Size every duct, precisely. SMACNA gauge auto-selector, ASHRAE friction loss, multi-segment static pressure budget, and inspector-ready PDFs in your pocket.
Ductulator: Duct Pro HVAC Calc is coming soon to the App Store.
All Ductulator: Duct Pro HVAC Calc guides
At a friction rate of 0.10 in. wg per 100 ft, 400 CFM needs about 9.7 in. of round duct, so you install 10 in. stock or a 14x6 rectangular equivalent. Drop the friction rate to 0.06 and the same 400 CFM needs roughly 11 in.; push it to 0.15 and about 9 in. will do. Friction rate is the pressure budget you are allowed to spend per 100 ft of effective duct length, so a lower budget forces a bigger duct to move the same air. That is why there is no single correct duct size for a given CFM until you have calculated the friction rate for your specific system.
Use the Huebscher circular-equivalent formula, De = 1.30 x ((a x b)^0.625) / ((a + b)^0.250), which ASHRAE and ACCA Manual D both publish as a table. Equal area fails because a flattened rectangle has far more wetted perimeter than a circle of the same area, and that extra surface adds friction. A 14x6 duct has 84 sq in. of area, the same as a 10.3 in. round, but it only performs like a 9.8 in. round. The flatter the aspect ratio, the wider that gap gets, which is why 3-to-1 and worse rectangles are avoided.
Friction rate is the pressure drop the duct system is allowed to consume per 100 equivalent feet, expressed in inches of water gauge, and it is calculated as (Available Static Pressure x 100) / Total Effective Length. It is a computed output of your specific equipment and duct layout, not a value you pick off a chart. Most residential systems land between 0.06 and 0.18 in. wg per 100 ft once the real numbers are run. ACCA Manual D treats anything below about 0.06 as a sign the duct runs are too long or the blower too weak for the layout.
Start with the External Static Pressure the manufacturer's blower table lists for your air handler at the design airflow and blower tap, typically 0.5 in. wg for many residential units. Then subtract the pressure drop of every component the manufacturer did not include in that rating: the filter at its dirty rating, the cooling coil wet, the heat exchanger if applicable, plus registers, grilles, dampers, and any balancing devices. What is left is Available Static Pressure, the amount of pressure you actually have to push air through the sheet metal. Skipping the deductions is the single most common reason a duct design comes out undersized.
Total Effective Length is the straight duct footage of one run plus the equivalent length of every fitting on that run, and it is what you divide Available Static Pressure by to get friction rate. You measure the single longest effective supply run and the single longest effective return run, then add those two together. You do not add up every duct in the house, and you do not use an average. ACCA Manual D calls this the critical path, because the worst run sets the pressure budget that every other run then has to live inside.
It depends entirely on the fitting geometry. A smooth radius 90 with a centerline radius of about 1.5 times the diameter typically counts as 15 to 30 equivalent feet, while a square-throat 90 with turning vanes runs closer to 45 to 60, and a sharp mitered square 90 without vanes can be charged 75 to 80 equivalent feet. Flex duct bent through 90 degrees without a support saddle is worse still. This is why a duct run measuring 40 physical feet routinely comes out at 200 or more effective feet once the fittings are counted.
For residential systems, ACCA Manual D targets roughly 700 to 900 FPM in supply trunks, 600 to 900 FPM in branch runouts, and 500 to 700 FPM in return trunks. At the return grille face, keep the velocity at or below about 500 FPM, and at supply registers below roughly 750 FPM. Light commercial work tolerates higher numbers, often 1,000 to 1,500 FPM in trunks, because the duct is usually further from occupied space. Velocity is a cross-check on the size the friction rate gave you, not a substitute for it.
Regenerated noise in sheet metal duct generally becomes audible above about 900 FPM, is clearly noticeable above 1,200 FPM, and turns into an obvious rush or whistle past 1,500 FPM. Grilles and registers complain much earlier because the air is accelerating through fixed slots close to the listener, which is why 500 FPM is the practical ceiling at a return grille face. Sharp fittings, closed balancing dampers, and undersized returns raise local velocity well above the average you calculated for the duct. If a system whistles only when a door closes, the return path, not the duct size, is usually the culprit.
At 0.10 in. wg per 100 ft, a 6 in. galvanized round duct carries about 110 CFM, while 6 in. flex pulled fully taut carries roughly 94 CFM, a gap of about 15 percent. The difference comes from the helical wire and the fabric liner: ASHRAE puts galvanized roughness at 0.0003 ft and fully extended nonmetallic flex at 0.003 ft, roughly ten times higher. Those flex numbers assume the manufacturer's fully extended condition, which almost never survives real installation. Texas A&M bench testing found a 6 in. flex duct at only 4 percent linear compression down to about 70 CFM at the same 0.10 in. wg.
If you size flex from a rigid-duct chart, yes, go up about 7 percent in diameter, which usually means no change on a small branch and one nominal size up on a trunk 14 in. and larger. Compression is the bigger problem. The Air Diffusion Council tells you to double the friction rate at 15 percent linear compression and quadruple it at 30 percent, which at fixed available static works out to roughly 71 percent and 50 percent of the extended airflow. Bench data is harsher: Texas A&M measured a 6 in. duct at only 4 percent compression moving about 70 CFM at 0.10 in. wg against about 110 CFM for rigid metal. The International Mechanical Code (Section 603.10 in recent editions) caps flexible duct support spacing at 4 ft with sag limited to 1/2 in. per foot of spacing, and the ADC Flexible Duct Performance and Installation Standard adds hanger straps at least 1.5 in. wide. Pulling flex tight and supporting it properly is usually cheaper than upsizing the whole system.