Duct Sizer Airflow
The ductulator, solved four ways — built for diagnosing the duct you already have.
Inputs are seeded with an example — edit them to your numbers.
Input
The default direction is the diagnostic one: the duct already exists, a traverse or a BAS trend says what it’s moving, and the question is how hard it’s working. The other three are the classic ductulator spins — size to a friction rate, size to a velocity, or read off what a size you’re stuck with can honestly carry.
Output
Worked example — the 12-inch that outgrew its job
A renovated floor keeps starving its VAV boxes. The main serving it is 12 in. round, and a traverse says it’s moving 1,200 CFM — the seeded inputs are that duct.
- Velocity first:
A = π × (12 ÷ 24)² = 0.785 ft², soV = 1200 ÷ 0.785 = 1529 FPM— top of the mains band, loud but livable. Velocity alone doesn’t flag the problem. - Friction does:
0.26 in. w.c. per 100 ft— about three times the 0.08 the duct was likely sized to. Over this 80-ft run that’s0.26 × 0.8 = 0.21 in. w.c.spent on straight duct alone, before the first elbow or takeoff. - Flip the solve to Diameter: at 0.08 in/100 ft, 1,200 CFM wants a 15.3 in. round. Flip it to Airflow: a 12 in. at 0.08 was only ever good for about 626 CFM. The renovation roughly doubled the load on a duct nobody resized.
- Nobody’s re-ducting a finished ceiling — but the “why is the static so high / why won’t the boxes make flow” argument now has a number in it, and the honest options (a parallel run at the next reno, resetting expectations on the static setpoint, rebalancing to what the run can actually deliver) all start from that number instead of from blame.
Input
The conversion is Huebscher’s equivalent diameter — the same friction per foot at the same airflow, which is what a friction chart needs. It is deliberately not an area match: corners rub, so the rectangle carries more cross-section than the round it stands in for. Add the airflow and the output shows what that does to velocity.
Output
Worked example — same friction, ten inches of ceiling
The friction tab said the starved run really wants a 15.3 in. round. Above this ceiling there are 10 inches of clear depth — fifteen inches of pipe is not happening. The seeds carry the story on.
- Flip the solve to Missing rectangular side: to match De = 15.3 in. with one side pinned at 10 in., the Huebscher formula wants the other side at 20.2 in. The sheet-metal answer is 20 × 10 — flip back to Equivalent round and check it:
De = 1.30 × (20 × 10)^0.625 ÷ (20 + 10)^0.25 = 1.30 × 27.42 ÷ 2.34 = 15.2 in.— close enough at chart precision. - Now the gotcha the formula is built around. Equal friction is not equal area: the 20 × 10 gives
200 ÷ 144 = 1.39 ft², while a 15.2 in. round is only1.26 ft²— the rectangle needs about 10 % more cross-section for the same friction, because its corners contribute rubbing wall without contributing much flow. - So at the same 1,200 CFM the two run at different speeds:
1200 ÷ 1.39 = 863 FPMin the rectangle against1200 ÷ 1.26 = 952 FPMin the round. Same pressure gradient per foot, slower air. - The field consequence: never pick a rectangular size by matching area to the round you wanted — an area-match comes out undersized on friction. And anything read from velocity — pitot VP on a traverse, noise expectations at a takeoff — shifts when a run changes shape, even though the friction chart calls the two sizes “equivalent.”
Where these numbers come from — and where they stop
The friction math: Altshul-Tsal — the closed-form friction-factor fit ASHRAE Fundamentals offers in place of the Colebrook equation its duct friction chart is computed from — at galvanized-duct roughness (ε = 0.0003 ft) and standard air (0.075 lb/ft³ — sea level, ~70 °F): f′ = 0.11 × (12ε/D + 68/Re)^0.25, corrected to f = 0.85f′ + 0.0028 below 0.018, with Re ≈ 8.50 × D × V and Δp = (12 × f × 100 ÷ D) × 0.075 × (V ÷ 1097)² per 100 ft. Sanity anchor: 12 in. at 1,000 FPM computes 0.12 in. w.c./100 ft — the point every published chart agrees on. This tool is US-native (in. w.c./100 ft, FPM, CFM, inches) — the units every ductulator and friction chart is printed in — and the m/s readout rides along.
The sizing conventions, loosely held: low-pressure design practice clusters around 0.08–0.10 in. w.c. per 100 ft for supply duct, with mains landing near 1,000–1,500 FPM and branches near 600–900 FPM, return runs slower still. Those are noise-and-economics conventions, not physics and not code — a mechanical-room main can run harder, duct over a conference room can’t run close — so treat them as the band where nobody argues, not as limits.
Straight galvanized duct only. This page prices the straight runs; in a real system the fittings — elbows, tees, transitions, balancing dampers — usually cost more static than the duct between them, which is why a run can compute at 0.3 in. w.c. and measure at 1. Flex duct is worse than either: its listed friction assumes it’s pulled fully tight, and real-world compressed flex runs several times higher — treat hard-pipe numbers as the floor for flex, never the answer.
And a sizing answer is a theory, not a command. This page is a learning aid and a second opinion — a way to check whether a duct explanation is even plausible before anyone cuts metal or moves a setpoint. Acting on it has stakes in both directions: an undersized run drives static toward the duct high-limit’s territory, and an oversized one on a forward-curved fan lets the wheel ride out its curve into airflow — and amps — the motor may not have (the affinity-laws page tells that story from the speed side). The fan curve, the motor nameplate, and the high-limit do the governing.