Rebar Size Chart: Every Standard Bar in Nine National Standards

Nine different bars are called ten

Count them. A drawing that specifies a number ten bar could mean any one of nine things, depending on which country the drawing came from and which designation system the detailer had in mind that morning.

An American number ten bar in the inch and pound system is 1.270 inches across and carries 819 square millimetres of steel. An American number ten bar in the soft metric system is 9.5 millimetres across and carries 71. Those are both ASTM A615 designations. They live in the same standard, in adjacent columns of the same table, and they differ by a factor of eleven and a half.

Meanwhile a Canadian 10M is 100 square millimetres, a Japanese D10 is 71.33, and a plain ten millimetre bar in Britain, Europe, India, China or New Zealand is somewhere between 78.5 and 78.6 depending on which national table you open. That is nine entries, spanning 71 to 819 square millimetres, all going by the number ten.

The tool at the top of this page has all nine of them, along with ninety two other standard bar sizes from nine national standards. Type a designation into the lookup box and it will tell you every bar in the world that answers to that name and how far apart they are. Everything else on this page exists because I kept finding that the published charts disagreed with each other, and I wanted to know which of them was right.

The short answer, in most cases, is that none of them were copying the standard. They were recomputing it.

A bar number is a local address, not a global one

There is a reasonable instinct that a bar size ought to mean something physical. Sixteen millimetres ought to be sixteen millimetres. Number five ought to be five of something.

That instinct is right about half the time and it is the half that gets people into trouble, because when a designation does encode a physical dimension it encodes it in the units and the conventions of the country that issued it, and those do not survive a border crossing.

Four different logics are in play across the nine standards on this page.

The first is diameter in eighths of an inch. This is the American system for the small bars. A number three bar is three eighths of an inch across, a number four is four eighths, a number eight is one inch exactly. Clean, memorable, and only true up to the number eight.

The second is diameter in millimetres. Europe, India, China and New Zealand all use it. A twelve millimetre bar is twelve millimetres across. This is the only system where the number means what it appears to mean with no further explanation required.

The third is a legacy imperial size that no longer describes the bar. This is the Canadian system and it is the one that catches people. A 15M bar is not fifteen millimetres across. It is sixteen. The number is a historical artefact from an imperial size that the bar replaced, and the physically meaningful number for a Canadian bar is not the diameter at all, as I will get to.

The fourth is the rounded metric equivalent of an imperial diameter. Japan does this, and so does the American soft metric column. A Japanese D10 is 9.53 millimetres across, which is three eighths of an inch to two decimal places. The American metric designation for a number three bar is also ten, for the same reason, arrived at independently.

Which is why nine bars are called ten. Three separate logics land on the same number for three physically different bars, plus a fourth bar that arrives there by counting eighths of an inch up to ten.

The American table, and where each column actually comes from

ASTM A615, the specification for plain carbon steel reinforcing bars, carries its dimensions in Table 1. The same table appears unchanged in A706, the low alloy weldable specification, so the geometry below applies to both. Only the chemistry and the mechanical properties differ.

Here is the inch and pound column, as printed:

Notice that the printed precision is not uniform. The weights for bars three through eleven carry three decimals, and the weights for the fourteen and eighteen carry two. That is not a transcription slip on my part. The standard prints 7.65 and 13.60, not 7.650 and 13.600, and a chart that silently pads them to three decimals is quietly telling you the value is more certain than the standard claims it is.

The area column is where the recomputation problem starts. Take the number four bar. Half an inch across, so the geometric area is pi times a quarter squared, which is 0.19635 square inches. The standard prints 0.20. That is a rounding of nearly two percent, and it is the largest single discrepancy in the table.

Now take the weight. Steel at 490 pounds per cubic foot works out to 3.40272 pounds per foot for every square inch of section. Multiply the printed area of 0.20 by that and you get 0.6805 pounds per foot. The standard prints 0.668. Multiply the true geometric area of 0.19635 instead and you get 0.66818, which rounds to 0.668 exactly.

So the weight column is computed from the true geometry and the area column is a rounded convenience. If you have ever wondered why a rebar weight calculator and a rebar area calculator give you slightly inconsistent answers, that is why. They are not derived from each other.

I ran this check across all twelve rows. Computing weight from the true diameter reproduces every published weight to within 0.17 percent, and the worst offender is the number nine bar at 0.07 percent. Computing weight from the printed area misses by up to 1.9 percent. The weight column is the trustworthy one.

Why the big bars have strange diameters

A number three bar is three eighths of an inch. A number eight is eight eighths, which is one inch. So far the numbering system is exactly what it looks like.

Then a number nine bar turns out to be 1.128 inches across, which is 9.03 eighths. Close enough. A number ten is 1.270, which is 10.16 eighths. Also close. A number eleven is 11.28 eighths. Drifting. And a number fourteen bar is 13.54 eighths of an inch, which is not fourteen by any reasonable rounding, and a number eighteen is 18.05.

The explanation is that bars nine and above are not round bars that were assigned numbers. They are replacements for square bars, and they inherited the square bar's cross sectional area rather than its dimension.

Work it through. A one inch square bar has an area of exactly 1.0000 square inches. The round bar with that same area has a diameter of the square root of four over pi, which is 1.12838 inches. The standard prints 1.128. Its weight at 3.40272 pounds per foot per square inch is 3.4027, and the standard prints 3.400.

Every one of the large sizes falls out of that calculation:

The number fourteen row is the proof. Its diameter is 13.54 eighths, so it cannot be named for its diameter. It is named for the one and a half inch square bar it replaced, and one and a half inches is twelve eighths, so it is not named for that either. The most plausible reading is that the numbering was extended to keep the sequence roughly proportional to the size, and 13.54 rounded up rather than down.

There is a practical consequence buried in this. The gaps in the sequence, the missing twelve, thirteen, fifteen, sixteen and seventeen, are not sizes that were withdrawn. They never existed. The jump from #11 to #14 is a jump from 1.56 to 2.25 square inches, a forty four percent step, and there is nothing in between. If a design needs 1.8 square inches in a single bar, the answer is that no such bar exists and the detail has to change.

The number twenty bar that most charts have never heard of

Current editions of ASTM A615 contain twelve rows, not eleven. There is a No. 20 bar, designated [64] in metric, 2.500 inches across, 4.91 square inches of section, and 16.69 pounds per foot. It is confirmed present in the 2020 edition and in every edition since.

Almost no published rebar chart includes it. Search for a rebar size chart and you will find the sequence ending at #18 more or less universally, which was correct for a long time and is now out of date.

Before anyone details with one, the standard itself attaches an unusual warning to that row. The gist is that design codes may not recognise the No. 20 bar, that structural members reinforced with it may need approval from the building official, and that special detailing is required to ensure adequate performance at service and factored loads.

