# Final Flow Rate Calibration

 Its been a bit of a trial getting to a good method to measure collector air flow rates.   This page shows the method I finally settled on, and am using for the solar air heating collector design tests.

## Flow Rate Measurement

For most of the testing, I use a Kestrel wind meter (see below) to set up the flow rates.  Both the reference collector and the collector being tested have an opening near the end of the outlet duct cut to fit the kestrel meter -- I just insert the Kestrel  into the opening until the turbine is in the center of the duct, then, using the hold function to retain the reading, pull the Kestrel out and read it.  I normally use an average of three readings, but the readings are pretty consistent.

Since the the reference collector and the collector being tested have identical ducting and fan arrangements, when the Kestrel reads the same for both, the flow rates should be very close.

Flow rate = (Duct Area)*(Kestrel Velocity) * (Correction factor)

Where Correction Factor corrects for velocity profiles etc (see below)

For example, if a flow velocity of 690 fpm is read, and the 5 inch duct has an area of 0.136 ft^2, then the flow is:

Flow rate = (0.136 ft^2)*(690 ft/min) *(0.82) = 76.9 cfm

 Kestrel wind meter Kestrel being used to read the 5 inch exit duct flow velocity.

## Correction Factor

The Kestrel wind meter is subject to some errors having to do with the velocity profile across the width of the duct, and with the Kestrel blocking a bit of the flow path just by being there.  I also think that the lower density of the air both to our high altitude (5000 ft) and the higher temperature of the air may effect the Kestrel readings in that its readings are based on how fast the little turbine rotates, which in turn depends on the air forces on the turbine blades which depend (at least to some degree) on the air density.

So, to calibrate the Kestrel velocity readings, I use the fill time time for a larger (54 cf) poly bag.   I believe that the fill time for the bag gives accurate flow rates, and I use these flow rates calculated from bag fill times to derive a correction factor for the Kestrel readings.

I've worked on the bag system to try to make it as precise as possible:

- A flat plate on the end of the exit duct allows the bag to be quickly and precisely coupled to the fitting on the bag.  This allows for good timing of the start of the fill process, and for low flow resistance as the bag fills.

 This fixture allows a quick and accurate coupling of the end of the outlet duct to the bag, and a good seal of the bag to the duct end. Note the weather stripping and the clamp ridge at the bottom of the duct end plate. Just holding the bag fitting to the duct end plate allows a good, leak free and low resistance coupling of the bag to the duct end.

- The bag is carefully rolled flat before each inflation to get all the air out of it.

 Rolling the bag flat before inflation. The bag is rolled over the cardboard tube to empty all of the air out of it.

- The end point for the bag fill is determined by hooking up the manometer to read the static pressure near the end of the exit duct.  As the bag approaches the very end of its fill, the manometer pressure rises rapidly.  I use a rise of 0.1 inches of water as the point to click the stop watch at the end of the fill.

If you don't have a manometer, the end of the fill is pretty easy to time without it.  The bag sort of snaps into its final form, and makes the timing of the end of fill pretty precise without the manometer.

The bag must cause some back pressure when its being filled, and this must in turn reduce the flow rate slightly.  But, the back pressure is very low -- less than 0.01 inches of water, so I think the effect is small.

The flow rate is determined from the bag fill time as:

Flow Rate = (Bag Volume) / (Fill Time)

So, with a bag volume of 53.8 cf, and a fill time of (say) 35 seconds or 0.58 minutes

Flow Rate = (53.8 ft^3) / (0.58 minute) = 92.8 cfm

The chart just below shows the results of timing bag fills and equivalent Kestrel velocities for several different flow rates.  The Kestrel Correction Factor is determined from these fills at different flow rates.

<chart>

Gary January 9, 2011