How to use this chart

Pick the friction rate your system is designed to, find your duct diameter in the left column, and read across. The CFM column is what that duct will carry at that friction rate; the velocity beside it is how fast the air is moving to do it, and the flag says whether that velocity is inside the limits ACCA publishes for residential duct.

One thing to settle before you read anything: 0.10 in. w.g. per 100 ft is a default, not a standard. It is the rate almost every published duct chart quietly assumes, and it is the rate this page prints, but Manual D calculates the friction rate for your system from its available static pressure and its total effective length. If your system's rate is 0.07, the 0.10 column is telling you about somebody else's house. Section 4 covers how to get your own number.

Duct CFM chart — round, flex and rectangular at 0.10 in. w.g. per 100 ft

at or below Manual D recommended velocity above recommended, at or below maximum above Manual D maximum

Duct CFM chart at a friction rate of 0.10 in. w.g. per 100 ft, standard air (0.075 lb/ft³, 70 °F, 29.92 in. Hg). Round: galvanized steel, absolute roughness ε = 0.0003 ft. Flex: fully extended, ε = 0.003 ft. Velocity flags are ACCA Manual D Table 3-1 supply limits as reproduced by ACCA — rigid 700 fpm recommended / 900 fpm maximum, flex 600 / 700. Computed with Darcy–Weisbach and a Swamee–Jain friction factor; not copied from any published chart.
Duct diameter	Round CFM	Round fpm	Round vs limit	Flex CFM	Flex fpm	Flex vs limit
4 in	37	424	OK	31	357	OK
5 in	68	497	OK	57	418	OK
6 in	111	564	OK	93	474	OK
7 in	168	628	OK	141	527	OK
8 in	240	688	OK	201	577	OK
9 in	329	745	Over rec.	276	625	Over rec.
10 in	436	800	Over rec.	366	670	Over rec.
12 in	709	903	OVER MAX	594	757	OVER MAX
14 in	1,069	1,000	OVER MAX	895	837	OVER MAX
16 in	1,525	1,092	OVER MAX	1,276	914	OVER MAX
18 in	2,084	1,179	OVER MAX	1,744	987	OVER MAX
20 in	2,755	1,263	OVER MAX	2,305	1,056	OVER MAX

The flex column is fully extended flex — pulled taut, supported, no sag. That is a laboratory condition, not an attic. Section 5 is what compression does to these numbers, and it is a bigger effect than everything else on this page combined.

The same chart at 0.05 and 0.08 in. w.g. per 100 ft

Duct CFM at friction rates of 0.05 and 0.08 in. w.g. per 100 ft, standard air (0.075 lb/ft³). Round: galvanized steel, ε = 0.0003 ft. Flex: fully extended, ε = 0.003 ft. Same ACCA Manual D supply velocity flags as the table above.
Duct diameter	Round CFM @0.05	fpm	vs limit	Flex CFM @0.05	fpm	vs limit	Round CFM @0.08	fpm	vs limit	Flex CFM @0.08	fpm	vs limit
4 in	25	287	OK	22	247	OK	33	374	OK	28	317	OK
5 in	46	337	OK	40	290	OK	60	439	OK	51	372	OK
6 in	75	384	OK	65	330	OK	98	499	OK	83	422	OK
7 in	114	428	OK	98	367	OK	148	555	OK	125	469	OK
8 in	164	469	OK	140	402	OK	212	608	OK	179	514	OK
9 in	225	509	OK	192	436	OK	291	659	OK	246	556	OK
10 in	298	547	OK	255	468	OK	386	708	Over rec.	326	597	OK
12 in	486	619	OK	415	529	OK	629	800	Over rec.	530	674	Over rec.
14 in	734	687	OK	626	586	OK	948	887	Over rec.	798	747	OVER MAX
16 in	1,048	751	Over rec.	893	640	Over rec.	1,352	968	OVER MAX	1,138	815	OVER MAX
18 in	1,434	812	Over rec.	1,222	691	Over rec.	1,849	1,046	OVER MAX	1,555	880	OVER MAX
20 in	1,898	870	Over rec.	1,615	740	OVER MAX	2,445	1,121	OVER MAX	2,056	942	OVER MAX

Rectangular duct is keyed by its dimensions rather than a diameter, so it has its own table in section 6.