That is a remarkable thing for a materials standard to say about one of its own products. It amounts to: this bar exists, we specify it, and you may not be allowed to use it. ACI 318 does not necessarily recognise the size, so a designer working to that code cannot simply reach for a #20 the way they would reach for a #18.

The tool includes the #20 in every table for completeness, and flags it in the development length table for exactly this reason. Treat its numbers as arithmetic rather than as a design option.

One more note on the American size list. Grade 75 used to be part of A615 and is not any more. Current editions cover Grades 40, 60, 80 and 100. If you are reading a chart that lists Grade 75 under A615, that chart predates the change.

The soft metric column, and the argument about 199

Somewhere in the 1990s the American reinforcing industry adopted metric designations without changing a single bar. This is called soft conversion, and it means the physical product is identical and only the label changed. A number five bar became a number sixteen bar. A number eight became a number twenty five. Nothing in the mill changed.

The metric designation is the inch diameter converted to millimetres and rounded to a whole number. Five eighths of an inch is 15.875 millimetres, which rounds to sixteen, hence #16. One inch is 25.4, hence #25. That rule fits all twelve rows without exception.

Then there is the area column, and here published charts disagree with each other. Some print 199 square millimetres for the #16 bar and some print 200. Some print 129 for the #13 and some print 127.

I chased this because it seemed like the kind of thing that should have one answer. It does. The standard prints 199 and 129. Confirmed against three independent full text copies of ASTM specifications, which agree with each other exactly.

The interesting part is why the other numbers exist, because the disagreement is not random. It comes from two different derivation routes.

Route one: take the published inch area and convert it. There are 645.16 square millimetres in a square inch, so the #13 bar at 0.20 square inches becomes 129.03, which rounds to 129. Correct.

Route two: take the rounded millimetre diameter and compute the area from it. The #13 is 12.7 millimetres, so pi times 12.7 squared over four is 126.68, which rounds to 127. That is where 127 comes from.

Route one, converting the published inch area, reproduces ten of the twelve rows. Route two reproduces six. So route one is the dominant rule and 129 is right.

Except at the #16 bar, where route one gives exactly 200.00 and the standard prints 199. And at the #64 bar, where route one gives 3167.7 and the standard prints 3167. Both of those rows match route two instead.

The derivation panel in the tool runs this audit live and colours the two failures red. It is the only place I know of where you can see the whole table checked against both rules at once.

I want to be careful about what this means. ASTM does not publish a derivation formula. Table 1 is a table of fixed values with no stated method. What I have described is a fit to the published numbers, not a quotation from the standard, and I would not want anyone building a spreadsheet on the assumption that the rule is official. The practical instruction is simpler and does not depend on the rule at all: use the printed values and do not recompute them. A chart that recomputes will disagree with the standard at the #13, and a chart that applies the dominant rule blindly will disagree at the #16.

The mass column is the well behaved one. Every metric mass is the inch and pound weight multiplied by 1.48816, and that fits all twelve rows exactly. 0.668 becomes 0.9941, printed as 0.994. 1.043 becomes 1.5521, printed as 1.552.

What nominal means, and the ninety four percent rule

Every dimension in the table above is a nominal dimension, and the standard's definition of that word is more specific than the everyday sense of approximately.

The definition is that the nominal dimensions of a deformed bar are those of a plain round bar having the same weight per unit length as the deformed bar. That is a definition by mass, not by measurement. You cannot measure a bar's nominal diameter with a caliper, because the caliper will catch the ribs and read high, and between the ribs it will read low. The nominal diameter is the diameter of an imaginary smooth bar that weighs the same.

This matters for two reasons. The first is that it explains the whole structure of the table. Weight is the primary measured quantity, area and diameter are derived from it, and that is why the weight column reconciles with geometry and the area column does not quite.

The second is the tolerance, which is one sided and stricter than most people assume. A615 requires that the mass of the test specimen be at least ninety four percent of the tabulated value. There is no upper limit. Overweight is explicitly stated never to be grounds for rejection.

Read that carefully. It is not plus or minus six percent. It is a floor at ninety four percent with nothing above. A mill has no commercial incentive to run heavy, since steel costs money and the bar is sold by weight or by length depending on the contract, so real production sits below nominal rather than scattered around it.

The practical consequence for anyone doing a takeoff is that a nominal tonnage is a purchasing estimate rather than a prediction of what arrives on the truck. On a job with a thousand tonnes of steel, the difference between nominal and the permitted floor is sixty tonnes.

Other standards handle this differently and the tool has a tolerance panel showing the four whose clauses I could read directly. BS 4449 and EN 10080 both allow plus or minus 4.5 percent above eight millimetres and plus or minus 6.0 percent at eight and below, which is a two sided band rather than a floor. IS 1786 uses a size graded band on a batch basis: plus or minus seven percent up to and including ten millimetres, five percent from over ten up to and including sixteen, and three percent above sixteen.

I have left the Canadian, Australian, Japanese and Chinese tolerances out of the tool. They exist, but I could not source them well enough to print, and a guessed tolerance is worse than an absent one.

The ribs are specified, not decorative

The deformations on a reinforcing bar look like a manufacturing flourish and they are not. A615 sets three numbers for every bar size: a maximum average spacing, a minimum average height, and a maximum gap.

The direction of those limits catches people out. Spacing has a maximum, because ribs that are too far apart do not grip. Height has a minimum, because ribs that are too shallow do not grip either. Gap has a maximum, because the longitudinal ribs interrupt the transverse ones and too much interruption reduces the bearing area. It is easy to get spacing backwards and assume it has a minimum, and several published summaries do.

The limits are proportional to the bar. Maximum average spacing is seven tenths of the nominal diameter across all twelve sizes, to the precision printed. Minimum average height runs at about four percent of the diameter on the smallest bars and five percent in the middle of the range. Maximum gap is 0.38268 of the diameter, which is the sine of 22.5 degrees, meaning the gap may subtend at most an eighth of the circumference.

That last one is a nice piece of design. The limit is not an arbitrary dimension, it is a fraction of the perimeter expressed as a chord, which keeps it dimensionally consistent across the whole size range.

Reading the marks rolled into the bar

Every deformed bar carries four marks rolled into one face. In order down the bar they are: the producing mill, the bar size, the type of steel, and the minimum yield strength.

The steel type letter is the one worth memorising. S is carbon steel to A615. W is low alloy steel to A706, which is the weldable one with restricted chemistry. Seeing both S and W on the same bar means it is dual certified to both specifications, which is common now because A706 inherently satisfies A615.

The grade marking runs on two parallel systems and the bar carries one or the other. Either a number, or a count of continuous longitudinal lines running through at least five deformation spaces.

