Duct Traverse Calculator Airflow
The whole traverse, not one point — and the root goes before the average.
V̄ = 4005 × mean(√VP) — then multiplies by duct area for CFM. Averaging the pressures first and rooting once reads high, every time.
Ak-factor flow converts a diffuser face reading through the catalog’s effective area, Q = Ak × V, forward or back-solved against a hood.
Inputs are seeded with an example — edit them to your numbers.
Input
25 readings
Paste the readings straight off your log — spaces, commas, or line breaks all separate points, so one textarea line per duct row keeps the grid readable. If your instrument already converts to FPM (hot-wire, rotating vane), switch the reading type and paste velocities; the average is then taken directly, no roots involved.
Output
Worked example
A 24 × 12 in. supply duct, traversed on a 5 × 5 grid — the 25 seeded readings, laid out one textarea line per duct row, just as you’d log them.
- Root each reading first:
√0.0196 = 0.14,√0.0625 = 0.25,√0.0841 = 0.29… 25 roots in all. - Average the roots: they sum to 5.41, so
mean(√VP) = 5.41 ÷ 25 = 0.2164. - Mean velocity:
V̄ = 4005 × 0.2164 = 867 FPM. - Area and flow:
A = (24 ÷ 12) × (12 ÷ 12) = 2.0 ft², soQ = 867 × 2.0 = 1734 CFM.
Now the wrong order, for comparison: average the pressures first — mean(VP) = 1.2177 ÷ 25 = 0.0487 — and root once: 4005 × √0.0487 = 4005 × 0.2207 = 884 FPM, which is 2 % high, or 34 CFM of phantom air on this one duct. The square root bends down, so the root of an average always sits at or above the average of the roots: the wrong order can only ever read high, and the gap widens the more uneven the profile. On the honest side of the comparison, the two agree only when every reading is identical — which a real duct never is.
Input
Ak is the manufacturer’s effective flow area for that diffuser with a stated instrument held in a stated position — not the neck’s πr², not the face dimensions. Conventions differ by instrument, which is why the second mode exists: hood one diffuser, back-solve the Ak your anemometer actually sees, and carry it to the rest of the identical units.
Output
Worked example
A job has twenty identical louvered-face diffusers. The catalog lists Ak = 0.65 ft² for a rotating-vane anemometer at the face.
- The flow equation:
Q = Ak × V = 0.65 × 400 = 260 CFMat the measured 400 FPM face velocity. - Now the calibration direction: hood the first diffuser — it reads
273 CFMwhile the anemometer averages420 FPM. Back-solve:Ak = 273 ÷ 420 = 0.65— the catalog value checks out for your instrument, so the remaining nineteen convert with one multiply each.
The trap this tab exists to flag: Ak is not a geometric area. The same diffuser carries a different Ak for a hot-wire probe at the neck than for a vane at the face, and the neck’s πr² (0.79 ft² for a 12 in. neck) is a third number entirely — multiply by that instead of the 0.65 and this example reads roughly 20 % high. Use the catalog value for your model and your instrument, or back-solve your own.
Grids, plane location, and standard air
How many points: for rectangular duct, divide the cross-section into a grid of equal cells and read the center of each — at least 25 points (5 per side), going to 6 per side when a side passes 30 in. and 7 when it passes 36 in. For round duct, traverse two perpendicular diameters through ports 90° apart, 6 to 10 points per diameter. Either shape: the standards-grade spacing (log-Tchebycheff for rectangular, log-linear for round) crowds the points toward the walls to catch the boundary layer — the position tables live in ASHRAE 111 and your instrument manual, and matter more as the duct gets smaller.
Where the plane goes: the grid can’t rescue a bad plane. Aim for at least 7.5 straight duct diameters downstream of the last elbow, damper, or takeoff and 3 upstream of the next one. When the run doesn’t exist — usually — add points, expect scatter, and treat a reading that goes negative as the duct telling you the flow is reversing or tumbling at that spot: relocate the plane rather than averaging it in.
Standard air: the 4005 folds standard-air density (0.075 lb/ft³ — sea level, ~70 °F) into V = √(2 · ΔP / ρ), the same constant as the airflow tool. At altitude the air is thinner, the same VP means a higher true velocity, and this math under-reads by a few percent per few-thousand feet; most balancing instruments correct for density internally — when yours does, trust it over this page. Velocity-type readings inherit whatever density handling the instrument that produced them applied. This tool is US-native (in. w.c., FPM, CFM) — the frame pitot charts and Ak catalogs are published in — and the m/s readout rides along.