GFRP modulus, the CE exposure factor, what a bend costs, and why this page prints no lap length.
← Back to the calculatorSteel ruptures at 6 to 12% strain and carries a yield row; GFRP has none, so the row reads n/a. ACI 440 then cuts the strength you are allowed to use to 0.7 of the guaranteed value wherever the concrete is exposed to earth and weather, and a bend costs another 40 to 50%.
The lap is not on this page. Development and lap splice length for GFRP is governed by ACI 440.1R-15 Chapter 8, and that length moves with the bar’s surface treatment and with the manufacturer’s guaranteed tensile strength, so there is no single length to print. This page gives no number for it. What it does give is the three numbers that decide whether the bar belongs on the drawing at all — the exposure factor, the bend, and the absence of a yield point.
All three come from one table. ACI 440.1R-15 prints steel and GFRP side by side; the copy this page works from is Table 4.1 of FHWA/TX-20/0-6952-1, which reproduces it. Nothing on this page is a single value taken from a supplier, because the material does not have one.
Five rows decide this product. ACI 440.1R-15 prints them for steel and GFRP in the same table, which is the only fair way to read them:
| Property | Steel (A615) | GFRP |
|---|---|---|
| Nominal yield stress | 40 to 75 ksi | n/a |
| Tensile strength | 70 to 100 ksi | 70 to 230 ksi |
| Elastic modulus | 29.0 ×10³ ksi | 5.1 to 7.4 ×10³ ksi |
| Yield strain | 0.14 to 0.25% | n/a |
| Rupture strain | 6.0 to 12.0% | 1.2 to 3.1% |
Three rows are the whole argument. The modulus row makes the bar 3.9 to 5.7 times softer than steel — 29.0 ÷ 7.4 = 3.9 at the stiff end, 29.0 ÷ 5.1 = 5.7 at the soft end. The yield row says n/a, which is not a missing value: the bar is linear elastic to rupture and there is no stress at which it stops being elastic. The rupture strain row puts the end of the line at 1.2 to 3.1%, where steel runs to 6 to 12%.
The ranges are not independent, and this is where most pages go wrong. Divide the tensile strength by the modulus and the answer is the rupture strain: 70 ÷ 5,100 = 1.37%, and 230 ÷ 7,400 = 3.11%. The low-modulus bar is the low-strength bar and the high-modulus bar is the high-strength one. A 200 ksi bar at 5,100 ksi does not exist, because it would rupture at 3.9% and it would not be glass. So “70 to 230 ksi” is not a range you pick a middle from — it is two different products, and a page that quotes 150 ksi as the strength of fiberglass rebar is quoting a bar nobody sells.
Design does not start from either end of the range. It starts from the guaranteed tensile strength in the manufacturer’s data for the bar on the truck, which is why that number has to travel with the submittal rather than live on a web page. ASTM D7957/D7957M-17 is the product standard for solid round GFRP bar; ACI 440 then applies its environmental factor to the guaranteed value, and what comes out is the design strength. Every number in the next section is that one multiplication.
CE is a table of six numbers and glass gets the worst of it:
| Fiber | Not exposed to earth and weather | Exposed to earth and weather |
|---|---|---|
| Carbon | 1.0 | 0.9 |
| Glass | 0.8 | 0.7 |
| Aramid | 0.9 | 0.8 |
The factor is not a safety factor and it is not a durability allowance stacked on top of one. It is a conversion from the strength the bar has in a laboratory to the strength it keeps in the environment it is cast into: 0.8 where the concrete is not exposed to earth and weather, 0.7 where it is. The same bar in a footing and in an interior slab therefore has two design strengths, and the difference between them is 12.5% of the number you started with.
Glass is the only fiber that takes the 30% cut, and the reason is chemistry rather than marketing. The alkali in cement paste attacks the glass fibers directly, and hydration products nucleate and grow in the spaces between the individual filaments, so the bundle loses strength from the inside where nothing can inspect it. Carbon takes 1.0 and 0.9, aramid 0.9 and 0.8, glass 0.8 and 0.7. That is why the same page that sells GFRP on corrosion also has to hand back 30% of its strength before the design starts.
| Guaranteed | Design, CE 0.8 | Design, CE 0.7 | Against Grade 60 |
|---|---|---|---|
| 70 ksi | 56 ksi | 49 ksi | 0.82× |
| 100 ksi | 80 ksi | 70 ksi | 1.17× |
| 230 ksi | 184 ksi | 161 ksi | 2.68× |
Read both ends of that table. A 70 ksi guaranteed bar in ground contact designs to 49 ksi, which is 18% below the 60 ksi of a Grade 60 steel bar. The 230 ksi bar designs to 161 ksi, 2.7 times it. Both are sold as fiberglass rebar, and even after the factor the spread between them is still 3.3 to 1. Any page that answers “how strong is fiberglass rebar” with one number has answered for one bar and sold you another.
One more thing is priced in the same direction and usually left out: ultraviolet light and moisture before the bar is cast reduce its tensile strength, because the polymer matrix degrades. A pallet that sat in the sun for a season is not the bar that was certified, and that is a storage instruction, not a design factor.