Grade 40 is identified by the absence of a mark, which is an uncomfortable convention because an unmarked bar and a bar whose marks you cannot read look the same. Grade 40 is also furnished only in sizes three through six, so a large unmarked bar is a bar to ask questions about rather than a Grade 40.

The metric number marks are the first digit of the metric grade, so 420 becomes 4, 550 becomes 6 and 690 becomes 7.

I have to flag a genuine conflict here rather than paper over it. The line counts above follow the ASTM text. CRSI, which is a strong secondary source, publishes a sentence that assigns three lines to both Grade 80 and Grade 100 and four lines to Grade 120. That contradicts the standard, which gives 80 three lines and 100 four. I have gone with the standard, but if you are identifying a bar in the field on the basis of four lines, confirm it against a current copy rather than against this page.

Two further details in this area I could not corroborate and so will not assert: the letter C as an alternative Grade 100 marking, which appears in one source and nowhere else, and the metric number marks, which come from the same single extraction. They are probably right. Probably is not good enough for a marking convention.

The 2020 tensile change that most charts have not caught up with

Open almost any rebar grade chart and it will tell you that ASTM A615 Grade 60 has a minimum tensile strength of 90,000 psi. That was true for decades. It stopped being true in the 2020 edition.

Grade 60 dropped by ten thousand psi and Grade 80 by five thousand. Grades 40 and 100 were left alone.

Read on its own that looks like a weakening of the specification, and it would be if it were the only change. It is not. The same edition introduced a new requirement of a completely different kind: the ratio of actual tensile strength to actual yield strength must be at least 1.10, measured on the test specimen.

That is the clever part, and it is worth understanding because it changes what the specification is doing. The old regime specified two independent minimums and relied on the gap between them to guarantee a margin. If a bar over yielded, coming in at 75,000 psi against a specified 60,000, it still only had to reach 90,000 tensile, so its actual margin over yield was fifteen thousand rather than thirty.

Under the new rule that same bar must reach 82,500 psi, because 75,000 times 1.10 is 82,500. The margin now scales with the actual strength of the steel rather than with the specified minimum. A bar that over yields has to over fracture to match.

The stated reason for the change is alignment with A706, which already used 80,000 and 100,000. Since A706 inherently meets the requirements of A615, dual certified W bar can be accepted as either.

Any chart still printing 90,000 for A615 Grade 60 is describing a specification that has not existed for several years. It is also, unhelpfully, describing a number that is still perfectly plausible, since most real Grade 60 bar comfortably exceeds 90,000. The error will not show up as a failure. It will show up as a rejected mill certificate somewhere down the line.

Canada starts from the area, and everything else follows

Every standard on this page derives its table from one primary quantity and computes the rest. Most of them start from the diameter. Canada does not.

CSA G30.18 starts from a round nominal area and works backwards.

Look at the area column. One hundred, two hundred, three hundred, five hundred, seven hundred, a thousand, fifteen hundred, twenty five hundred. Those are not the outcome of a calculation. They are the design.

The diameters are then back calculated. The round bar with an area of 100 square millimetres has a diameter of 11.284 millimetres, printed as 11.3. Two hundred gives 15.958, printed as 16.0. All eight rows fall out of that square root.

And the masses are the round areas multiplied by 0.00785 with no rounding at all. One hundred times 0.00785 is 0.78500 exactly, and the standard prints 0.785. Fifteen hundred times 0.00785 is 11.77500, printed as 11.775. Every single row is exact.

Compute the mass from the diameter instead and the agreement breaks. The 10M at 11.3 millimetres gives 0.7873, not 0.785. The 30M gives 5.5119 against a published 5.495. So the chain runs strictly one way: round area, then diameter, then mass.

This is arguably a better system than everyone else's. Design work in reinforced concrete is done in areas. A slab needs so many square millimetres of steel per metre, and a Canadian detailer can do that arithmetic in their head because the bars come in round hundreds. Nobody working in millimetre diameters can do that.

The cost is that the designation has come completely adrift from the geometry. A 15M bar is sixteen millimetres across. A 25M is twenty five point two, which is close enough to be misleading. A 45M is forty three point seven.

On currency: the 2021 edition is the third edition and it kept all eight sizes. What changed was the grade list, which dropped 500R and added 600W. Current grades are 400R, 400W, 500W and 600W, where R is the general purpose designation and W the restricted chemistry one with enhanced ductility and weldability.

The ten M trap

This one deserves its own section because it is the most likely to actually hurt somebody.

A Canadian 10M bar carries 100 square millimetres of steel. An American number ten bar in the metric designation carries 71, because that designation refers to a number three bar. They are both small bars, both used for the same sorts of things, and both called ten.

The Canadian bar has 41 percent more steel.

That is a large enough difference to matter and small enough that nothing will look wrong. Substitute one for the other in a slab and the bars will fit the same clearances, sit in the same chairs, and produce a member that is under reinforced by nearly a third with no visible sign of it.

The inverse error is more dramatic and much less likely, because an American number ten bar in the inch and pound designation is 819 square millimetres, which is 8.2 times the Canadian 10M. Nobody accidentally substitutes a bar eight times heavier. It will not fit the chairs, it will not bend on the same machine, and the delivery will weigh eight times what the schedule says.

The dangerous confusions are always the ones near enough in size to pass inspection. 10M against #10 metric, at 41 percent, is exactly in that band.

Britain and Europe are two different lists, and the difference bites

People treat BS 4449 and EN 10080 as the same thing. They are not, and they do not carry the same sizes.

EN 10080 lists fourteen preferred diameters: 4, 5, 6, 8, 10, 12, 14, 16, 20, 25, 28, 32, 40 and 50 millimetres.

BS 4449 lists 6, 7, 8, 9, 10, 12, 16, 20, 25, 32, 40 and 50, with the 6, 7 and 9 flagged as preferred for the manufacture of welded fabric only. Its stated bar range is six to fifty millimetres.

There is no 14 millimetre bar in BS 4449, and no 28. Both are EN 10080 preferred diameters. A chart with a single column headed Europe or UK is wrong for one side of that pairing, and a specification calling for 14 mm bar to BS 4449 is calling for something that standard does not list.

There is a second and more subtle problem with the European tables, and it is the best example on this page of a standard failing to follow its own instructions.

Both standards state the rule explicitly. EN 10080 says the nominal mass per metre values are calculated from the nominal cross sectional area using a density of 7.85 kilograms per cubic decimetre. BS 4449 says the same in different words.

Apply that rule literally, using the printed area, and you get the wrong answers. Five rows disagree:

All five are explained if the mass came from the unrounded area rather than the printed one. Twelve millimetres has a true area of 113.097, and 113.097 times 0.00785 is 0.8878, which rounds to 0.888. The printed 113 gives 0.8871, which rounds to 0.887.

