7 to 12 mils of film, 2% and 5% of damage, and 25 in of extra lap on a #8.
← Back to the calculatorThe film is 7 to 12 mils (A775), the damage limits are 2% and 5% per foot (D3963 7.3), and the factor is ψe 1.5, which drops to 1.2 only when cover and spacing both clear the bar.
The 25 in is the ordinary case, not the worst one: 4000 psi, 2 in of clear cover, 6 in of clear spacing, bottom bars. Everything above is a number you can put on a drawing or check with a gauge. The pages that rank for this query do not have any of them: they explain what fusion-bonded epoxy is, list the standards, and hand the numbers to a certificate.
The three questions this page answers are the ones that come up at the bar — how thick is the film and how much of it has to be in spec, how much damage is too much, and how many inches the coating adds to the lap. The first two are ASTM. The third is ACI, and it is the one that changes the bar list.
ASTM A775 asks for a dry film thickness of 7 to 12 mils. In the units a coating inspector actually writes down that is 0.007 to 0.012 in, or 178 to 305 micrometres, and one mil is 25.4 micrometres.
The clause is worth reading twice, because it is not a tolerance band on every reading. A775 states that, for acceptance, at least 90% of all recorded thickness measurements of the cured coating shall be 7 to 12 mils. So the test is a distribution: ten readings with one at 5 mils can still pass, and ten readings all at 13 mils fail even though every single one of them is thicker than the minimum. That sentence is the difference between a coating that gets rejected and a coating that gets a second look, and it is the sentence no product page quotes.
Two practical consequences follow. First, thickness is measured with a magnetic gauge on the flat between deformations, and the ends of the bar read high because the spray builds up there — which is why a bar sampled at its ends looks better than it is. Second, the film changes the geometry: 12 mils on a #8 adds 0.024 in to the diameter, so two #8 bars detailed at 6 in clear become 5.976 in apart, and a 2 in cover becomes 1.988 in. Neither number is enough to fail a detail, and both are the kind of thing that turns up when someone measures the cage instead of the drawing.
ASTM D3963 7.3 is the clause that gets quoted and almost never used, because it is written in percentages of surface area and nobody at a bar knows what 2% of a foot is. Two limits, both per 1 ft section of coated bar:
Converted to square inches, on a bar whose surface in one foot is π × db × 12:
| Bar | Bar surface in 1 ft | 2% — damage | 5% — patched | ¼ in spots to reach 2% |
|---|---|---|---|---|
| #3 | 14.14 | 0.28 | 0.71 | 4.5 |
| #4 | 18.85 | 0.38 | 0.94 | 6.0 |
| #5 | 23.56 | 0.47 | 1.18 | 7.5 |
| #6 | 28.27 | 0.57 | 1.41 | 9.0 |
| #7 | 32.99 | 0.66 | 1.65 | 10.6 |
| #8 | 37.70 | 0.75 | 1.88 | 12.1 |
| #9 | 42.52 | 0.85 | 2.13 | 13.6 |
| #10 | 47.88 | 0.96 | 2.39 | 15.3 |
| #11 | 53.16 | 1.06 | 2.66 | 17.0 |
Read the last column. On a #4, six quarter-inch nicks in one foot reach the 2% line, and fifteen reach 5%. On a #3 it is four and a half spots; on a #8 it is twelve. A drag mark from a chain, a shear burr, a weld spatter burn — each one is a fraction of a spot, and they add up along the bar faster than anyone expects, because the bar is only 18.85 in² of surface per foot at #4.
Two more things about the repair itself. The patch material has to be a two-part epoxy meeting A775, A934 or D3963 — a spray-can product cannot meet any of them, whatever the label says. And the cut ends that are excluded from the percentage counts are still required to be patched, so a bar with two cut ends has two patches on it that never appear in the 2% arithmetic.
