Measuring Air Flow with a Differential Pressure Method

What is the application?

I've worked with human respiratory measurement...fixed orifice is very non linear and the results pretty noisey. Are you going to get a large temperature variation? One advantage of respiratory work is it was relatively consistent temperature/humidity etc.

I tended to work empirically to gather data and then fit curves/equations to the data.

As Del said, the air flow through an orifice is dependent on so many variables.

shape of the orifice, particualrly leading edge of orifice, length of orifice etc...

Gas composition temperature and ambient pressure plus humidtiy and such like...

Generally its a square law characteristic, but the only way to calculate the differential pressure against flow rate is some extremely complex calculations which at best will only give you a 'ball park' figure +/- 20% say...

Even the designers of the orifice plates will only give an estimation of flow rate measured... depends on pipework and so many other things....

That's why for reasonably accurate work you will need to calibrate them in-situ...

John.

http://en.wikipedia.org/wiki/Orifice_plate indicates a relationship between the differential pressure and the velocity, which infers volumetric flowrate. Plenty of equations to play with there. Temperature is a significant feature, though not a linear one.

Everything on orifice plate measurement depends on the density. As others are posting, this is related to the pressure, temperature, and compensation of the fluid being measured. As long as you know these variables fairly well or they are fairly constant, the flow is related to the SQRT(Diff. Press * Density), times some other fairly constant values.

Differential Pressure sensing is related to volumetric airflow, which means it is also related to mass airflow because it is volumetric. Mass airflow is used when other materials are in the airflow such as fuel, natural gas, etc.

Unfortunately because your measurement is indirect and as others have pointed out the inconsistancies of the orifice plate, these errors would be magnified due to the indirect measuring. It doesn't mean it can't be accomplished though. Through digital calibration, these variances can be "zeroed out", but that isn't just a simple task.

As far as temp variation, that shouldn't be an issue considering you said that you measure at room temp and a few degrees variation should not make a noticable difference.

Airflow sensors I have worked on were mass type and used a titanium wire that changed resistance due to the flow and the materials.

Hi - we use an orifice meter on one of our test benches, the others are primarily nozzle metered. Check out the ASHRAE document dealing with orifice airflow - ASHRAE Standard 41.7-1984 (RA 2000). The equation M (lb/hr) = 359 * K * Do^2 * Y * Sqrt(hw/V) gives airflow in mass flow units. Of course you are factoring in air density as the others have noted. To get the proper air density, you need air temperature as well as barometric pressure, orifice inlet pressure, and humidity. So, the temperature is already a contributor to the flow function. This is all great so long as you are in proper flow ranges, density ranges, dry liquid (<2% liquid content), etc for the orifice under consideration. I can give more info if required.

What is your application? What temperatures are you operating at, orifice/line sizes, etc.?

AndyC.

Hi Andy,

Thank you for your explanation. my application is for the human respiration. If your can provide more info, it would be great.

I use the mass flow meter and a precise pressure transducer(PPT) to measure the orifice pressure-flow characteristics at different ambient temperature. The P-F curves are diferent. I'm not sure whether it is affected by air temperure through the orifice or by the error of the flow meter.

Thank you very much.

Are you measuring in one direction (Inhale only) or bi-directional (inhale/exhale)?

What diameter tube and orifice are you using?

Is the tube straight or are there curves and bends? Can you get 20 diameters straight upstream and 10 diameters straight downstream?

Any checkvalves in the flow stream could disrupt your measurements.

As Del mentioned earlier, for equipment this small the relation between DP and volumetric or mass flow will probably be empirical and not lineal.

If you are doing serious work on respiratory you will ideally want access to what is best best described as a breath simulator. The ATS (American Thoracic Society) has standardised waveforms for testing, these are in digital format and are used to drive a computer controlled piston giving reproducible simulated loops (breaths in and out) with agreed values of PRF FEV1 FVC etc.

The problem with respiratory work is that some measurements e.g.PEF (peak expiratory flow) were originally derived on arbitrarily defined equipment, (e.g pen recorders with 1uF capacitor added for smoothing) these measurements the became the 'gold standard' due to the work of Dr Martin Wright. Thus one need to take care and not simple measure a peak value and assume that is right. (e.g. You will need to do some averaging of digital values over say 10ms or some analogue smoothing to get the correct reproducible values)

If you doubt my expertise check out Martin Wright, Clement Clarke, Mini-Wright, VM1, ATS etc.

I was Chief Designer at Clement Clarke for about 12 years and although I have forgotten a good deal I can doubtless point you in some useful directions.

(I find it odd that you chose not to respond to my initial post as I have expertise in this field)

Note:-

There are two competing scales for PEF measurement .

The 'Wright' (historical/empirical scale) prevelant in Europe.

The ATS scale prevalent in the USA and newer markets.

Most of the research, early work and non-US papers will refer to the 'Wright' scale.

There is still controversy about this.