How the CFM in this chart is calculated

Duct friction is the Darcy–Weisbach equation. ASHRAE publishes it in the inch-pound form that returns inches of water directly, as Equation 19 of the “Duct Design” chapter of the ASHRAE Handbook–Fundamentals:

Δp f = 12 f L / D h · ρ (V / 1097)² Δp f friction loss in total pressure, in. of water f friction factor, dimensionless L duct length, ft D h hydraulic diameter, in. V velocity, fpm ρ density, lbm/ft³

A friction rate is just this equation with L = 100. ASHRAE publishes no separate “per 100 ft” formula, which is worth knowing, because charts routinely quote a per-100-ft figure as though it came from a distinct constant somewhere.

The friction factor f here is the explicit Swamee–Jain approximation to Colebrook–White, which is the same correlation the duct size calculator and the duct sizing chart on this site already use, so all three pages agree with each other on the same physics. It was cross-run against an iterative Colebrook–White solution and against ASHRAE's own explicit Altshul–Tsal approximation (Equations 20 and 21 of the same chapter): the three agree within 0.8 % at every diameter and rate on this page, and Colebrook against Swamee–Jain within 0.35 %.

f = 0.25 / [ log₁₀( ε / 3.7D + 5.74 / Re^0.9 ) ]² Swamee–Jain Constants used for every value on this page: ρ air density 0.075 lb/ft³ ASHRAE standard air μ dynamic viscosity 1.204 × 10⁻⁵ lb/(ft·s) g 32.174 ft/s² 1 in. w.g. 5.202 lb/ft²

Standard air is ASHRAE's term for air weighing 0.075 lb/ft³, which approximates dry air at a temperature of 70 °F and a barometric pressure of 29.92 in. Hg (ASHRAE Terminology). That is what the whole chart is drawn for.

Roughness — why the material changes the answer

Absolute roughness ε is the one input that separates a metal duct from a flex duct of the same size. ASHRAE's “Duct Design” Table 1 sorts duct materials into roughness categories:

Duct roughness categories as published in ASHRAE Handbook–Fundamentals, “Duct Design,” Table 1. The two rows in bold are the values used on this page.
Category	ε, ft (mm)	Materials
Smooth	0.0001 (0.03)	Uncoated carbon steel, clean; PVC plastic pipe; aluminum
Medium Smooth	0.0003 (0.09)	Galvanized steel, longitudinal seams, 4 ft joints
Average	0.0005 (0.15)	Galvanized steel, continuously rolled, spiral seams, 10 ft joints; spiral seam with 1–3 ribs
Medium Rough	0.003 (0.9)	Fibrous glass duct, rigid; fibrous glass liner with facing material
Rough	0.01 (3.0)	Fibrous glass liner, spray coated; flexible duct, metallic (0.004–0.007 ft fully extended); flexible duct, all types of fabric and wire (0.0035–0.015 ft fully extended); concrete

ASHRAE's own friction chart is drawn at the Medium Smooth value: This chart is based on standard air flowing through round galvanized ducts with beaded slip couplings on 48 in. centers, equivalent to an absolute roughness of 0.0003 ft. Notice the consequence — spiral-seam galvanized duct is an Average material, so it sits slightly outside the envelope the chart itself claims. Very few published duct charts make that distinction, or any distinction at all.

The validity envelope ASHRAE states for its own chart, which is worth having in front of you: No corrections to Figure 9 are needed for (1) duct materials with a medium smooth roughness factor, (2) temperature variations in the order of ±30 °F from 70 °F, (3) elevations to 1500 ft, and (4) duct pressures from −20 in. of water to +20 in. of water relative to the ambient pressure. Outside those four conditions — a hot attic in Denver, say — the numbers need correcting, and this chart does not correct them for you.

The flex column on this page uses ε = 0.003 ft rather than Table 1's “Rough” 0.01 ft. That is a deliberate choice with a real conflict behind it, and section 5 states the conflict, marks the flex figures for verification, and prints the 0.01 ft alternative so you can see the size of the disagreement rather than taking our word for it.

Independent calibration

A published friction chart read-back gives a check that does not come from our own arithmetic. A PE continuing-education course from CED Engineering reads its friction chart as putting the intersection of the 0.1 in.-wc line and the 1,000 CFM line at a round duct diameter of 13.5 in. The model behind this page gives 13.65 in for the same point — +1.1 %, which is inside the precision of reading a value off a log-log chart at all.