ACI 318-19 Table 25.4.2.5 gives epoxy-coated bar a coating factor ψe that multiplies development length. It has exactly two values, and the choice between them has nothing to do with how thick the film is:
For a #8 that means 3 in of cover and 6 in of clear spacing. For a #11 it means 4.2 in and 8.5 in — which is why the big bars almost never earn the 1.2, and why the 25 in in the headline is the ordinary case rather than the worst one. The cover half of that test has its own table on concrete cover, and the spacing half is what the spacing page counts. Because the lap is 1.3 ld, ψe enters through ld rather than through the lap: the coated lap is 1.3 times the coated ld, rounded up at each step.
| Bar | Class B, uncoated | ψe 1.2 | Extra | ψe 1.5 | Extra |
|---|---|---|---|---|---|
| #3 | 16 in | 19 in | +3 | 24 in | +8 |
| #4 | 21 in | 25 in | +4 | 30 in | +9 |
| #5 | 25 in | 30 in | +5 | 38 in | +13 |
| #6 | 30 in | 37 in | +7 | 46 in | +16 |
| #7 | 45 in | 52 in | +7 | 65 in | +20 |
| #8 | 50 in | 60 in | +10 | 75 in | +25 |
| #9 | 56 in | 68 in | +12 | 85 in | +29 |
| #10 | 64 in | 76 in | +12 | 95 in | +31 |
| #11 | 71 in | 85 in | +14 | 106 in | +35 |
So a #8 lap at 4000 psi is 50 in black, 60 in at 1.2 and 75 in at 1.5. The dollars follow from the pounds: 25 in of #8 is 2.08 ft × 2.670 lb/ft = 5.56 lb, which is $2.78 at $1,000 a ton. A #11 is worse in every column — 71 in becomes 106 in, 35 in more, 15.50 lb, $7.75 — and it is the bar where the lap starts to govern the cut list.
Multiply it out on a frame and the number stops being small. Eight columns, four #8 bars each, spliced at three elevations:
laps = 8 × 4 × 3 = 96 extra per lap = 25 in = 2.08 ft extra bar = 96 × 2.08 ft × 2.670 lb/ft = 534 lb cost = 534 lb × $0.50/lb = $267 extra lap length = 96 × 25 in = 2,400 in = 200 ft
One cap matters here and it is easy to miss. The product ψt × ψe may not exceed 1.7. A top bar takes ψt = 1.3, so an epoxy-coated top bar would compute to 1.95 and is capped at 1.7: its lap is 50 × 1.7 = 85 in, not 97. Without the cap the same detail gets 12 in of bar that nobody needed, and the bar list grows quietly.
The most-read page on this query says epoxy-coated bar needs “15% more bar” for development. The two factors in the code are 1.2 and 1.5, which are 20% and 50% — neither of them is 15%, and the page does not show the arithmetic behind its number. It is being used as an argument for galvanized bar, which is a fair argument: galvanized takes ψe = 1.0, so it costs no extra lap at all.
The honest version of that comparison is 50 in against 60 or 75 in, or $2.78 a lap at $1,000 a ton. That is the number that belongs in a decision, and it is smaller than the argument implies — the reason to use coated bar is chloride exposure and a 50-year deck, not the price of the steel it adds.
One more query belongs here because it is what people type: is epoxy coated rebar banned. Nothing in ASTM A775 or ACI 318-19 prohibits it — the code charges it a factor rather than forbidding it. What does exist is owner-level restriction on some bridge work, and it goes back to field studies from the 1990s that found corrosion starting at coating damage and at holidays rather than in the middle of an intact film. That is the same 2% and 5% problem this page is about, and it is why the repair limits matter more than the coating spec on its own.
There is a third option that costs no lap and no coating: stainless steel rebar. It is uncoated, so ACI 318-19 gives it ψe = 1.0, the same as black bar, and the lap on a #8 stays 50 in instead of 60 or 75. Bond is not the obstacle — 191 beam-end tests comparing A615, 316LN and 2205 duplex bar found no difference from carbon steel, which is the finding FHWA/TX-20/0-6952-1 cites before concluding that design codes treat stainless and carbon laps the same. The cost sits in the alloy, priced per pound of bar, rather than in inches of bar on the drawing.
And there is a route that takes the steel out of the question rather than coating it: GFRP bar, which does not rust, does not yield, and whose design strength is 0.7 times its guaranteed value wherever the concrete is exposed to earth and weather.
The film almost never fails in the middle of a bar. It fails where something touched it, and every one of those contacts is measured against the same number — 2% of the surface of a 1 ft section — so the field version of this section is a size, not a rule. A US dime is 0.705 in across, which covers 0.39 in². That is more than the whole 2% on a #4 (0.38 in²) and a little over half of it on a #8 (0.75 in²). One dime of bare steel on a #4 is the rejection line, not a warning.
| Bar | 2% of one foot | In dimes |
|---|---|---|
| #3 | 0.28 in² | 0.7 |
| #4 | 0.38 in² | 1.0 |
| #6 | 0.57 in² | 1.4 |
| #8 | 0.75 in² | 1.9 |
| #11 | 1.06 in² | 2.7 |
What spends it, in the order it happens on a job:
ASTM A775 accepts 7 to 12 mils, which is 0.007 to 0.012 in, or 178 to 305 micrometres. The clause is written as a distribution, not a single number: at least 90% of the readings on a bar have to land inside that band. So one thin spot on a coated bar is not automatically a rejection, and a gauge reading of 5 mils does not tell you anything until you know how many of the other readings looked like it.
