GFRP rebar does not rust and weighs about a quarter as much as steel (specific gravity 1.92 against 7.85). In a University of Kentucky program for FHWA, GFRP bars reached 612 MPa mean ultimate strength against 487 MPa yield and 653 MPa ultimate for steel, but their modulus was 40 GPa against 196 GPa, about one-fifth of the stiffness, with no yield plateau. It is not a one-for-one substitute: crack width and deflection usually govern, so bar size, spacing, cover and laps are recalculated to ACI CODE-440.11 or the governing code.
GFRP rebar is compelling when chloride corrosion, magnetic interference or handling weight drives lifecycle cost. Steel remains the benchmark where ductility, high stiffness, fire resistance, field fabrication or a familiar design-and-repair ecosystem controls. The materials are not one-for-one substitutes.
Every number below is tied to a named public agency, university or accredited laboratory. Research values are benchmarks, not guaranteed F1 Composite product properties; project procurement still requires grade- and lot-specific qualification data.
24.5%
of steel's equal-volume mass
Derived from 1.92 vs 7.85 specific gravity in the Kentucky/FHWA program.
40 vs 196
GPa elastic modulus
Mean/adopted GFRP vs steel values in the same controlled test program.
11 bridges
15–20 years in service
USDOT-funded core and extracted-bar durability investigation.
2 decks
GFRP vs epoxy-coated steel
Adjacent MnDOT bridges monitored side by side for roughly four years.
University of Kentucky researchers tested four 15 mm-diameter GFRP bars and three 16 mm-diameter epoxy-coated steel bars. GFRP reached a 612 MPa mean ultimate strength; steel yielded at 487 MPa and ultimately reached 653 MPa. The comparison shows why GFRP ultimate versus steel yield is a misleading marketing shortcut.
Scope before headline
These are measured means for different bar sizes and one legacy product set: valuable for explaining mechanics, not for sizing a current project. See Kentucky / FHWA report, Tables 2.3–2.5 ↗.
Tensile stress
GFRP ultimate
n=4
612
Steel yield
n=3
487
Steel ultimate
n=3
653
Elastic modulus
GFRP
40
Steel
196
Steel was 4.9× stiffer in this test.
Lower GFRP stiffness affects deflection, crack width, bar spacing and post-crack member response even when tensile capacity is adequate.
Failure behavior · schematic
GFRP: elastic → rupture
No steel-like yield plateau or plastic hinge behavior.
Steel: elastic → yield → plastic strain
Ductile deformation provides redistribution and warning.
Independent test-unit check
Its ISO/IEC 17025 quality-system, IAS-accredited, FDOT-qualified lab tested three 2019 production lots of No. 8 GFRP bar for a seawall project.
458.6 kN
3 production lots
Table 1 values are product-, size- and project-specific. ACI / Concrete International record ↗.
Values are intentionally attached to their test scope. A design value must come from the governing code and the qualified bar, not from a generic comparison table.
| Topic | GFRP rebar | Steel rebar | Evidence |
|---|---|---|---|
| Corrosion mechanism | Nonmetallic; does not undergo the electrochemical rusting that expands and spalls concrete. Resin, glass, bond, moisture, alkali and temperature still require qualification. | Normally passive in alkaline concrete. Chlorides or carbonation can disrupt passivity; corrosion products can crack and spall the cover. | FHWA corrosion primer + ASTM D7957 ↗ |
| Specific gravity | 1.92 in one controlled bridge-research test program. | 7.85 in the same report: equal-volume GFRP mass is 24.5% of steel (derived). | Kentucky / FHWA, Table 3.1 ↗ |
| Tensile result | 612 MPa mean ultimate strength, n=4 tested bars. | 487 MPa mean yield and 653 MPa mean ultimate strength, n=3 tested bars. | Kentucky / FHWA, Tables 2.3–2.4 ↗ |
| Elastic modulus | 40 GPa mean / adopted value in the same comparative test program. | 196 GPa adopted value: 4.9× the tested GFRP stiffness. | Kentucky / FHWA, Table 2.5 ↗ |
| Stress–strain behavior | Approximately linear-elastic to rupture; no steel-like yield plateau. | Yields before ultimate failure, providing plastic deformation and warning. | MnDOT Ch. 6 + DoD UFC Appendix G ↗ |
| Longitudinal thermal expansion | 9.18 µε/°C average for the tested No. 4 GFRP bar. | 11.88 µε/°C average for the tested No. 4 Grade 60 bar. | FHWA-HRT-05-081, 3 chamber runs ↗ |
| Electrical / magnetic response | Electrically insulating and nonmagnetic: useful near MRI, rail, power and sensing equipment. | Electrically conductive and ferromagnetic. | FDOT owner guidance ↗ |
| Detailing and fabrication | Factory-formed bends; field bending is prohibited in current DoD guidance. Serviceability often controls. | Familiar ductile detailing, field fabrication and mechanical-splice ecosystem, subject to the governing code. | UFC 3-301-01, Appendix G ↗ |
Chloride ingress
Passive film breaks
Rust expands
Crack · delaminate · spall
FHWA explains that chlorides can disrupt steel's passive film; corrosion products then generate pressure that cracks and spalls the concrete cover. Coatings and concrete quality can delay the chain, but do not change its underlying mechanism. FHWA corrosion fundamentals ↗.