Which friction rate should you use — 0.05, 0.08 or 0.10?

This is the part most duct charts skip, and skipping it is why their numbers disagree by more than half.

Manual D does not prescribe a friction rate. It calculates one. The friction rate is available static pressure divided by total effective length, times 100 — where available static pressure is the blower's total external static pressure less all the component pressure losses, and total effective length is the sum of the longest supply path and the longest return path. Two houses with the same equipment and different duct runs get different friction rates, and the same house gets a different rate when you add a better filter.

ACCA's own blog makes the point in its title: Calculating Friction Rate — It's Not a Constant. It observes that designers usually assume 0.1 iwc/100′, and argues against doing that. A widely cited worked example elsewhere puts it bluntly: if the calculated rate comes out just over 0.07 and you use 0.1 instead, your ducts come out undersized.

So what are the three rates on this page?

0.05 — supply trunk or plenum. Component guidance, not a whole-system rate.
0.08 — supply run-outs. Same source, same status. (The matching return-duct guidance in that source is 0.02.)
0.10 — the traditional single-rate default. Described in the PE course literature as a very common friction rate for a reasonably well designed system, and by ACCA as the number people assume when they have not calculated one.

None of these three is an industry-standard rate, and this page will not call any of them one. Several published duct charts describe 0.1 as “the industry-standard friction loss rate.” ACCA's own blog contradicts that directly. Manufacturer engineering literature puts the practical band at roughly 0.05–0.2 in. w.g. per 100 ft and calls 0.05–0.10 the most cost-effective range for an equal-friction design — a range, chosen per system, not a constant.

On Manual D's own permitted band the sources disagree and we are not going to pretend otherwise. The lower bound is unanimously 0.06. The upper bound is cited as 0.18 by an ACCA-attributed brochure and two independent write-ups, and as 0.16 by a third. Manual D itself is paywalled; three-to-one favours 0.18, and if your calculated rate lands near either end, that is the point to check a licensed copy rather than a web page.

Flex duct: why it carries less air, and what compression costs you

Two separate penalties sit on flex duct, they come from different places, and a chart that mixes them will be wrong in both directions. This page keeps them apart.

Penalty one: roughness, both ducts taut

At ε = 0.003 ft, fully extended flex carries 84–86 % of the CFM of galvanized steel at the same diameter and the same friction rate. That ratio is essentially flat from 4 in to 20 in, which makes the rule of thumb reliable: taut flex carries roughly one sixth less air than metal of the same size, so going up one nominal size gets it back. At ASHRAE Table 1's “Rough” value of 0.01 ft the ratio falls to 70–73 %.

Verify against standard. The flex column on this page is marked for verification, and here is exactly why. ASHRAE Table 1 classifies fabric-and-wire flexible duct as Rough, ε = 0.01 ft, with a measured range of 0.0035–0.015 ft fully extended. This page publishes at 0.003 ft, which is below that band. The reason is that laboratory measurement of taut flex does not support the Table 1 design value: testing at Texas A&M's Energy Systems Laboratory under the ASHRAE Standard 120 method found that results for the maximum stretch case and rigid duct showed agreement within 2% — that is, taut flex measured almost as smooth as metal. Our 0.003 ft sits between that finding and Table 1, and matches the value already used across this site's duct pages. It is a judgement, not a standard, and it is the one number here most likely to move against a licensed ASHRAE copy. The table below shows what the conservative Table 1 reading would give instead.

Sensitivity case — flexible duct at ASHRAE Table 1 “Rough”, ε = 0.01 ft, standard air. This is *not* the value used in the main chart above; it is printed so the size of the roughness disagreement is visible. Same equations and constants otherwise.
Duct diameter	CFM @0.05	fpm	CFM @0.08	fpm	CFM @0.10	fpm
4 in	18	207	23	264	26	296
5 in	33	244	42	310	47	348
6 in	55	278	69	354	78	396
7 in	83	310	105	394	118	442
8 in	119	340	151	433	169	485
9 in	163	370	208	470	232	526
10 in	217	398	276	505	309	566
12 in	354	451	449	572	503	641
14 in	535	500	679	635	760	711
16 in	765	548	970	695	1,086	778
18 in	1,047	593	1,329	752	1,488	842
20 in	1,387	636	1,760	807	1,970	903

And the honest framing, because the laboratory result is easy to misread: fully extended, board-supported flex is a laboratory condition. The Air Diffusion Council's installation standard exists precisely because field flex is not that. Taut flex measures nearly as smooth as metal in a lab; the chart derates it because installed flex is not taut — and the next section is what that actually costs.

