Duct Size Calculator — Online Ductulator

Velocity or friction rate? Manual D uses both, in a specific order

People talk about these as two competing methods. They are not. ACCA Manual D — the ANSI standard for residential duct design, and the method the IRC points to in section M1601.1 — uses both, with a clear division of labor:

1. The friction rate sizes the duct. One rate, derived from your specific system, applied to every duct in it.
2. Velocity caps the result. Manual D Table 3-1 sets maximum air speeds by duct role. If the friction-sized duct exceeds its limit, you go up a size regardless of what the friction math allowed.

The flip point — why Table 3-1 has to exist

Look at the flex duct chart above and compare the 600 FPM column against the 0.08 in/100ft column:

Around 12 inches they cross, and that crossover is the whole reason the velocity table exists. On small branches friction binds first, so the velocity cap never comes into play and Table 3-1 is irrelevant. On large trunks friction would happily let you keep loading the duct, and the velocity cap is the only thing standing between you and a system you can hear in every room.

This is also why small branch runs are where residential systems fail. A 6 inch flex run sized on velocity alone looks like it will carry 118 CFM. Sized on friction at 0.08 it carries 70. If you designed for 118 and built it out of 40 feet of flex with a couple of bends, the air was never going to show up. Take the lower number.

Where the friction rate actually comes from

Here is the part most free duct calculators get wrong, including this one until recently. They offer you a friction rate of 0.10 as a default, as though it were a property of residential systems.

It is not a constant. In Manual D it is a derived quantity, specific to the system in front of you:

Friction Rate = Available Static Pressure × 100 / Total Effective Length

Available Static Pressure (ASP) is what is left of the blower's rated external static after every device in the air path takes its cut — the wet cooling coil, the filter, the supply registers, the return grilles, the balancing dampers, and anything else in the stream. On a system with a 0.50" blower, a 0.25" wet coil and a 0.10" filter, more than two-thirds of the budget is gone before a single foot of duct is considered.

Total Effective Length (TEL) is the longest supply path plus the longest return path, in equivalent feet — measured straight run plus the equivalent length of every elbow, boot, takeoff and transition on that path. Fittings routinely account for more effective length than the straight duct does, which is why TEL surprises people the first time they compute it honestly.

The consequence is that two houses with identical airflow can have legitimately different friction rates. A compact system with a strong blower might land at 0.14. A sprawling one with a high-MERV filter and a 200 foot effective length can land near 0.03 — and at 0.03, the ducts Manual D demands are so large that the correct answer is usually to change the design rather than build them.

Use the "Friction rate — from my actual system" mode above to compute yours, then apply it to the sizing. It takes a minute and it is the difference between a Manual D duct size and a plausible-looking guess.

Flex duct: the compression problem

Everything in the flex chart above assumes the duct is pulled tight. Almost none of it is.

Flexible duct is a wire helix with a liner stretched over it. Fully extended, the liner is taut between the wire coils and the interior is merely rough. Let it relax, and the liner sags into the space between every coil, turning the inside of the duct into a corrugated washboard. Each ring becomes a small obstruction, and there are dozens of them per foot.

These are approximate multipliers drawn from published flexible duct compression testing, offered as order of magnitude rather than precision. The direction and the scale, however, are not in question.

Fifteen percent compression is not a rare abuse case. It is what you get when someone cuts a 25 foot length for a 21 foot run and does not trim the excess — which is routine, because trimming takes time and the extra material is already paid for. That single decision roughly doubles the pressure drop of the run. You cannot size your way out of a compressed flex duct. A 15 percent compressed 8 inch behaves worse than a fully extended 6 inch in some respects, and no chart in the world will tell you that.

The related sin: excess length coiled in the attic

Leftover flex looped into a lazy spiral above the ceiling adds length, adds bends, and usually adds compression all at once. If a run measures 14 feet and there is 22 feet of duct in the attic, the run is 22 feet long and it has an unaccounted-for coil in it. Cut it.