So the printed area column and the printed mass column were computed to different precisions, and a website that faithfully applies the standard's own stated method arrives at numbers the standard does not print. The 12 millimetre and 40 millimetre rows are the reliable tells for whether a chart was copied or recomputed. Use 0.888 and 9.86.

On grades, EN 10080 is unusual in that it sets none at all. It says explicitly that technical classes should be defined in accordance with it, and leaves the actual figures to national standards. It cannot be used on its own. BS 4449 supplies the British ones:

All three are 500 MPa steels. The letter is a ductility class, not a strength class, and it is the number that actually distinguishes them in a seismic or redistribution context. B500A is a cold worked wire product with very little post yield capacity. B500C is the ductile one and is the only class with an upper bound on the tensile to yield ratio, which matters because steel that is too strong relative to its yield point defeats capacity design.

Below eight millimetres the B500A requirements relax further, to a ratio of 1.02 and an elongation of 1.0 percent.

These map onto the Eurocode 2 ductility classes A, B and C. Note that one widely circulated industry guide prints B500C as capped at 1.30 and adds an upper bound to A and B that the standard does not impose. Two verbatim copies of the standard give under 1.35 for C and no upper bound for A and B.

India, and a table that disagrees with itself

IS 1786 covers 4, 5, 6, 8, 10, 12, 16, 20, 25, 28, 32, 36 and 40 millimetres. There is no 14 millimetre bar, and there is a 36, which Europe does not list.

The mass column is identical to the European one, which makes sense since the underlying geometry is the same. The area column is not, and this is where it gets awkward.

IS 1786 prints 78.6 for the 10 millimetre bar where exact geometry gives 78.54. It prints 201.2 for the 16 where geometry gives 201.06, 314.3 for the 20 where geometry gives 314.16, and 1257.2 for the 40 where geometry gives 1256.64. Every one of them is slightly high.

The standard also states its own density rule: the density of steel shall be taken as 0.00785 kilograms per square millimetre of cross sectional area per metre. Apply that to the standard's own printed areas and three rows fail against the standard's own printed masses:

The area and mass columns cannot both be right. The mass column matches exact geometry and matches the European tables; the area column does not match either. I have reproduced what the standard prints, and I would treat the mass column as the reliable one, but I want to be honest that all my sources for the area column trace back to scans of the same document. If you are working to IS 1786 and the third significant figure of the area matters, check a purchased copy.

The grades are the part of IS 1786 most people actually want, and its approach to tensile strength is unlike anything else on this page. Rather than specifying a fixed tensile minimum, it specifies a percentage above the actual measured proof stress, with an absolute floor underneath.

The D suffix is the higher ductility variant. Same proof stress, higher minimum elongation, a larger required margin over yield, and an additional requirement of at least five percent total elongation at maximum force that the plain grades do not carry.

Notice that IS 1786 arrived at the ratio idea long before ASTM did. The 2020 American change that introduced a 1.10 tensile to yield ratio is doing the same job that the Indian percentage above actual proof stress has been doing for years.

A 2017 amendment added Fe 650 and Fe 700, and lists Fe 415S and Fe 500S alongside the D grades. I could not confirm what the S denotes or what its requirements are, so I have left both out of the tool.

Australia and New Zealand: one standard, two different size lists

AS/NZS 4671 is a joint standard, and it contains two separate dimension tables because the two countries do not use the same sizes.

The Australian table runs 12, 16, 20, 24, 28, 32 and 36 millimetres. That is a four millimetre increment series, and it contains neither 10 nor 40.

The New Zealand table runs 6, 10, 12, 16, 20, 25, 32 and 40. That is a European style series.

The two share only 12, 16, 20 and 32. An Australian N24 or N28 has no New Zealand equivalent, and a New Zealand 25 has no Australian one.

There is a small rounding difference between them too, which is a nice illustration of how these tables get built. New Zealand prints the 40 millimetre area as 1260, to three significant figures. Europe prints 1257, to four. Same bar, same geometry, different house style.

The designation format is worth getting right because the shorthand everyone uses is not what the standard says. The full designation has four parts: shape, then grade, then ductility class, then size. A deformed ribbed bar of grade 500 MPa normal ductility steel at 16 millimetres is D500N16.

D is deformed, R is plain round, I is indented. The number is the lower characteristic yield stress in megapascals, so 250, 300 or 500. The letter after it is the ductility class: L for low, N for normal, E for earthquake.

So when an Australian drawing says N12, that is shorthand with the shape and grade dropped. The full designation is D500N12. R10 is short for R250N10. There is also an S designation in the Australian market meaning swimming pool grade, which is a corrosion related product classification and not a ductility class at all.

The mechanical properties are where the ductility classes earn their keep:

The 500L row is striking. A uniform elongation of 1.5 percent and a tensile to yield ratio of only 1.03 describes a steel with almost nothing beyond yield. It has its uses, in fabric and in members where ductility is not being relied on, and it is emphatically not interchangeable with 500N despite carrying the same strength number.

The table above is from the 2001 edition. The 2019 edition exists and I could not get a reliable copy of its equivalent table. Two separate reads of the one accessible partial copy returned contradictory values, including a 500L uniform elongation of twelve percent, which would make low ductility steel more ductile than seismic grade and is therefore certainly a misread. I have kept the 2001 values and labelled them, and I would not use any 2019 figure I have seen quoted online without checking it.

The same caveat applies to the dimension tables. The market clearly supplies N10 and N40 in Australia today even though the 2001 table lists neither, so the size list has probably been extended. I could not confirm how.

Japan: imperial diameters wearing metric names

JIS G 3112 is the only standard on this page that prints its own formulas, which makes it the easiest one to verify and the most pleasant to work with.

Look at those diameters. 9.53 is three eighths of an inch. 12.7 is a half. 25.4 is one inch. 38.1 is one and a half. 50.8 is two. The Japanese series is the American series with metric labels, arrived at by the same soft conversion logic and landing on almost the same designations, which is why a D16 and an American #16 are within half a percent of each other.

The stated derivation is: perimeter equals 3.142 times the diameter, area equals 0.7854 times the diameter squared rounded to four significant figures, and unit mass equals 7.85 thousandths of the area rounded to three significant figures.

Two details there are quietly interesting. JIS uses 0.7854 rather than pi over four, and 3.142 rather than pi. Those are truncations of the constants, written into the standard as the constants. And the mass is computed from the rounded four figure area, which is the opposite of the European convention where the mass comes from the unrounded value. I checked all twelve rows and the rule reproduces every published mass exactly.

China: the longest size list, and a one digit disagreement with Europe

GB 1499.2 was revised in 2024 and, notably, dropped the T from its designation, moving from a recommended standard to a mandatory one. It covers HRB400, HRB500 and HRB600, the fine grain HRBF variants, and the seismic E grades.