Steel has a yield plateau, and the plateau is the reason steel structures give warning. The table puts yield at 0.14 to 0.25% strain and rupture at 6 to 12%, so a steel bar runs 24 to 86 times more strain at constant stress after it yields. That reserve is what lets a steel section redistribute moment, deflect visibly and still hold load while somebody gets out of the way.
GFRP has none of it. The yield row is n/a, the curve is a straight line from zero to rupture, and failure is the end of that line. Three consequences follow, and none of them is a safety factor away:
The second thing glass does not have is patience. A GFRP bar held under constant tension can fail suddenly after a period the literature calls the endurance time — creep rupture, or static fatigue — and of the three fibers ACI 440.1R-15 compares, glass is the most susceptible, carbon the least. Steel has no equivalent at ambient temperature. It is a sustained-load check that does not appear anywhere in the strength table, so a page that only quotes tensile strength has left out the failure mode that arrives without a load change.
Every bend in a GFRP bar is a weaker section, and ASTM D7957/D7957M-17 — the product standard — lists tensile properties for the straight bar and does not carry a separate value for the bent portion. The bent portion carries 40 to 50% less strength than the straight portion, because the fibers are turned through the radius and the load path on the inside of the bend is short. It is a property of the product, not of the workmanship.
Run it through a 100 ksi bar. Straight, it is 100 ksi guaranteed and 70 ksi design at CE 0.7, the factor for concrete exposed to earth and weather. Through a bend, the same bar is 50 to 60 ksi guaranteed and 35 to 42 ksi design at the same factor — a number that straddles the 40 ksi of a Grade 40 steel bar. A hooked anchor, a stirrup corner and a 90° dowel are all that number, and they are all at the ends of members where the moment is highest.
The second half of the rule is the one that bites on site: a GFRP bar cannot be bent after it is formed. The bar is pultruded straight and the bends are made before the resin cures, so bending is a factory operation and a field bend is a scrapped bar, not a repaired one. Order bent bars to shape and to the drawing, and do not plan on adjusting one to fit.
Development and lap splice length for GFRP is governed by ACI 440.1R-15 Chapter 8, and this page gives no number for it. The reason is in the product rather than in the page: the length moves with the bar’s surface treatment and with the manufacturer’s guaranteed tensile strength, so there is no single length to print. A sand-coated bar, a helically wrapped bar and a ribbed bar are three different bond problems, and the Chapter 8 equations were fitted to an experimental database that is mostly GFRP but is not one product.
What the literature does say is directional, and it all points the same way:
Four things have to come from the supplier before the bar is detailed: the guaranteed tensile strength, the surface treatment, the modulus for deflection, and the manufacturer’s own development and lap data. The first two set the lap, the third sets the section, and none of the four is on a bar schedule by default.
ACI 440.1R-15 puts GFRP at 70 to 230 ksi, a three-fold range across products, so a single number is wrong for the material. Design does not use either end: it uses the guaranteed tensile strength in the manufacturer's data for the bar being supplied, multiplied by the environmental factor CE — 0.8 where the concrete is not exposed to earth and weather, 0.7 where it is.
5,100 to 7,400 ksi, against 29,000 ksi for steel, so a GFRP bar is 3.9 to 5.7 times softer than a steel bar of the same size. That ratio, not the strength, is what usually decides the section, because it sets how much the member deflects under service load.
No. ACI 440.1R-15 prints n/a in the yield row for GFRP: the bar is linear elastic to rupture. Steel yields at 0.14 to 0.25% strain and then runs to 6 to 12% at constant stress, which is 24 to 86 times more strain after yield; GFRP breaks at 1.2 to 3.1% strain with no plateau before it.
Because the alkali in cement paste attacks glass fibers, and hydration products nucleate and grow between the individual filaments, so a bundle loses strength from the inside over time. The 0.7 applies where the concrete is exposed to earth and weather; 0.8 applies where it is not. Carbon takes 1.0 and 0.9, aramid 0.9 and 0.8, so glass is the one fiber with a 30% cut.
The bent portion of a GFRP bar carries 40 to 50% less strength than the straight portion, because the fibers are turned through the radius. A 100 ksi bar is 50 to 60 ksi in the bend, and 35 to 42 ksi in the bend after CE 0.7. GFRP also cannot be bent after it is formed: bends are made before the resin cures, so a field bend scraps the bar.
Development and lap splice length for GFRP is governed by ACI 440.1R-15 Chapter 8. That length moves with the bar's surface treatment and with the manufacturer's guaranteed tensile strength, so this page gives no number for it. Get the manufacturer's development and lap data with the submittal.
On tension capacity sometimes, on stiffness no. A #5 GFRP bar at 70 ksi design strength carries 21.7 kip against 18.6 kip for a Grade 60 #5, but it is 3.9 to 5.7 times softer, so the member deflects far more at the same load and deflection usually governs the design.
It does not rust, and chlorides do not attack it the way they attack steel, which is the reason it exists. It degrades differently: alkali and moisture reduce the tensile strength and stiffness of the glass over time, and that degradation is exactly what the CE factor prices in. Ultraviolet light and moisture before the bar is cast also cost strength, so a pallet left in the sun is not the same bar as one kept covered.