Why flex carries about 70% of what sheet metal does

Compare the two friction columns at any diameter. At 0.08 in/100ft, 10 inch flex carries 277 CFM and 10 inch galvanized carries 386. That ratio — right about 72 percent — holds remarkably steady across every size in the table.

It is entirely a roughness effect. ASHRAE puts fully extended flexible duct in the "rough" category at ε = 0.01 ft, against 0.0003 ft for galvanized steel with a longitudinal seam. That is a factor of thirty in absolute roughness.

The practical translation: swapping sheet metal for flex at the same nominal size is roughly a 30 percent capacity cut before anyone compresses anything. If a plan calls for 8 inch metal branches and the installer runs 8 inch flex because it is faster, the system that gets built is not the system that was designed.

Sizing a return grille

Returns are undersized more often than supplies, and the reason is that returns are ugly. A return big enough to be quiet is a large object on a wall or ceiling, and there is constant pressure to make it smaller.

The math is short. You need enough free area — actual open space between the blades, not the size of the frame:

gross face area (sq in) = CFM × 144 / (face velocity × free area fraction)

Two numbers drive it:

• Face velocity. For residential returns, 400 FPM is a good target and 500 FPM is the practical ceiling. Past that you hear it, and you hear it constantly because the return runs whenever the system runs. Supply registers tolerate more — 500 to 750 FPM — partly because throw needs some velocity to work.
• Free area. A stamped face grille is only about 65 to 70 percent open. Put a filter behind it and the effective figure drops again. Sizing on gross dimensions rather than free area is the single most common way returns end up undersized by a third.

For a 3 ton system at roughly 1,200 CFM through a 70 percent free grille at 400 FPM, you need about 617 square inches gross — a 25 × 25, or two 20 × 16 grilles. A single 20 × 20 is not enough, and a single 20 × 20 is what gets installed.

What undersized ductwork actually looks like

The symptoms are diagnostic if you know the pattern:

• The far rooms are weak, the near rooms are fine. Friction is spent along the path, so the run with the most path loses the most.
• Total external static pressure exceeds the equipment rating. This is the measurement that settles the argument. Most residential air handlers are rated at 0.5 in. wg total external static. Field measurements routinely find 0.8 to 1.2. The blower is then delivering far less air than nameplate, at every register, all the time.
• Frozen evaporator coils in cooling. Not enough air across the coil means the coil runs colder than design until it ices.
• Limit trips in heating. The same low airflow lets the heat exchanger overheat.
• Noise at the returns. Almost always the first complaint, almost always dismissed.

Note what is not on that list: how the air feels at the register. Hand-at-the-vent is not a measurement, and a high-velocity blast from an undersized register can feel stronger than adequate airflow from a correctly sized one.

The honest limitation of every duct calculator, including this one

This tool sizes a duct for an airflow you give it. It cannot tell you what that airflow should be.

Getting the CFM right requires a room-by-room load calculation — ACCA Manual J or equivalent — which accounts for orientation, glazing, insulation, infiltration, internal gains and the local design conditions. The airflow to each room follows from that room's load, not from its floor area and not from a per-ton rule of thumb.

The "400 CFM per ton" figure in the calculator hint is a total-system sanity check, nothing more. Dividing it up among rooms by guesswork is how you get a house where the master bedroom is always five degrees off.

Similarly, sizing individual ducts correctly does not by itself produce a working system. This page will compute a Manual D friction rate and size a duct with it, which is the core of the method — but a full Manual D design also requires the fitting-by-fitting equivalent length takeoff that produces TEL in the first place, equipment selection per Manual S, and the balance between every path in the system. A system built from individually correct ducts can still be badly unbalanced.

Worth knowing: the IRC also permits prescriptive duct sizing from Table M1601.1.1 — sized by equipment tonnage and outlet count — without any Manual D calculation. That exception is legal, and it is a large part of why so much installed residential duct has never seen a friction rate.

If the stakes justify it — a new system, a major renovation, a problem nobody has been able to solve — that is engineering work, and it is what my practice does. This page will get you a defensible duct size. It will not get you a duct design.