spares left over from when I used to repair and calibrate instruments.  The tiny thermistors used as the "hot wires" were prone to damage from shock and vibration, and I seemed to often end up having to repair and recalibrate the things.

Interesting, I haven't come across these before. I presume they use some sort of differentiator to compare two inputs?? A static and a dynamic??

[Jeremy Harris]

28 minutes ago, Barnacles said:

Interesting, I haven't come across these before. I presume they use some sort of differentiator to compare two inputs?? A static and a dynamic??

 

 

Pretty much, but they usually work by balancing the current needed to maintain two hot wires (which are usually very, very tiny bead thermistors now) at the same temperature.  One of these will be a held at close to ambient temperature in a sheltered location, the other will be at the top of the probe that's poked into the moving airstream.  The flow of air past the tiny exposed sensor cools it, and measuring how much current needs to pass through it to bring it back to the reference temperature is directly proportional to flow rate.  The handy thing about using very, very tiny bead thermistors is that you can measure their resistance to ensure that the temperatures remain equal.  They are also extremely sensitive, and able to measure very low air flow velocities.

 

In the flask variometers I used to repair, the tiny bead thermistors were placed on opposite sides of a tiny hole in a plate made from double sided PCB material (maybe a 1mm diameter hole, or thereabouts).  This plate had a glass pipe bonded either side of it and was connected by flexible pipes from the glass tubes to the static port(s) on the fuselage and a lightweight air flask.  With no air moving into, or out of the flask, both would remain in equilibrium.  If the glider started to gain altitude, then air would flow from the flask out of the static port, as air pressure dropped outside.  This would cause one thermistor bead to cool faster than the other and indicate that the glider was climbing.  The same would happen the other way around when descending.  They were tricky things to repair, even when I was a lot younger, with sharper eyesight, as the thermistor beads were maybe 0.1mm in diameter, with even finer wires coming out either side, and both beads had to be exactly centred on the small hole, with the wires soldered in place to the board, so that stronger wires could be fed off to the instrument electronics.

[Barnacles]

18 minutes ago, Jeremy Harris said:

 

 

Pretty much, but they usually work by balancing the current needed to maintain two hot wires (which are usually very, very tiny bead thermistors now) at the same temperature.  One of these will be a held at close to ambient temperature in a sheltered location, the other will be at the top of the probe that's poked into the moving airstream.  The flow of air past the tiny exposed sensor cools it, and measuring how much current needs to pass through it to bring it back to the reference temperature is directly proportional to flow rate.  The handy thing about using very, very tiny bead thermistors is that you can measure their resistance to ensure that the temperatures remain equal.  They are also extremely sensitive, and able to measure very low air flow velocities.

 

In the flask variometers I used to repair, the tiny bead thermistors were placed on opposite sides of a tiny hole in a plate made from double sided PCB material (maybe a 1mm diameter hole, or thereabouts).  This plate had a glass pipe bonded either side of it and was connected by flexible pipes from the glass tubes to the static port(s) on the fuselage and a lightweight air flask.  With no air moving into, or out of the flask, both would remain in equilibrium.  If the glider started to gain altitude, then air would flow from the flask out of the static port, as air pressure dropped outside.  This would cause one thermistor bead to cool faster than the other and indicate that the glider was climbing.  The same would happen the other way around when descending.  They were tricky things to repair, even when I was a lot younger, with sharper eyesight, as the thermistor beads were maybe 0.1mm in diameter, with even finer wires coming out either side, and both beads had to be exactly centred on the small hole, with the wires soldered in place to the board, so that stronger wires could be fed off to the instrument electronics.

Fantastic explanation, many thanks indeed Jeremy!

[Barnacles]

1 hour ago, Barnacles said:

Thanks for that. Makes sense that the cheap AFMs aren't really ideal for accurate measurements of such systems. As my house is quite leaky at the moment I suppose I'm aiming more for proportionally distributing the air flow based upon the room size rather than trying to achieve a particular volume per hour. That said, if I could get somewhere close that would be great.

 

I've just looked at the spec sheet for the UT363S and it quotes a resolution of 0.01 in the range 0-99 (although it maxes out at 30m/s) with an accuracy of +-(5%+0.5m/s) which could put the readings out by as much as 30% so on a typical reading of 2m/s (which at 70mm aperture gives around 110m3 p/h) I could be getting anywhere between about 80 and 140m3 p/h. TBH that's probably close enough for my set up. As long as I'm somewhere around 100m3 p/h rather than 1000m3 p/h I'm a happy chap ?.

? Just had a head-scratch moment. When I rechecked I couldn't make the numbers stack up. Realised I was using 70mm as the radius in my calculations rather than 35mm! ?

[Seeoda]

I did this recently and I want to show the method to help others. I used Volume flow rate (m^3/s) = Speed (m/s) * Cross-sectional area (m^2)

 

Very easy with ChatGPT