Posted by Morris Dovey on August 4, 2004, 5:44 am
It's fairly diffuse because there's a /lot/ of turbulance in the
Yes! I combined this info from all three responses and wrote a
simple little C program to model the panel's behavior.
It was an instructive exercise. In order for the panel to work
really well, the interval (t) would need to be reduced to about
five seconds. Given that the length of the flow path is nearly
seventeen feet, this indicates that air will need to flow through
the panel at a speed greater than three feet/second.
Since I don't think the current input baffle geometry will allow
that, a redesign is in order. While I doubt that I'll actually
achieve that speed, I'm sure I can substantially improve on the
current performance. I should at least gain enough information to
begin accounting for baffle dimensions to the model.
This is becoming interesting (-:
DeSoto, Iowa USA
Posted by Gary on August 4, 2004, 1:34 pm
Morris Dovey wrote:
Here is a data point for you.
My free convection air panel is 5.5 inches thick, and 46.5 inches wide
( 256 in^2 cross section). The panel is 95 inches tall. It has two
vents on top that are 4 inches by 18inches each (144 in^2 total, or
56% of the cross section area). It has two intake vents on the bottom
that are the same as the top vents. There are two layers of ordinary
window screen suspended in the middle of the panel for its full
length. There are no twists or turns, except that the air makes a 90
degree turn entering and a 90 degree turn exiting.
The top vent exit velocity under sunny conditions is 140 ft/min, and
the corresponding temperature is around 120F.
Nick's advice on entry and exit vent sizes was that you will continue
to gain on velocity and efficiency until the vent areas approach the
cross sectional area of the collector (which sounds right to me).
Posted by Gary on August 4, 2004, 2:16 pm
You might want to try making the next prototype with cardboard,
foamboard, OSB... and put it together with duct tape or foam glue --
i.e. cheap, quck, and easy construction. That way you can get some
results quickly, and if they are not what you want, you can change
things quickly and easily til you get what you want??
The Polycarbonate or figerglass glazing sheets that the lumber yards
sell for patio covers are good for glazing quick prototypes.
Posted by nicksanspam on August 4, 2004, 2:31 pm
Vertical's best for natural airflow. The more height the better.
You might improve the entry and exit with gradual transitions from
the smaller vent to the larger cross-sectional area, and add an
internal "turning vane," something like this:
| t t|
| t t |
Posted by Gary on August 4, 2004, 2:59 pm
I should have made it clear that the 140ft/min is measured in the exit
vent flow -- the average velocity inside the collector would be more
like (140fpm)(144 in^2 / 256 in^2) = 79 ft/min i.e. the air flow
through the vents is faster because the area is smaller.