Its diameter range is six to fifty millimetres in fifteen sizes: 6, 8, 10, 12, 14, 16, 18, 20, 22, 25, 28, 32, 36, 40 and 50. That is the most complete list on this page. It includes 18 and 22 millimetre bars that Europe does not list, and 14 and 28 that Britain does not.

The Chinese table computes theoretical weight from the rounded four significant figure area, which is the Japanese convention rather than the European one. It produces two visible disagreements with Europe:

Neither is wrong. They are the same bar computed under two different rounding conventions, and on a large order the difference is real money even though it is a tenth of a percent. This is the kind of thing that turns into an argument between a supplier and a quantity surveyor who are each reading a different standard and each convinced the other cannot do arithmetic.

Nine names for the same piece of steel

The mirror image of the ten problem is more encouraging. Filter the full dataset down to bars within two percent of 200 square millimetres and nine entries come back, from nine different standards, and they agree to within 1.3 percent of each other.

Five different names, one bar. The world converged on five eighths of an inch and then spent a century arguing about what to call it.

The same convergence happens at 32 millimetres, where D32, the American #10, N32 and seven national 32 millimetre bars all sit between 794 and 819 square millimetres, and at 36 millimetres where 35M, the American #11 and the Chinese, Indian and Australian 36 all sit between 1000 and 1020.

What that convergence does not guarantee is interchangeability. Two bars with the same area can have entirely different grades, entirely different ductility classes, and entirely different weldability. A B500A and a D500N are both nominally 500 MPa 16 millimetre bars and one of them has an elongation at maximum force of 2.5 percent while the other has 5.0. The tool says so when it reports a convergence, and it is worth repeating: matching the steel area is the easy part.

Choosing a bar size for a slab

The size chart is the reference half of this page. The design half starts here, because knowing that a #5 bar has 0.31 square inches of section only helps once you know how many of them a slab needs.

Slab and wall reinforcement is specified as an area per unit width rather than as a bar count, and the conversion is the one piece of arithmetic every detailer does daily. A bar of area A at a spacing s provides A times the width divided by s. In imperial units, with a foot as the width, that is:

As per foot = A times 12 divided by s, with s in inches.

In metric, with a metre as the width, it is A times 1000 divided by s with s in millimetres.

The imperial version has a pleasant property that makes it easy to sanity check. At twelve inch spacing the arithmetic cancels and the answer is just the bar area. A #4 at twelve inches gives 0.20 square inches per foot, which is the #4 bar area. A #5 at twelve gives 0.31. Halve the spacing and you double the steel, so a #5 at six inches gives 0.62.

The metric version has the same property at 1000 millimetre spacing, which nobody uses, so the useful mental anchor is different. A 15M at 200 millimetres gives 1000 square millimetres per metre, because 200 square millimetres times five bars per metre is a thousand. Canadian round areas make this arithmetic easy in a way that no other system does.

The slab tab of the tool does this in both directions. Give it a required area and it returns, for every bar in the chosen standard, the widest spacing that will satisfy it. Widest, because wider spacing means fewer bars, and fewer bars means less labour, and the cheapest compliant answer is the one you want.

It also checks two things that a plain area calculation does not.

Shrinkage and temperature steel, and the change nobody mentions

Concrete shrinks as it cures and moves as it warms. Steel placed to control the cracking that causes is called shrinkage and temperature reinforcement, and it is specified as a ratio of the gross concrete area rather than as a structural requirement.

Under the 2019 edition of ACI 318 the ratio is 0.0018. Flat. No grade dependence, no reduction for high strength steel.

That is a change from earlier editions and it is a change that very few published tables have absorbed. The 2014 edition used a small table: 0.0020 for deformed bars below 60,000 psi, 0.0018 at 60,000 and above, and for higher strengths a reducing value of 0.0018 times 60,000 over fy with a floor at 0.0014.

The 2019 edition deleted the grade dependence entirely. The reasoning given is that the mechanics of shrinkage cracking do not care how strong the steel is, so a higher yield strength buys you nothing for crack control and permitting a reduction was never justified.

The consequence for anyone working from an old table is specific. Under the old rules, Grade 100 steel could be reduced to 0.0018 times 60 over 100, which is 0.00108, floored at 0.0014. Under the current rules it is 0.0018. That is a 29 percent increase in the shrinkage steel for a high strength slab, and if your spreadsheet still has the reducing formula in it, your slabs are light.

Going the other way, a Grade 40 slab under the old rules needed 0.0020 and now needs 0.0018, so an old table is conservative there rather than dangerous. The tool uses the flat 0.0018 and says so, and warns that a project running on the 2014 edition needs 0.0020 for Grade 40.

The companion limit is maximum spacing, which is the lesser of five times the member thickness and eighteen inches. Both matter. A thin slab is governed by the five h rule and a thick one by the eighteen inch cap, and the crossover is at a thickness of 3.6 inches.

This second constraint is why the slab tab does not simply return the widest spacing that satisfies the area. Early in building the tool it did exactly that, and then flagged the answer as non compliant if the spacing broke the limit. That is a useless answer: it tells you what you cannot do without telling you what you can. The tool now searches for the widest spacing that satisfies the area and stays inside the spacing cap, and only falls back to reporting a non compliant answer when no spacing in the table satisfies both, at which point the honest advice is to use a smaller bar.

Development length, where the table and the equation are the same equation

A reinforcing bar only works if it is bonded to the concrete over enough length to develop its yield force. That length is the development length, and ACI 318 gives you two ways to find it.

The general equation, in the 2019 edition, is:

ld = (3 over 40) times [ fy divided by (lambda times the square root of f prime c) ] times [ psi t times psi e times psi s times psi g, divided by the confinement term ] times db

The confinement term is (cb plus Ktr) divided by db, capped at 2.5. cb is the smaller of the cover to the bar centre and half the centre to centre spacing. Ktr accounts for transverse reinforcement crossing the splitting plane and may always be taken as zero. There is a flat twelve inch minimum.

The modification factors are:

The product of psi t and psi e need not exceed 1.7, which caps the worst case at a top cast epoxy bar.

The second route is a small table of four simplified equations, which most people use because it avoids having to know cb and Ktr. The equations look like this: for good spacing and cover, fy over 25 lambda root f prime c times db for #6 and smaller, and over 20 for #7 and larger. For all other cases, three fy over 50 for the small bars and three fy over 40 for the large ones.

Here is something I have not seen stated anywhere, and it makes the whole structure obvious once you see it. The four simplified equations are the general equation evaluated at two specific confinement values.

Take three fortieths, apply psi s, and divide by a confinement term of 1.5:

• Small bars: 3 over 40, times 0.8, divided by 1.5, equals 0.04, which is one twenty fifth.
• Large bars: 3 over 40, times 1.0, divided by 1.5, equals 0.05, which is one twentieth.