Penalty two: compression, and the disagreement that matters

ADC publishes a friction-rate multiplier for longitudinal compression, and it is the number a contractor can cite in a submittal:

Install ducts fully extended. Do not install in the compressed state or use excess length as this will noticeably increase friction losses.

When a flexible duct is fully extended, it is said to have no more than 4% longitudinal compression and the published friction rate may be used for duct sizing calculations (0 - 4% = 1 x Friction Rate).

For 15% longitudinal compression the friction rate can increase by a factor of two (15% = 2 x Friction Rate).

For longitudinal compression of 30% the friction rate can increase as much as four times (30% = 4 x Friction Rate).
ADC, Flexible Duct Performance & Installation Standards, 5th Edition

Lawrence Berkeley National Laboratory measured the same thing and got a much larger number. Its published finding is that moderate compression in flexible ducts, typical of that often seen in field installations, could increase the pressure drop by a factor of four, while further compression could increase the pressure drop by factors close to ten — with around 15 % described as the moderate field case and around 30 % as the extreme case, the same two compression levels ADC uses. A separate laboratory programme at Texas A&M, run to the ASHRAE Standard 120 test method, reported the same direction: some configurations show over ten times the pressure loss found in rigid ducts or fully stretched flexible ducts of the same diameter, and that the existing published design data does not include pressure loss data for flexible ducts that are compressed beyond approximately 4%.

Turning a friction multiplier into a capacity loss needs the exponent linking pressure drop to flow. It is not 2. Measured off this page's own flex model at 0.08 in./100 ft it runs from about 1.89 at 4 in to about 1.95 at 20 in, so the table below uses n = 1.90 and derates CFM by the friction multiplier raised to −1/1.90.

Flex compression derating — two published bases, side by side, deliberately not averaged. ADC column: the trade installation standard's friction multipliers. LBNL column: the laboratory pressure-drop correction factor from the published 6 in model, PDCF = 1 + 25.35 rc, where rc is the change in length divided by the fully stretched length. CFM derate = multiplier raised to −1/1.90. Multiply any flex figure in the chart above by the derate to get the compressed capacity.
Longitudinal compression	ADC friction multiplier	CFM derate (ADC)	LBNL PDCF, 6 in	CFM derate (LBNL, 6 in)
0–4 % (“fully extended”)	1×	1.00	2.01×	0.69
15 % (“typical field install”)	2×	0.69	4.80×	0.44
30 % (“extreme”)	4×	0.48	8.61×	0.32

Read those two derate columns against each other, because the gap is the point. Take a 6 in flex run at 0.10 in./100 ft — 93 CFM taut. At the 15 % compression both sources call a typical field installation, the trade standard says it still moves about 64 CFM. The laboratory model says about 41 CFM. That is a difference of more than a third of the airflow, between two sources that are both published, both quotable, and both about the same duct. We are not going to average them into a number that is nobody's answer. If you are sizing to a submittal, the ADC column is the one you can cite; if you are trying to work out why a room is not getting air, the laboratory column is the one that predicts what you are measuring.

Note also that LBNL's own 6 in model returns a pressure-drop factor of 2.01 at just 4 % compression — directly contradicting ADC's “0 − 4 % = 1 × friction rate” at exactly the small end where residential branch runs live.

Fittings: what a bend costs in flex

Compression is not the only thing that eats a flex run's budget. From the same ADC standard, expressed as equivalent length of straight duct:

A 90-degree bend ≈ 20 lineal feet of flexible duct.
A gradual 45-degree bend ≈ 10 lineal feet.
A 180-degree offset ≈ 40 lineal feet.
Support spacing must hold maximum centreline sag to ½ in per foot of spacing between supports — sag is compression by another name.

A 25 ft flex run with two 90-degree bends is, for pressure purposes, closer to 65 ft. That is why the calculated friction rate in section 4 uses total effective length rather than tape-measure length.