ASTM D3963 7.3 sets two limits per 1 ft section of bar: damaged area before repair may not exceed 2%, and the area covered by patching material may not exceed 5%. Sheared and cut ends are excluded from both counts. The 2% is not a damage allowance — D3963 requires all visible damage to be repaired — it is the point past which the bar may be rejected, which is a different thing and the one that gets misread in the field.
No, it costs bond. ACI 318-19 Table 25.4.2.5 gives epoxy-coated bar a coating factor ψe of 1.2 or 1.5 that multiplies development length, so the same bar needs 20% or 50% more embedment. The coating is a corrosion barrier; the smooth, inert film is exactly why the code charges for it.
For a #8 at 4000 psi with 2 in of cover, the Class B lap goes from 50 in to 75 in — 25 in more, which is 5.56 lb of extra bar per lap, about $2.78 at $1,000 a ton. With 3 in of cover and 6 in of clear spacing the factor drops to 1.2 and the same lap becomes 60 in, 10 in longer and 2.23 lb heavier.
1.2 when clear cover is at least 3db and clear spacing is at least 6db; 1.5 in every other case. For a #8 that means 3 in of cover and 6 in of clear spacing, and for a #11 it means 4.2 in and 8.5 in — which is why the larger bars almost always get 1.5. The factor is a property of the geometry around the bar, not of the coating thickness.
Only if the specification allows it, and the coating on the outside of the bend will crack. Shop bending before coating is the normal route and it is what ASTM A934 covers: prefabricated bent bar, coated after bending. A field bend leaves bare steel and cracked film on the tension face, and both have to be patched under D3963 before the pour.
No. D3963 and the A775 appendix ask for a chop saw or power shears, and the cut end then has to be patched with a two-part epoxy material. A flame cut burns the coating back from the end and leaves steel that no patching material will cover properly. The cut ends are excluded from the 2% and 5% area limits, but they are not excluded from the repair requirement.
Thirty days, then it needs an opaque cover. The coating is degraded by UV and the damage shows up as chalking and then as a film that no longer bonds, so A775's appendix and D3963 both call for the bars to be stored off the ground, separated from uncoated bar, and covered if they will be outdoors more than 30 days. A tarp that traps condensation is worse than no tarp.
No, and the code treats them differently. Galvanized bar to ASTM A767 takes ψe of 1.0, the same as uncoated, while epoxy-coated bar takes 1.2 or 1.5. That is the whole basis of the “epoxy needs more bar” argument: on a #8 at 4000 psi it is 50 in against 60 or 75 in for the same lap.
A second, textured epoxy layer applied over the smooth coating to give the bar back some of the grip the fusion-bonded film took away. ASTM A1124/A1124M-23 is the specification, and the published tests claim laps up to 60% shorter than plain coated bar. ACI 318-19 was written before that standard existed, so unless the specification or an evaluation report says otherwise, the 1.2 or 1.5 factor is still what the drawing has to carry.
Not to the number you order by. A615 and A775 both weigh and price the bar on its nominal uncoated section, so a #8 is 2.670 lb/ft black or coated and the bar list does not move. The film is real but small: a #8 has 37.7 in² of surface per foot, 12 mils over that is 0.45 in³ of epoxy, and at 1.4 to 1.6 g/cm³ that is 0.023 to 0.026 lb — 0.9 to 1.0% of the bar. At the 7 mil minimum it is half that. What does change is the diameter an inspector measures: 0.024 in more on a #8.
Yes, by 0.024 in on a #8 at 12 mils, because the film is part of the outside surface. It never matters until a detail sits exactly on the 6db line that decides ψe. Two #8 bars at 6 in clear have 5.976 in of gap once both are coated, and 6db is 6.000 in, so the 1.2 bracket is gone and the lap goes from 60 in to 75 in — 15 in a lap. On the 96-lap frame above that is 120 ft, 320 lb, $160 at $0.50/lb, for a detail that was drawn correctly and built to the drawing. Detail 6.5 in clear and the question never comes up.