A W14 can be twenty-two inches deep
The number in a W designation is a rolling group, not a depth. Most people who use these tables know that in the abstract and assume the error is small, the way it is small in a metric table. It is not small here.
The W14 group runs from 13.66 inches deep to 22.84 inches deep. That is nine inches of spread on a family of sections all called W14, and W14x808 is over half again as deep as the number in its own name. W27 reaches 32.5 inches. W12 reaches 16.8.
The reason is the same one that moves a British serial size, just running much further. Rolling a heavier section in a group does not pack more steel inside it, it adds metal to the outside of the flanges, so the heavy members of a group are deeper and wider as well as heavier. The heavy end of W14 is not a beam at all, it is a column, and W14x808 has flanges nearly five inches thick.
And the nominal is not a floor either. Every W group contains a section shallower than its own name: the lightest W24 measures 23.48. So a designation is wrong in both directions, which is worse than being consistently optimistic, because you cannot even build in a habit to correct for it.
The letter does not tell you beam or column, and 28 sections are neither
This is the real difference from British steel, and it goes deeper than notation. A British designation states the job: a universal beam is deep and narrow because it spans, a universal column is nearly square because it carries load down, and the letters are the answer. W is every wide flange section there is. You read the flange width against the depth or you do not know what you are holding.
The proportion does separate them cleanly, and it is worth saying how cleanly rather than offering a rule of thumb. In the British table every universal beam has a flange between 0.30 and 0.66 of its depth, every universal column between 0.88 and 1.05, and Britain rolls nothing in between. Those are hard edges, so they can be used to say what a British reader would call an American section.
Applied to the W list: 201 sections fall in the beam band and 47 in the column band. The other 28 are neither, and they are not exotic. W12x53, W10x33, W14x82 and W8x24 are everyday steel sitting in a proportion the UK does not roll at all. So the honest answer for those is that there is no UB or UC to compare them to, which is more use than forcing them into the nearer bucket and being quietly wrong.
W360x134 is not a metric section, it is W14x90
Every shape here has a metric designation as well, and it is the same piece of steel under another name rather than a comparable section. W360x134 is W14x90. W690x802 is W27x539. The mill prints both lists in one catalog precisely because they are the same rolling.
That matters because a metric W designation looks like something it is not. A reader who meets W360x134 on a Canadian drawing and reasons about it as a metric section, comparing it to a 356x171x67 UB because the first numbers are close, has misread it twice: wrong section, wrong standard. The search above takes either designation and shows both on the row.
It also catches the same trap the British page warns about, from the other side. The metric name carries a nominal group too, so W690x802 is nominally 690mm and actually 826. The group label is approximate in both unit systems.
Four of these sizes are Canadian, and they sit right next to American ones
W18x41, W18x45, W24x56 and W24x61 are drawn from the Canadian CSA list rather than ASTM A6. The catalog footnotes them and it is easy to miss, which is a problem, because each one sits a pound a foot away from an American section of a completely different shape:
| Section | Standard | Flange width | Weight |
|---|---|---|---|
| W24x55 | ASTM A6 | 7.005 in | 55 lb/ft |
| W24x56 | CSA | 8.9 in | 56 lb/ft |
| W24x61 | CSA | 8.93 in | 61 lb/ft |
| W24x62 | ASTM A6 | 7.04 in | 62 lb/ft |
One pound a foot apart in the name, nearly two inches of flange apart in reality. Somebody reading W24x56 has every reason to think it is a slip of the pen for the W24x55 they know, and it is not. This is where the US and Canada stop being one market: they share the designation system completely and the size list only mostly, so a Canadian drawing can legitimately call for a section that is not in the American list.
Same weight is same price, and the deeper one is not automatically better
Weight is what steel is bought, craned and carted by, and the weight is already in the designation, which is the one genuinely friendly thing about American naming. So the useful question is what the same pounds per foot can buy, and the spread is startling: W12x230 and W44x230 both weigh 230 lb/ft, and the W44 has eight and a half times the second moment of area.
Depth is not free, though, and it would be easy to write this section as though it were. The deeper section of a matched pair almost always has a smaller weak-axis radius of gyration, and an unrestrained beam does not fail by bending, it fails by buckling sideways, which the weak axis governs. Going from W21x55 to W24x55 buys 18 per cent more bending stiffness and gives up 23 per cent of the weak-axis radius to get it. In a column the trade is worse still, because weak-axis slenderness is what sizes it in the first place.