Rectangular duct and equivalent diameter

A rectangular duct is matched to a round one by Huebscher's equivalent diameter, published as Equation 25 of the ASHRAE “Duct Design” chapter, with a and b in inches:

De = 1.30 (ab)^0.625 / (a + b)^0.25

ASHRAE gives the provenance in its own words: Huebscher (1948) developed the relationship between rectangular and round ducts that is used to determine size equivalency based on equal flow, resistance, and length. The primary is R. G. Huebscher, Friction equivalents for round, square and rectangular ducts, ASHVE Transactions 54, pp. 101–118 (1948).

Equal friction and equal flow — not equal area. This is where published rectangular charts most often go wrong. The CFM of a rectangular duct is the CFM of the round duct of that equivalent diameter. But the velocity must be computed on the actual rectangular free area, which is larger than the round equivalent. A 12 × 8 in duct has De = 10.66 in and carries 517 CFM at 0.10 in. w.g./100 ft. Its round equivalent would read 835 fpm; the true velocity in the rectangular duct is 775 fpm. A chart that prints the round-equivalent velocity is printing a number that exists nowhere in the system. The velocities in the table below are on the real rectangular area.

Rectangular galvanized duct capacity at 0.05, 0.08 and 0.10 in. w.g. per 100 ft, standard air (0.075 lb/ft³), ε = 0.0003 ft. De is the Huebscher equivalent diameter; CFM is the CFM of the round duct of that De; velocity is computed on the actual rectangular free area. Flag column is the ACCA Manual D supply rigid limit (700 fpm recommended / 900 fpm maximum) applied at 0.10.
Rectangular size	Aspect	De (in)	CFM @0.05	fpm	CFM @0.08	fpm	CFM @0.10	fpm	vs limit @0.10
3.25 × 10 in	3.08:1	6.00	75	334	98	434	111	492	OK
3.25 × 12 in	3.69:1	6.49	93	345	121	448	137	506	OK
3.25 × 14 in †	4.31:1	6.93	111	353	145	458	164	518	OK
6 × 6 in	1.00:1	6.56	96	384	125	498	141	563	OK
8 × 6 in	1.33:1	7.55	140	421	182	546	206	618	OK
10 × 6 in	1.67:1	8.40	187	448	242	581	274	657	OK
12 × 6 in	2.00:1	9.14	234	469	304	607	343	686	OK
14 × 6 in	2.33:1	9.80	283	485	366	628	414	709	Over rec.
8 × 8 in	1.00:1	8.75	208	468	270	607	305	686	OK
10 × 8 in	1.25:1	9.76	280	504	362	652	409	736	Over rec.
12 × 8 in	1.50:1	10.66	354	531	458	687	517	775	Over rec.
14 × 8 in	1.75:1	11.46	430	553	556	714	627	806	Over rec.
16 × 8 in	2.00:1	12.19	507	570	655	737	739	832	Over rec.
18 × 8 in	2.25:1	12.86	585	585	756	756	853	853	Over rec.
20 × 8 in	2.50:1	13.48	664	597	857	771	967	870	Over rec.
22 × 8 in	2.75:1	14.06	743	608	959	785	1,082	886	Over rec.
24 × 8 in	3.00:1	14.61	823	617	1,062	797	1,198	899	Over rec.
10 × 10 in	1.00:1	10.93	379	546	490	706	553	797	Over rec.
12 × 10 in	1.20:1	11.96	482	579	623	748	704	844	Over rec.
14 × 10 in	1.40:1	12.89	589	606	761	782	858	883	Over rec.
16 × 10 in	1.60:1	13.73	697	628	901	810	1,016	914	OVER MAX
20 × 10 in	2.00:1	15.23	920	662	1,187	854	1,339	964	OVER MAX
12 × 12 in	1.00:1	13.12	617	617	797	797	899	899	Over rec.
16 × 12 in	1.33:1	15.11	900	675	1,161	871	1,309	982	OVER MAX
20 × 12 in	1.67:1	16.80	1,194	716	1,539	923	1,735	1,041	OVER MAX
24 × 12 in	2.00:1	18.28	1,494	747	1,926	963	2,171	1,086	OVER MAX

† Aspect-ratio validity. ASHRAE publishes no algebraic aspect-ratio limit on Equation 25; it shades a “not recommended” region in its circular-equivalents table, which is a buildability constraint rather than a validity one. The only quantified accuracy statement we could source is from a peer-reviewed conference paper by Hassan (AIVC 2002), reporting a study by Zou (2001): for normal ventilation duct aspect ratios, a/b < 4, the deviations are always less than 1% for Reynolds numbers between 4,000 and 10⁷. Every row in the table at or below 4:1 is therefore good to about 1 %. The 3.25 × 14 in row is 4.31:1 and sits above that band; it is marked, and its figures should be treated as slightly less certain than the rest.