GFRP cannot rust, but glass–resin interfaces and bond can still change under hot, wet, alkaline or sustained-load exposure. That is why ASTM D7957 qualification, environmental reduction factors and lot traceability matter. ASTM D7957/D7957M-26 scope ↗.
Controlled material test
4 GFRP + 3 steel bars
A direct laboratory comparison measured tensile strength, yield behavior, modulus and specific gravity. It is unusually useful because both materials were tested inside one program; it is still one 2000-era product set, not a universal catalog value.
Open the public report ↗Side-by-side bridge decks
~4 years monitored
Adjacent 2018 bridge decks (one GFRP, one epoxy-coated steel) both behaved as designed. GFRP strains were slightly higher but not notably so, and crack patterns were generally similar. The authors call four years a snapshot, not a 75-year validation.
Open the public report ↗In-service extraction
11 bridges · 15–20 years
Concrete cores and extracted bars were examined by microscopy, chemistry and mechanical testing. Results were encouraging, but the report's 100-year strength projection is modeled from limited field evidence and has been debated. It is not a completed 100-year exposure test.
Open the public report ↗Read the published ASCE discussion ↗MnDOT's 100-year bridge-deck analysis used an initial installed deck cost of $36/ft² for steel and $42/ft² for GFRP. Its result changes with assumed deck life and whether an interim GFRP repair is included.
This is a 2023 project scenario using agency cost inputs, a 1.22% discount rate and specific repair assumptions. It is neither current market pricing nor a guaranteed savings model. MnDOT Report 2023-13, Chapter 7 ↗.
100-year present value
Epoxy-coated steel
65-year life scenario
$87.14
GFRP · no interim repair
65-year life scenario
$67.22
GFRP · repair at year 30
65-year life scenario
$85.21
GFRP often earns preference
Steel often earns preference
Fire and seismic statements above identify current U.S. DoD restrictions, not a universal worldwide ban. Always apply the jurisdiction's governing code and owner specification. UFC 3-301-01, Appendix G ↗.
ACI CODE-440.11-22
Design and construction code covering strength, serviceability, development, splices, durability, inspection and elevated-temperature considerations.
ASTM D7957/D7957M-26
Product specification for qualification and lot acceptance, including physical properties, tensile force/modulus/strain, shear, bond and bent-bar requirements.
ASTM D7205/D7205M-26
Determines short-term static tensile force, strength, strain, modulus and stress–strain response; it does not establish sustained-load or fatigue performance.
ASTM A615/A615M-26
The current ASTM product specification for carbon-steel reinforcement. Project design requirements still come from the applicable concrete code.
Minimum decision-grade submittal
Bar identification and measured area · guaranteed tensile force by size · elastic modulus and rupture strain · bond and transverse shear · resin / glass declaration · cure and glass-transition temperature · environmental durability data · bend qualification · lot traceability · handling and inspection plan.
It can have higher ultimate tensile strength per unit area, but that does not make it a universally stronger replacement. GFRP has roughly one quarter of steel's elastic modulus and no yield plateau. Compare ultimate-to-ultimate values, then design for serviceability, bond, environmental reduction factors and the governing code: never compare GFRP ultimate strength only with steel yield strength.
No. Bar area, spacing, cover, development length, lap details, crack-width control and deflection must be recalculated. Lower stiffness frequently means that serviceability controls even when ultimate tensile capacity is high. Use ACI CODE-440.11 and project-qualified ASTM D7957 bars where those documents are applicable.
No. It removes the steel rust-expansion mechanism, but concrete still cracks from shrinkage, temperature, restraint and loading. In the MnDOT side-by-side bridge study, the GFRP and epoxy-coated-steel decks developed generally similar surface and full-depth crack patterns during the first four years.
No. The MnDOT case had a higher initial deck cost for GFRP ($42 versus $36 per square foot), but lower modeled present value when the assumed service life was long enough. Those numbers depend on discount rate, repair timing, deck type, labor and local material prices; they are a sensitivity study, not a market quote.
The strongest U.S. field program cited here examined bars from 11 bridges after 15–20 years and found encouraging condition. That is not the same as 75–100 years of field validation. Longer horizons rely on accelerated testing, environmental reduction factors and models, so project exposure, resin system, glass, cure, sustained stress and test documentation matter.
GFRP cannot be welded, and current DoD guidance prohibits field bending. Bends and stirrups should be factory formed and qualified, then protected from damage in handling. Field cutting may be permitted under the project specification using appropriate tools and dust controls.
Send the dimensions, quantity, service conditions and destination. We reply within one business day.