So the tool lists every section at a given weight with both figures side by side and recommends nothing. Which one is right depends on restraint that no table can know about.
What was wrong in the catalog, and what was left out
Section tables get copied from site to site and errors travel with them, so every row here was checked against its own geometry before it was published. The checks are identities rather than estimates: depth must equal the flat web depth plus two k, a radius of gyration must be the root of second moment over area, a plastic modulus cannot be smaller than an elastic one, and the area times the density of steel must reproduce the weight printed in the section's own name. Then every row was converted to millimeters and compared against the separately typeset metric table in the same catalog.
That found four faults in the source. W6x12 is printed 3.03 inches deep when its own columns require 6.03, and it is corrected, because two independent columns agree and a W6 three inches deep does not exist. W14x370 and W36x441both print a weak-axis second moment their own modulus and radius contradict, and in both the right figure is 1,990; for W14x370 the catalog's own metric table prints it correctly.
W6x9 is absent. Its printed modulus needs one second moment of area, its printed second moment needs another modulus, and its printed radius of gyration agrees with neither, while its plastic modulus is below its elastic one, which cannot happen. Three columns give three different answers, so there is nothing to correct it to. A wrong section modulus on a beam somebody sizes is worse than a gap, so it is a gap.
On the four Canadian sizes the inch and metric tables disagree by under two per cent on the strong-axis properties, and each is internally consistent, so there is no way to arbitrate between them from the geometry. The inch figures are the ones shown here. A Canadian working in millimeters will find slightly different numbers in the same catalog, and knowing that is better than assuming one of us has fat fingers.
Common questions
How deep is a W14x90?
14.02 inches, which is close to its name. But that is luck rather than a rule: the W14 group runs from 13.66 inches deep to 22.84 inches deep, so the 14 is a rolling group and not a dimension. W14x808 is 22.84 inches, over half again as deep as the number in its own designation.
Do all W24 sections measure 24 inches?
No, and neither more nor less consistently. The lightest W24 is 23.48 inches deep and the heaviest is 27.99. The extra weight goes on the outside of the flanges rather than inside the section, so a heavier member of a group is deeper and wider than a lighter one. An opening set out at exactly 24 inches can be wrong in either direction depending which W24 is delivered.
Does a W shape tell you whether it is a beam or a column?
No, and that is the biggest difference from British steel. A British designation says UB or UC, so the letters answer the question. W covers both, and W14x30 and W14x132 share a letter and a nominal depth while doing different jobs. You read the flange width against the depth instead, which the tool above does for you.
Is W360x134 a metric beam or an imperial one?
Both, because it is the same piece of steel. W360x134 and W14x90 are one rolled section under two designations, not two comparable sections. The tool shows the metric name on every row and finds a section if you type either one, because assuming a metric W designation is a different section is a real and expensive mistake.
What is the difference between W24x55 and W24x56?
About two inches of flange. W24x55 has a 7.005 inch flange and comes from ASTM A6; W24x56 has an 8.9 inch flange and comes from the Canadian CSA list. They are one pound a foot apart in the name and are not the same shape, so W24x56 is not a typo for the one you know. There are four of these Canadian sizes and the tool flags each one.
If two sections weigh the same, is the deeper one better?
For bending, usually by a lot. W12x230 and W44x230 are both 230 lb/ft, so both the same price, and the W44 has eight and a half times the second moment of area. But the deeper section normally has a smaller weak-axis radius of gyration, and an unrestrained beam fails by buckling sideways rather than by bending. W21x55 to W24x55 buys 18 per cent more bending stiffness and gives up 23 per cent of the weak-axis radius. Which matters depends on restraint, so the tool shows both figures and recommends neither.
What is the T dimension for, and why is it not the depth?
T is the flat web depth between the fillets, and it is what decides whether a stiffener, a clip or a bolt head fits inside the section. Take the depth, knock off two flanges and two fillet radii and there is a good deal less room than the headline number suggests. On a W14x90 the section is 14.02 deep and T is 10.113.
Where does this data come from?
Nucor-Yamato Steel's own structural shapes catalog, April 2023, from a mill that rolls these sections, and its notes state that the decimal dimensions for flange width, depth and both thicknesses follow ASTM A6/A6M. Every row was reconciled against its own geometry before publication. That found three errors in the catalog, all corrected here with the proof, and one row that could not be resolved and is left out rather than guessed.