Velocity limits — when a duct is big enough but too loud

A duct that satisfies the friction rate can still be the wrong duct, because velocity is what you hear. ACCA publishes velocity limits in Manual D Table 3-1; the values below are that table as reproduced by ACCA in its own Table of Useful Air Distribution System Design Information.

ACCA Manual D Table 3-1 velocity limits, fpm, as reproduced by ACCA. Marked *verify against standard*: Manual D itself is paywalled, and at least one other circulating transcription of this table disagrees with this one. See the note below the table.
Duct	Supply rec. rigid	Supply rec. flex	Supply max rigid	Supply max flex	Return rec. rigid	Return rec. flex	Return max rigid	Return max flex
Trunk ducts	700	600	900	700	600	600	700	700
Branch ducts	600	600	900	700	400	400	700	700

Face velocities, from the same ACCA-reproduced table.
Where	Face velocity (fpm)
Supply outlet, sized for throw	700
Return grille	500
Filter grille	300

Three things in that table contradict what most people assume, and all three are useful:

Return branch recommended velocity is 400 fpm, not 600. This is probably the most commonly mis-stated number in residential duct design.
Filter grille face velocity is 300 fpm. That single number is why return filter grilles have to be physically enormous: 1,200 CFM through a filter grille needs four square feet of free area.
Flex maximum caps at 700 fpm everywhere — supply and return, trunk and branch.

And now the finding that the chart at the top of this page exists to make visible. At 0.10 in. w.g. per 100 ft — the rate essentially every published duct CFM chart uses as its default — round rigid duct is over Manual D's 900 fpm maximum at 12, 14, 16, 18 and 20 in, reaching 1,263 fpm at 20 in, which is 40 % over the ceiling. Flex is over its 700 fpm maximum from 12 in up and over recommended from 9 in up. At 0.08 the rigid breach starts at 16 in and the flex breach at 14 in. At 0.05, rigid duct is inside the maximum at every size from 4 to 20 in — but 20 in flex is still over 700 fpm, which means there is no friction rate at which a 20 in flex duct is velocity-clean under Manual D. The rectangular table breaches at 16 × 10, 20 × 10, 16 × 12, 20 × 12 and 24 × 12 at 0.10.

This is not a defect in the physics, and it does not mean the CFM figures are wrong. It means the equal-friction method and the velocity ceiling are two separate constraints, and on large duct at a high friction rate they stop agreeing. A chart that prints the CFM and hides the velocity is showing you a duct that satisfies one constraint and quietly fails the other.

Verify against standard. Manual D is paywalled and we do not hold a licensed copy. A state licensing training handout reproduces “Table 3-1” with a different and internally inconsistent arrangement — it rates the trunk below the branch, which is backwards. The face velocities agree exactly across both versions (700 / 500 / 300). We use the ACCA-attributed brochure version above because it is self-consistent and because ACCA published it; but neither version is the primary, and these limits should be checked against a licensed Manual D before anything is designed to them.

On noise, and what we are not going to claim

There is no authoritative published “above X fpm a duct gets noisy in a house” figure that we could source. The industry encodes the noise limit implicitly, as Manual D's Maximum column. A secondary summary of a paywalled ASHRAE applications chapter does circulate with specific velocity thresholds attached; we are not printing those, because the attribution could not be verified against the standard itself. What is worth carrying is the caveat, which is both citable and true:

Velocity limits are commonly used as a surrogate for limiting duct breakout noise. Many argue it is a poor indicator since noise is more likely to result from turbulence than velocity.
PDHonline Course M246, HVAC Ducting: Principles and Fundamentals

Which matches what people actually experience: a straight run at 900 fpm is often quieter than a sharply turned one at 600.

Why published duct CFM charts disagree

Before you take any duct CFM chart at face value — this one included — it is worth knowing how far apart the published ones are. We transcribed the 10 in round rigid figure from the six duct CFM chart pages that currently rank for these queries, along with whatever basis each one states for its numbers. They are anonymised here because the point is the pattern, not the pages.

