I took some comparative data of some of the exhaust valves used in double hose regulators.
To take the data I used a Magnehelic and a Dwyer flow meter.
To supply the air flow I used a blower (from a new shop vac that I bought specifically for this type of test) to generate relatively high flow rates.
Notice that I am using air flow at 1 atmosphere, but I am increasing the air flow to higher that realistic values in order to replicate the Reynolds numbers expected at normal diving depths. I don’t have the facilities to test at hyperbaric pressures to increase the air density, but I can increase the air flow and obtain the same Reynolds numbers.
Everyone can understand comparative testing even if you are not following the significance of non-dimensional Reynolds number scaling.
Let me just say that the test conditions are controlled to create a realistic and comparative conditions.
I adjusted the air flow to accurately reproduce the same conditions during all data collection. Only one variable was changed at a time following well established scientific methods for data collection.
Notice that this data is only for steady-state flow. The breathing cycle is closer to a sine wave / transient type of flow.
I tried to also take initial cracking effort (initial valve opening) data, but the numbers tends to be too low to accurately measure. The only observation I can add is that duckbills do tend to stick at times, but I think we have all observed that.
The Magnehelic that I was using only reads up to 5 inWC, therefore I only took data with a pressure delta below or equal to that level. You will see a couple of data points were I could not reach the 10 cfm level without exceeding the 5 inWC limit.
The pressure differential data is just the raw data as I read it. Some of it was expected, but some was a bit of a surprise.
The data was very easy to take with one exception, the silicone duckbill fluttered a lot with 10 cfm of flow. This caused the needle to move wildly. The reading shown was my best average of the fluctuating gauge.
![Image](http://i17.photobucket.com/albums/b60/wavejockey/DHexhaustflowresistance7-19-2013_zps3f10f572.jpg)
Here are the test samples.
The Argonaut can is a 3D printed prototype. It is intentionally blurred in the pictures for a couple of reasons… There will be more pictures of the new cans in the near future.
![Image](http://i17.photobucket.com/albums/b60/wavejockey/Picture001modified_zpsa70d48e4.jpg)
Here is part of the test set –up.
Attached in this picture is the can with no valve used as a baseline for comparison testing.
![Image](http://i17.photobucket.com/albums/b60/wavejockey/Picture002modified_zps53beb8b3.jpg)
![Image](http://i17.photobucket.com/albums/b60/wavejockey/Picture009_zpsb6107939.jpg)
![Image](http://i17.photobucket.com/albums/b60/wavejockey/Picture003_zpsa49ea1e8.jpg)
Here is the Scubapro 109 connected for testing.
![Image](http://i17.photobucket.com/albums/b60/wavejockey/Picture008_zpse6530e94.jpg)
I bought this little vacuum cleaner specifically to use id as a blower for this type of testing (either for blowing or to generate suction/ vacuum).
![Image](http://i17.photobucket.com/albums/b60/wavejockey/Picture013_zps7e5c4c09.jpg)