At a glance
- Design with the characteristic value, never the mean: characteristic = mean − 3σ from at least 20 coupons (ACI 440.2R §4.3), a 99.87%-confidence floor.
- Raw MPa across products can mislead: fabric is quoted on net-fiber area, plate on the full cured laminate. Multiply strength × thickness — the force per unit width is what’s comparable.
- A dry-fiber number like 5,800 MPa is not for design — use the mechanical-property table, then apply CE and the debonding cap.
A CFRP data sheet is full of large numbers, and using the wrong one quietly wrecks a design. This guide teaches you to read one correctly — mean versus characteristic, which cross-section a strength is measured on, which test method backs it, and how the headline number becomes a usable design value. Examples use FidStrong’s real FSC fabric and FSL plate data. Start, if needed, with what CFRP strengthening is.
Mean or characteristic — which number do you design with?
Always the characteristic value. ACI 440.2R §4.3 defines it as the mean minus three standard deviations, computed from at least 20 test specimens (FidStrong tests 25 per property), giving 99.87% confidence the true strength exceeds it. The mean is just the raw coupon average and overstates what you can rely on. One exception: the modulus Ef is reported and used as the mean — it is not reduced — because stiffness scatters far less than strength. So on any row, expect a mean/characteristic pair for strength and elongation, and a single mean modulus.
Why do a fabric’s and a plate’s MPa look so different?
Because they are measured on different cross-sections. Fabric strength is reported on the net fiber area (fiber only, very thin); plate strength is reported on the gross cured laminate (fiber plus resin, thicker). The same FSC200A ply reads 4,500 MPa on net-fiber area (0.111 mm) or 1,248 MPa on the cured laminate (0.40 mm) — both true. The way to compare fairly is to multiply strength by thickness: 4,500 × 0.111 and 1,248 × 0.40 both give 499 kN/m, the force per unit width that actually governs design. See plate vs. fabric for the full comparison.
Can you design with the 5,800 MPa number?
No. Some sheets quote a raw dry-fiber tow strength — 5,800 MPa on a FidStrong plate sheet, for instance — explicitly marked “reference only, not for design.” It is the strength of loose fiber before it is made into a composite, it is not statistically reduced, and the cured laminate never reaches it. Design must use the cured-laminate (or net-fiber) mean and characteristic values from the mechanical-property table — for that plate, 2,700–3,400 MPa, not 5,800.
Which test method should back the numbers?
For fabric and plate, ASTM D3039, with the chord modulus taken over the 1,000–3,000 microstrain window. Watch for a trap: ASTM D7205 is a different method, for pultruded FRP bars, using steel-tube-and-grout anchors instead of bonded tabs — its numbers are not interchangeable with D3039 fabric or plate data even though both are “CFRP.” A credible sheet names the test method next to each property; if it doesn’t, ask.
How do you sanity-check a data sheet?
Two quick checks. First, for linear-elastic CFRP, elongation should roughly equal strength divided by modulus: FSC200A’s 4,500 MPa ÷ 240 GPa = 1.875%, matching its reported 1.8% mean. If the two don’t reconcile, the sheet may be mixing bases — fiber-area strength with laminate-area modulus. Second, nominal thickness tracks areal weight, not strength grade: every 200 g/m² grade shares 0.111 mm; 300 g/m² is 0.167 mm. A sheet where thickness changes with the strength tier is a red flag. The tow grade, incidentally, is not on a proper sheet at all — products are differentiated by measured performance, not a claimed T-number; more in carbon fiber raw materials.
From data sheet to design value
The headline is only the first rung. The design value ladders down: mean → characteristic (mean − 3σ) → multiply by the environmental factor CE (0.95 indoor, 0.85 outdoor for carbon, per ACI 440.2R Table 9.4) → and finally the debonding-strain cap, which depends on the specific joint (concrete strength, number of plies, stiffness, thickness) and usually governs before the material limit is reached. That is why there is no single “final MPa” on a sheet — the sheet gives you the top of the ladder, and the design computes the rest. Interpreting these values correctly is also how a specifier reads the epoxy adhesive properties that go with them.
FAQ
What’s the difference between mean and characteristic strength?
Mean is the raw average of test coupons. Characteristic (ffu*) is mean minus three standard deviations from at least 20 specimens per ACI 440.2R §4.3 — a 99.87%-confidence floor. Design always uses the characteristic value, never the mean, because it accounts for manufacturing variability between coupons.
Why do a fabric’s and a plate’s MPa look so different when both are carbon?
They’re reported on different cross-sections. Fabric strength is net-fiber-area (fiber only, thin); plate strength is gross cured-laminate (fiber plus resin, thicker). Multiply strength by thickness on either basis and the force per unit width converges — that force is the number that actually governs design capacity.
My sheet lists 5,800 MPa fiber strength — can I design with it?
No. That’s a raw dry-fiber reference figure, unachievable in the cured composite and not statistically reduced per ACI 440.2R §4.3. Design must use the cured-laminate or net-fiber mean and characteristic values from the product’s mechanical-property table.
Should ultimate elongation equal strength divided by modulus?
Yes, for linear-elastic CFRP — elongation ≈ strength ÷ modulus, at either the mean or characteristic level. If the two don’t reconcile, the sheet may be mixing bases (fiber-area strength with laminate-area modulus); ask the manufacturer which basis each column uses.
Which ASTM method should back the tensile numbers?
ASTM D3039, with chord modulus over 1,000–3,000 microstrain. ASTM D7205 is a separate method for pultruded FRP rebar, using steel-tube anchors instead of tabs — its results aren’t interchangeable with D3039 fabric or plate data, even though both are “CFRP.”
The characteristic strength is already reduced — why apply CE on top?
The characteristic value only accounts for manufacturing variability between coupons. ACI 440.2R Table 9.4’s CE factor (0.95 indoor, 0.85 outdoor for carbon) then accounts for the service environment — a separate reduction, applied afterward, before the debonding-strain cap potentially reduces usable capacity further.
FidStrong publishes FSC carbon fabric and FSL plate data sheets with mean and characteristic values on a stated basis, tested to ASTM D3039 (25 specimens per property), under ISO 9001, 14001, and 45001 quality systems. Interpretations here follow ACI 440.2R; always design to the current data sheet and governing code.