The tile is an orifice and the maths is old
Air moves through a hole because the pressure on one side is higher than on the other, and the velocity that pressure buys is the square root of twice the pressure over the density. In the units a floor manometer reads, an inch of water column is 5.20233 pounds per square foot, and with the 0.075 lb per cubic foot of standard air that works out at about 4,008 feet per minute for one full inch of water.
Nobody runs a plenum at an inch. At the 0.05 in wc in the default the velocity through the holes is 896 fpm, and a 24 by 24 tile that is 25 percent open presents one square foot of hole. Multiply by a discharge coefficient of 0.70 and the tile delivers 627 CFM. Hood a real 25 percent tile at that pressure and you will usually find something in that neighbourhood, which is the only reason to trust the arithmetic at all.
Doubling the pressure does not double the flow
Because of the square root, going from 0.05 to 0.10 in wc — which across a whole plenum is a large and expensive change — takes the tile from 627 to 887 CFM. That is 41 percent more air for twice the pressure, and the fan power to hold that pressure went up considerably more than 41 percent.
Swapping the tile is the cheaper lever in almost every case. Open area appears linearly rather than under a root, so a 56 percent grate in place of a 25 percent tile roughly doubles the flow at the pressure you already have. The catch is that it does it by taking pressure from everywhere else in the plenum, which is a shared resource nobody owns.
Matching the tile to the cabinet
A 6 kW cabinet at a 22 degree rise moves 862 CFM. One default tile carries 4.37 kW at that rise, so two tiles cover it with 393 CFM to spare. Push the same cabinet to 10 kW and the demand goes to 1,436 CFM: three tiles, or two grates, or a conversation about where that cabinet should be standing.
The rise is the input people get wrong. A rack of storage shelves with slow fans runs a wide rise and moves little air. A rack of dense compute with high static fans runs a narrow rise and moves a great deal. Using 20 degrees for everything can put the airflow demand out by half in either direction, and it is measurable in ten minutes with two probes.
Where this arithmetic stops being true
Close to a downflow cooling unit the air under the floor is moving fast horizontally, and fast-moving air has less static pressure available to push through a tile. Tiles within a few feet of a discharge can deliver a fraction of what this page predicts, and in bad cases they draw air downward into the plenum instead of up into the room. No open-area calculation catches that; only a flow hood does.
The plenum itself is the other unknown. Depth, blockage from abandoned cable, unsealed cutouts under cabinets and missing tiles in unused areas all bleed pressure. A room that has been running for fifteen years usually has more air escaping through holes nobody remembers than through the tiles that were designed to deliver it, and the first useful piece of work is a walk with a smoke pencil rather than a spreadsheet.
Questions people ask
How much air does a perforated floor tile deliver?
It depends on three things and only three: the open area, the pressure under the floor and how much of the theoretical flow the tile actually passes. At 0.05 in wc, a 24 by 24 tile that is 25 percent open with a 0.70 discharge coefficient gives 627 CFM. Raise the pressure to 0.1 and it becomes 887. Swap to a 56 percent grate at the original pressure and it becomes about 1,405. There is no single number, which is why the page asks for all three.
What plenum pressure should I run?
That is a design question for the room, not a number a page can hand you, and it is set by what the tiles have to deliver, what the cooling units can hold, and how much leakage the floor has. What this page does is invert the question: it tells you the pressure that would make the tiles you have match the airflow the cabinet actually pulls, so the conversation starts from a figure rather than from a habit.
Why is the tile in front of my hottest rack delivering nothing?
Most often because it is too close to a cooling unit discharge. The air under the floor there is moving sideways at speed, and static pressure is what pushes air up through a tile — where the velocity is high the static is low, and a tile can deliver very little or even pull air downward. Blocked plenum, a closed damper under the tile and a leaking floor elsewhere are the other three usual suspects. A flow hood settles it in a minute.
Is more airflow always better?
No. Air that goes past a cabinet rather than through it does no cooling, costs fan power on the cube of flow, and returns to the unit cold, which narrows the unit temperature difference and makes it do less work per unit of air. Matching the tile flow to what the equipment moves is the target; a large surplus over the top of a cabinet is a sign the containment is doing nothing rather than a sign of comfort.
Where does the discharge coefficient come from?
From the manufacturer flow curve if you have it, or from your own flow hood if you do not. Put a hood reading into the field on the form and the page works back the coefficient your tile and your plenum actually produce, which folds in the damper position, the hole geometry and the local conditions under the floor all at once. That number is worth far more than any datasheet figure, because it is measured in the room the question is about.