Published capacity for a 10 in round rigid duct across six ranking duct CFM chart pages, with the basis each page states for its own numbers. Transcribed 2026-09-12. Sources are anonymised deliberately; they are cited here as evidence that published charts disagree, and no value from any of them is used anywhere on this page.
Chart	Basis stated on the page	10 in round, as published
Chart A	None stated	270 CFM
Chart B	None stated	300 CFM (flex table) / 325 CFM (metal table)
Chart C	Column headings name a friction-loss figure; no units given, never “per 100 ft”	300 CFM (lower column) / 420 CFM (0.1 column)
Chart D	700 fpm target velocity; no friction rate at all	350 CFM
Chart E	0.1 in. w.c. per 100 ft, single straight run, no fittings	360 CFM
Chart F	0.1 in. w.c. per 100 ft	300 CFM (one table) / 400 CFM (a second table on the same page)
This page	0.10 in. w.g. per 100 ft; round galvanized, ε = 0.0003 ft; standard air ρ = 0.075 lb/ft³; Darcy–Weisbach with a Swamee–Jain friction factor; velocity 800 fpm, flagged over ACCA recommended	436 CFM

The published spread for one ordinary duct size is 270 to 420 CFM — a factor of 1.56. At 20 in the published spread runs from about 1,620 to 2,600 CFM, with three separate pages printing exactly 2,000 while citing three different bases or none. Specific defects in that set, without naming anyone:

Two of the six state no basis whatever — no friction rate, no velocity, no material, nothing. One of those two is the top-ranking page for the query.
One contradicts itself on the same page at the same stated rate: its by-diameter table and its quick-reference table give different capacities for a 10 in duct, both labelled 0.1 in./100 ft.
One publishes a non-monotonic set — its prose rates a 12 in duct below its own 10 in table entry, and its prose and its table disagree about the 8 in figure as well. Its table has three rows in total.
One is velocity-based, not friction-based. Constant-velocity capacity scales as d²; equal-friction capacity scales as roughly d^2.5. Its numbers are structurally non-comparable to the rest of the set and only converge with them by coincidence near 8–10 in.
Two of the six give a friction rate with units, properly. Those two are the only ones in the set whose numbers can be checked at all.

The reason we are not reconciling these, or averaging them, is that averaging unsourced numbers imports their error and launders it into something that looks authoritative. The only external number this page treats as corroboration is the friction-chart read-back in section 3, which is a published chart with a stated basis and which this page matches to +1.1 %.

Duct CFM chart FAQ

How many CFM in a 6 inch duct?

About 111 CFM in round galvanized steel at a friction rate of 0.10 in. w.g. per 100 ft, at 564 fpm. At 0.08 it is 98 CFM and at 0.05 it is 75 CFM. In fully extended flex the same 6 in duct carries 93 CFM at 0.10. Compressed flex carries far less — at the 15 % compression both ADC and the laboratory literature call a typical field install, that 93 CFM falls to roughly 64 CFM on the trade standard's multiplier, or roughly 41 CFM on the laboratory model.

How many CFM can 8 inch flex duct handle?

About 201 CFM fully extended at 0.10 in. w.g. per 100 ft, at 577 fpm — which is inside ACCA's 600 fpm recommended velocity for flex, so 8 in flex at that rate is a clean size. At 0.08 it is 179 CFM and at 0.05 it is 140 CFM. All three assume the run is pulled taut and supported; the word “handle” does a lot of work in this question, and compression is what usually decides the real answer.

Is flex duct the same as metal duct of the same size?

No. At the same diameter and the same friction rate, fully extended flex carries 84–86 % of the airflow of galvanized steel, and that ratio holds flat from 4 in to 20 in, so in practice flex needs to go up one nominal size to match metal. That is the taut comparison. Installed flex that is compressed 15 % carries between 69 % and 44 % of its own taut rating depending on whose data you use, which is a much bigger penalty than the roughness difference.

What friction rate should I use?

The one your system calculates, which is available static pressure divided by total effective length, times 100. If you have not calculated it, 0.10 in. w.g. per 100 ft is the traditional default and the rate this chart prints, 0.08 is common guidance for supply run-outs and 0.05 for supply trunks. Manual D's own permitted band starts at 0.06 and is commonly cited as reaching 0.18, with one source saying 0.16. None of these is an industry standard rate, whatever other charts call them — and assuming 0.10 when your system actually calculates to 0.07 will undersize the duct.