Now divide by a confinement term of 1.0 instead:

• Small bars: 3 over 40, times 0.8, divided by 1.0, equals 0.06, which is three fiftieths.
• Large bars: 3 over 40, times 1.0, divided by 1.0, equals 0.075, which is three fortieths.

Those are exactly the four table constants. I ran the identity across 192 combinations of bar size, grade and concrete strength and the two routes agree to the last digit every time.

So the simplified table is not an approximation and it is not a separate method. It is the general equation with the confinement term pinned at 1.5 for the favourable case and 1.0 for everything else. Which tells you precisely what using the table costs you: if your actual cover and transverse steel give a confinement term better than 1.5, the general equation is shorter by exactly the ratio, and the table is throwing that away. The cap at 2.5 means the most the general equation can ever save you over the favourable table case is a factor of 2.5 over 1.5, or forty percent.

The tool runs both routes side by side on every calculation and tells you which is which. When they agree it says so and explains why. When they disagree it says by how much and in which direction, and it will not claim agreement when both have simply been clamped to the twelve inch floor, which is a distinction the first version of the tool got wrong.

What high strength steel actually costs you

The grade factor psi g is new in the 2019 edition and it is the provision that governs whether high strength reinforcement is worth using.

The pitch for Grade 100 is straightforward. It yields at 100,000 psi against Grade 60's 60,000, so it carries 67 percent more force for the same bar. Fewer bars, less congestion, easier placing, less labour.

The catch is that development length scales with fy in the numerator and now also picks up psi g. Going from Grade 60 to Grade 100 multiplies fy by 1.667 and psi g by 1.30, so the development length goes up by a factor of 2.17.

You get 67 percent more capacity and pay 117 percent more embedment for it. In a member with plenty of length that is a clear win, because bar count drives cost and anchorage length is free. In a short member, a coupling beam, a joint region, anywhere the bar has to develop inside a limited dimension, it can be a straight loss.

The tool says this explicitly whenever you select a grade above 60, and gives you both multipliers so you can see the trade rather than being told the answer.

One related change worth knowing: the 2019 edition caps fy at 80,000 psi when computing minimum flexural reinforcement in beams, so Grade 80 and Grade 100 get identical minimums. The minimum itself is the greater of three root f prime c times bw d over fy, and 200 bw d over fy, and those two expressions are equal at a concrete strength of 4444 psi. Below that the 200 expression governs, above it the root expression does.

There is also a new floor on Ktr for high strength bars: for steel at or above 80,000 psi spaced closer than six inches on centre, Ktr shall not be taken less than half the bar diameter. One source writes that threshold as at or above 80,000 and another as above 80,000, which matters at exactly Grade 80, and I could not resolve it. The tool takes Ktr as an input and does not apply the floor automatically, so nothing turns on it here, but it is worth checking against your own copy if you are detailing Grade 80.

Lap splices

Splices are a straightforward multiplier on the development length, with a two class system and a twelve inch floor.

Both Class A conditions have to hold at once. Excess steel alone is not enough, and staggering alone is not enough. In practice most splices end up Class B, because getting a provided over required ratio of two means placing twice the steel the analysis calls for, which rarely happens except where minimum reinforcement governs.

The ld used in the splice calculation is computed without the excess reinforcement reduction that the code otherwise permits, which is easy to miss. And the 2019 edition removed that reduction entirely for hooked, headed and mechanically anchored bars, and prohibits it in seismic force resisting systems in the higher design categories.

Hooks, and the two changes in six years

When a bar cannot be developed straight, it gets hooked, and the hook development length is where the 2019 edition made its most consequential change.

The 2019 equation is:

ldh = [ fy times psi e times psi r times psi o times psi c, divided by (55 times lambda times root f prime c) ] times db to the power 1.5

with a minimum of the greater of 8db and six inches.

The exponent is the headline. Hook development now scales with the bar diameter to the power one and a half rather than linearly, so bigger bars are penalised progressively. For a #11 bar the exponent alone adds nineteen percent over a linear scaling.

The factors are:

The 2014 equation was ldh equals 0.02 psi e psi c psi r fy over lambda root f prime c, times db to the first power, with the same minimums. It looks similar and it behaves completely differently, because the factors were redefined rather than retuned.

In the 2014 edition, psi c was a cover factor worth 0.7 for well covered bars and psi r was a reduction worth 0.8 for confined hooks. Both were discounts. In the 2019 edition psi c became a concrete strength factor, psi r became a penalty of 1.6 rather than a discount of 0.8, and psi o was added as a further penalty of 1.25.

The net effect depends entirely on which discounts you were claiming before, and this is where most summaries of the change go wrong by quoting a single percentage.

Take a #8 Grade 60 bar in 6000 psi concrete, uncoated, normal weight. Under the 2014 rules with both discounts claimed, ldh comes out at 8.7 inches. Under the 2019 rules with no penalties, it is 14.1 inches. That is the 62 percent increase that gets quoted, and it is real.

But the same bar under the 2014 rules with no discounts claimed gives 15.5 inches, so against that baseline the 2019 hook is nine percent shorter. And a #4 Grade 100 bar in 3000 psi concrete comes out shorter under 2019 than under a fully discounted 2014 calculation.

So the honest statement is not that hooks got longer. It is that the 2019 edition removed the discounts and added penalties, and whether your particular hook grew or shrank depends on which factors you were entitled to claim. The tool computes both editions on every calculation, shows all three baselines, and explains which direction it moved and why.

Where the change unambiguously bites is the psi r penalty. A hook without adequate confining steel picks up a factor of 1.6, which is a sixty percent increase for nothing but detailing. The tool flags this and tells you what the length would be with confinement added, because adding ties is almost always cheaper than lengthening the hook.

And then the 2025 edition changed it again. Published in 2025, it moves the hooked bar provisions back towards the pre 2019 form, with adjusted modification factors including a new size effect factor, undoing much of the 2019 conservatism. I could not find the actual constants, so the tool computes the 2019 edition and says plainly that a project on the 2025 edition should not use its numbers.

I will note that the frequently repeated publication date of 31 March 2025 for ACI 318 in its 2025 edition is not the code's publication date. That is the date of a magazine article about it. The code was announced to subscribers in January 2025 and the printed version in July.

Bend geometry, and what a hook costs you in bar

Development length tells you how much embedment a hook provides. It says nothing about how much bar the hook consumes, which is what a bar bending schedule actually needs.

The code gives you two numbers per hook: a minimum inside bend diameter and a straight extension beyond the bend.

For hooks in longitudinal bars:

For hooks in stirrups, ties and hoops the bend diameters are tighter, because the bars are smaller and the bends are less critical:

That three inch floor on the 135 degree tail is worth flagging, because it is a change many charts missed. It arrived in the 2014 edition, specifically to eliminate the difference between a standard hook tail and a seismic hook tail. Before that, standard hooks got a bare 6db and only seismic hooks carried the three inch minimum. If a chart shows a 135 degree stirrup tail as 6db with no floor, it predates 2014.

