Friday, August 21, 2009

Mucking Around with Fans (Venturi Effect)

I watched a TV program the other day that talked about the venturi effect, and how it can be used to push a greater volume of air... applications include blowing up aircraft evacuation slides quickly, etc...
I wondered if this could be applied to PC cooling fans - to get more airflow out of a simple fan ??

The idea is something like the following (they seem to call them Eductors)

(image from this site)

First step was to work out a way to measure the effectiveness of a fan.
I opted for using a large plastic garbage bag, and seeing how long it would take to blow it up. Here is the bag blown up...

Typically it would take around 2 minutes to blow the bag up quite firmly. One of the problems with using this method to measure, was that it was difficult to determine when the bag was inflated. Difficult to measure.
Also, as the bag becomes more inflated, then internal pressure increases, which would impact the performance of the fan.
After mucking around, I came up with taking photos every ten seconds to measure the progress of inflation... seemed to provide a reasonable view of how the bag was inflating.

Anyway - the two options tested were....

Straight Through Connection

Basically had the fan directly connected to the bag.

Venturi (Eductor) Connection

Consists of a square section which has a small constriction (nozzle) for speeding up the fan air flow. This fits into the cardboard diffuser which connects to the plastic bag. Note that the nozzle and diffuser are only implemented along one axis.

To test the difference, I did two runs each, photos taken every 10 seconds. The results are shown in the following image, one row for each run.

(You will need to click the image to see clearly...)
It seems that the second and fourth runs are slower at blowing up the bag, and that the end point of the inflation is flatter than the other two runs.
For the first and third runs, the inflation seems quicker, and the end point is also fuller (i.e. greater pressure).

The first and third runs are with the straight through fan... :-(
Which shows that with this Eductor design (if you can call it a design :-) ), it gives a slower inflation, and can handle a lower pressure.

So - dont muck around with the fans - just use them as they are meant to be used...
Or try and test a proper Eductor design - if you know how to do it.

No comments:

Post a Comment