How many CFM per ton of cooling?

The usual residential design figure is 400 CFM per ton, which puts a 3-ton system at about 1,200 CFM of total supply air. That is an airflow target, not a duct size: 1,200 CFM through a single trunk at 0.10 in./100 ft needs something between a 14 and 16 in round duct, and at 16 in it would be running at 1,092 fpm, well over ACCA's 900 fpm maximum. That is exactly the case for splitting into trunks rather than sizing one large duct. To go the other direction — airflow to size — use the duct sizing chart.

Why is my duct big enough but still noisy?

Because friction rate and velocity are two different constraints and they stop agreeing on large duct. At 0.10 in. w.g. per 100 ft, round duct from 12 in up exceeds ACCA's 900 fpm maximum velocity while still satisfying the friction rate perfectly — a 20 in duct at that rate runs at 1,263 fpm. Flex is over its own 700 fpm ceiling from 12 in up. Velocity is also only a proxy for noise: turbulence from a sharp turn, a crushed takeoff or a boot too close to the register generates more noise than the same air moving straight, which is why a duct that reads fine on a chart can still whistle.

Does this chart work for rectangular duct?

Yes — the rectangular table gives 26 common sizes from 3.25 × 10 to 24 × 12 in at all three friction rates. Each is matched to a round duct by Huebscher equivalent diameter, which means equal friction and equal flow, not equal area. The velocity shown is calculated on the actual rectangular free area, which is the number that exists in your duct; charts that print the round-equivalent velocity instead are printing a number that is not happening anywhere.

Can I trust these numbers for a real design?

They are computed, stated and checkable, which is more than most published charts offer — every equation, constant and roughness value is named above, and the model reproduces an independently published friction-chart read-back to +1.1 %. Two groups of numbers here are explicitly marked verify against standard: the flex roughness value, and the Manual D velocity limits, both of which sit behind paywalled standards we do not hold a licensed copy of. For anything stamped, permitted or paid for, this chart is a sanity check and the licensed standard is the authority.

Sources

ASHRAE Handbook–Fundamentals, “Duct Design” chapter — Equations 19 (Darcy–Weisbach, I-P form), 20 (Colebrook–White), 21 (Altshul–Tsal) and 25 (Huebscher equivalent diameter); Table 1, Duct Roughness Factors; the validity note at Figure 9.
ASHRAE Terminology, “standard air” — the 0.075 lb/ft³ definition used for every value here.
ACCA, Calculating Friction Rate — It's Not a Constant — the friction-rate formula and the argument against assuming 0.1.
ACCA, Table of Useful Air Distribution System Design Information — Manual D Table 3-1 velocity limits and face velocities, as reproduced by ACCA.
Air Diffusion Council, Flexible Duct Performance & Installation Standards, 5th Edition — compression friction multipliers, bend equivalent lengths, sag limit.
Abushakra, Walker & Sherman (Lawrence Berkeley National Laboratory), Compression Effects on Pressure Loss in Flexible HVAC Ducts — the PDCF models and the factor-of-four / factor-of-ten finding.
Weaver & Culp (Texas A&M Energy Systems Laboratory), Static Pressure Losses in 6, 8, and 10-inch Non-Metallic Flexible Ducts — ASHRAE Standard 120 test method; taut flex within 2 % of rigid; compressed configurations over ten times the loss.
CED Engineering PE course M06-032, HVAC: How to Size and Design Ducts — standard air density, the friction-chart read-back used for calibration, and the 0.05 / 0.08 / 0.02 component guidance.
Hassan, AIVC 2002 conference paper — Huebscher (1948) provenance and the Zou (2001) aspect-ratio accuracy statement.
PDHonline Course M246, HVAC Ducting: Principles and Fundamentals — the velocity-as-noise-surrogate caveat.
McGill AirFlow, Duct System Design Guide, Ch. 2 — the practical friction-loss band for equal-friction design.

ASHRAE Handbook–Fundamentals Chapter 21 and ACCA Manual D are both paywalled. Every ASHRAE and Manual D figure quoted above comes from a university-hosted chapter copy, a national-laboratory document, ACCA's own blog and brochure, or peer-reviewed papers. Before this page is treated as authoritative, both should be checked against licensed copies.