For a #3 stirrup, 6db is 2.25 inches and the floor takes it to 3.0. That is a third more tail on the most common stirrup size in the industry.

I have to be straight about the stirrup table. The #3 through #5 rows are solid and multiply corroborated. The #6 through #8 rows are the weakest numbers on this page. The 6db bend diameter for those sizes rests on a single source in my checking, published summaries disagree about the 135 degree extension for the larger stirrup sizes, and I could not settle whether 180 degree hooks appear in the stirrup table at all. The tool warns about this in place and I would not detail a #7 stirrup hook from this page without opening the code.

Given the bend diameter and the extension, the bar consumed by the hook is pure geometry, and this is a calculation the tool does that the code does not.

The bar centreline sits half a bar diameter inside the outside of the bend, so the centreline radius is the inside bend diameter plus one bar diameter, all halved. The bend then sweeps through its angle at that radius, and the tail runs straight from the tangent point.

Bar consumed = the bend angle in radians times the centreline radius, plus the straight extension.

For a #5 bar with a 90 degree longitudinal hook: inside bend diameter 6 times 0.625 is 3.75 inches, so the centreline radius is (3.75 plus 0.625) over 2, which is 2.188 inches. Ninety degrees is pi over two radians, so the arc is 3.436 inches. The extension is 12db, which is 7.5 inches. Total bar beyond the tangent point: 10.94 inches, or 278 millimetres.

A word of warning on comparing that to a fabricator's table. Fabricators publish out to out detailing dimensions, usually labelled A or G, which measure the overall envelope of the hook rather than the length of bar in it. Those are different measurements answering different questions. The number above is what to add to a cutting length. The fabricator's number is what to check against a clearance.

A worked example

Take a real situation. A 200 millimetre thick suspended slab, Canadian standards, needing 600 square millimetres per metre of bottom steel in the short direction, cast in normal weight 30 MPa concrete with Grade 400 bar.

Step one, the area check. Shrinkage and temperature steel at 0.0018 of the gross area, for a 200 millimetre slab, is 0.0018 times 1000 times 200, which is 360 square millimetres per metre. The structural requirement of 600 governs, so 600 it is.

Step two, the bar and spacing. A 15M bar is 200 square millimetres. At 300 millimetre spacing it gives 200 times 1000 over 300, which is 667 square millimetres per metre. That clears 600 with eleven percent to spare. At 350 it gives 571, which does not clear it. So 300 millimetres is the widest spacing that works.

Step three, the spacing limit. The maximum is the lesser of five times the thickness and eighteen inches. Five times 200 millimetres is 1000. Eighteen inches is 457. So 457 governs, and 300 is comfortably inside it.

Try the same thing with 10M bar and it does not work as nicely. A 10M is 100 square millimetres, so it needs 167 millimetre spacing to reach 600, which rounds down to 150 in practice. That is 6.7 bars per metre against 3.3 for the 15M, so twice the bars, twice the tying, and the same steel. The 15M at 300 is the better answer, which is the general rule: use the largest bar that satisfies the spacing limits.

Step four, the weight. Say the slab is 8 metres by 12 metres and needs the same steel both ways for simplicity. In the 8 metre direction, bars run 12 metres long at 300 spacing across 8 metres, so 27 bars. In the 12 metre direction, bars run 8 metres long at 300 across 12 metres, so 41 bars. That is 27 times 12 plus 41 times 8, which is 324 plus 328, or 652 metres of 15M.

At 1.570 kilograms per metre that is 1024 kilograms, just over a tonne. Add ten percent for laps and waste and order 1130 kilograms.

Now the tolerance point from earlier. Under the standards that use a two sided band, the delivered mass could reasonably land anywhere within a few percent of that. Under the American one sided rule it will sit at or below nominal. Either way, a nominal takeoff is a purchasing figure and not a prediction.

The takeoff tab of the tool does this arithmetic from a bar list and computes the total two ways, once from the pounds per foot column and once from the kilograms per metre column, so that any inconsistency between the two published columns shows up rather than hiding. That cross check only means something for American bars, where the standard publishes both columns independently. For every other standard the pound figures on this page are conversions of the metric ones, so checking one against the other would only measure the conversion. The tool says so rather than printing a meaningless agreement.

Cover, spacing and the rest of the detailing envelope

Bar size interacts with three other constraints that decide whether a detail is buildable, and all three key off the bar diameter.

Minimum cover for cast in place non prestressed members:

The weather exposed rows apply to all member types, split only by bar size. Several online charts nest those rows under slabs and walls, which is wrong and produces an under covered beam.

The table number moved in the 2019 edition, from 20.6.1.3.1 to 20.5.1.3.1, so a cross reference to the old number tells you the source predates 2019. The 2019 edition also added a separate cover table for deep foundation members.

Minimum clear spacing between parallel bars in a horizontal layer is the greatest of one inch, one bar diameter, and four thirds of the maximum aggregate size. For longitudinal bars in columns, pedestals, struts and boundary elements it is the greatest of one and a half inches, one and a half bar diameters, and four thirds of the aggregate.

The word is greatest. At least one widely used university lecture note has this as least, which would produce a one inch clear spacing for a #11 bar and a placement that no vibrator could reach.

The aggregate term is the one people forget. With three quarter inch aggregate, four thirds gives one inch and the aggregate rule does not govern. With one and a half inch aggregate it gives two inches, and suddenly it governs everything up to a #16 bar, if such a thing existed.

Bundled bars are limited to four in a bundle. Bars larger than #11 are not bundled in beams. Development length increases by twenty percent for a three bar bundle and thirty three percent for a four bar bundle, and individual bars terminating within a span must be staggered by at least 40db. For spacing and cover purposes the bundle is treated as a single bar of equivalent total area, and the cover need not exceed two inches except where cast against ground.

Minimum wall reinforcement depends on bar size and grade in a way that rewards small bars:

Using #5 bar at Grade 60 or better buys a twenty percent reduction in the minimum against using #6 bar, in both directions. That is a real saving and it points the same way as the slab logic points the other way: in walls, smaller is cheaper.

Where the in plane shear demand exceeds half the design shear capacity, both ratios go to 0.0025 regardless, and special structural walls require 0.0025 in both directions unconditionally.

The values in that table are corroborated three ways, but I never got the imperial table itself in front of me, so the row ordering and column headings are my reconstruction rather than a transcription. The numbers I trust. The layout is my arrangement of them.

Crack control spacing for the reinforcement closest to the tension face is the lesser of 15 times (40,000 over fs) minus 2.5 cc, and 12 times (40,000 over fs). Here fs may be taken as two thirds of fy, and cc is the least distance from the surface of the bar to the tension face, which is a clear cover to the nearest bar rather than a cover to the centroid.

What I could not verify, and left out

Everything on this page that I could not source to my own satisfaction has been left off it rather than guessed at, and it seems more useful to list the gaps than to pretend there are none.

From the American side: the edition that first introduced the #20 bar, the letter C as an alternative Grade 100 marking, the metric grade number marks of 4, 6 and 7, the line counts for Grades 100 and 120 where ASTM and CRSI directly contradict each other, whether Grades 80 and 100 are furnished in every size, the elongation requirements, and the A706 mechanical property table. The A706 dimensional table is confirmed identical to the A615 one. Its strength table is not.

From the international side: whether CSA G30.18 prints a mass column in the form given here and what density it states, the exact IS 1786 area column against a paper copy, the meaning of the IS 1786 S suffix, the entire AS/NZS 4671 2019 mechanical property table, whether the Australian dimension table was extended after 2001, the Australian, Canadian, Japanese and Chinese mass tolerances, the Japanese D6 bar, the higher JIS grades, and whether EN 10080 lists any diameters beyond the fourteen given.

From the code side: every numerical detail of the 2025 edition, whether 180 degree hooks appear in the stirrup table, the bend diameter for #6 to #8 stirrups, the hooked bar epoxy factor where two sources conflict, the section number for the prohibition on bundling bars larger than #11 in beams, whether the Ktr floor triggers at or above 80,000 psi, whether interpolation of psi g between grades is a code allowance or a software convenience, and whether the shells and folded plates cover row survived the 2019 reorganisation.

That is a long list and I would rather it were short. It is long because reinforcing bar standards are mostly paywalled, and the freely available summaries of them contradict each other more often than you would hope. Where sources disagreed I have said so in place rather than picking one and moving on.

The numbers I have published are the ones I could get from a primary text or from several independent sources that agreed. The engines were checked against an independent implementation across five thousand one hundred and one cases and against the published worked examples I could find. That found four real defects in the tool before it went up, including one that printed a plausible looking number where it should have printed a comparison, so the checking was not ceremonial.

Common questions

What is the difference between a 10M rebar and a number 10 rebar?

They are three different bars. A Canadian 10M carries 100 square millimetres of steel. An American number 10 bar in the metric designation carries 71 square millimetres, because that designation refers to a number 3 bar. An American number 10 bar in the inch and pound designation carries 819 square millimetres. The Canadian 10M has 41 percent more steel than the American metric number 10, which is close enough in size that a substitution would pass a visual inspection while leaving the member badly under reinforced.

Why does a number 14 rebar have a diameter of 1.693 inches?

Because it is not named for its diameter. Bars from number 3 to number 8 are named for their diameter in eighths of an inch, but number 9 and above replaced square bars and inherited the square bar's cross sectional area instead. A number 14 replaced a one and a half inch square bar, which has an area of 2.25 square inches, and the round bar with that area is 1.693 inches across. That works out to 13.54 eighths of an inch, not 14.

Is the area of a number 16 metric rebar 199 or 200 square millimetres?

ASTM prints 199. Charts that print 200 have recomputed the value rather than quoting the table. The metric area column is normally the published inch area converted at 645.16 square millimetres per square inch, which for the 0.31 square inch number 5 bar gives exactly 200.00. The number 16 row is one of only two rows in the whole table where the standard does not follow that rule, and it instead matches the area computed from the rounded metric diameter of 15.9 millimetres, which is 198.56 and rounds to 199.

What is the minimum tensile strength of ASTM A615 Grade 60 rebar?

80,000 psi under the 2020 edition and later. It was 90,000 psi before that, and most published charts still print the old figure. The same edition cut Grade 80 from 105,000 to 100,000 psi and added a new requirement that the ratio of actual tensile strength to actual yield strength be at least 1.10, measured on the test specimen. That ratio requirement is what makes the lower absolute minimums safe, because a bar that over yields must now over fracture to match.

Is there a rebar size larger than a number 18?

Yes. Current editions of ASTM A615 include a number 20 bar, designated 64 in metric, 2.500 inches across with 4.91 square inches of section and a weight of 16.69 pounds per foot. Almost no published chart includes it. The standard attaches an unusual warning to that row, noting that design codes may not recognise the size and that a building official may need to approve its use, so it is not an ordinary design option in the way that a number 18 is.

How much steel does a number 5 bar at 12 inch spacing provide?

0.31 square inches per foot, which is the bar area itself. The area per unit width is the bar area multiplied by the width and divided by the spacing, so at a 12 inch spacing with a one foot width the arithmetic cancels and the answer is just the bar area. Halving the spacing doubles the steel, so a number 5 bar at 6 inches gives 0.62 square inches per foot.

Did ACI 318 change the hook development length?

Twice in six years. The 2019 edition replaced the linear form with one that scales as the bar diameter to the power 1.5, changed the constant from 50 to 55, redefined two modification factors from discounts into penalties and added a third penalty. The 2025 edition then moved the provisions back towards the pre 2019 form with a new size effect factor. Whether a particular hook grew or shrank in 2019 depends entirely on which of the older discounts you were entitled to claim, so a single percentage figure for the change is misleading.

What is the minimum shrinkage and temperature reinforcement ratio?

0.0018 under the 2019 edition of ACI 318, flat, with no grade dependence. Earlier editions used 0.0020 for steel below 60,000 psi and permitted a reduction for higher strength steel down to a floor of 0.0014. The 2019 edition removed the grade dependence entirely, on the reasoning that shrinkage cracking does not care how strong the steel is. A spreadsheet still carrying the old reducing formula will under reinforce a high strength slab by up to 29 percent.

Related Calculators and Reference

• Rebar Weight and Spacing Calculator for slab and wall layouts with bar counts and total weight
• Development Length and Lap Splice Calculator for the full ACI and Eurocode anchorage treatment
• Rebar Cutting List Calculator and Bar Bending Schedule for turning a detail into a cutting list
• Quantity Takeoff Calculator for concrete, rebar and formwork together
• Slab Design Calculator for one way and two way slabs
• Shear and Stirrup Spacing Calculator for beam transverse reinforcement
• RC Beam Design Calculator for singly and doubly reinforced sections
• Grade 60 Rebar Yield Strength and the Fe500, Fe415 and Fe550 grade pages
• Individual size pages for #3, #4, #5, #6, #7, #8, #9, #10 and #11 bar
• Canadian metric size pages for 10M, 15M, 20M, 25M and 30M bar
• Unit Weight of Steel and Modulus of Elasticity of Steel for the underlying material properties
• Length Converter for working between inch and millimetre bar schedules
• CRSI bar identification guide, the industry reference for reading the marks rolled into a bar