What Product Standard Governs Prestressing Steel Strand in China?
Prestressing steel strand in China is governed by GB/T 5224-2023 Steel Strand for Prestressed Concrete, which replaced the 2014 edition and took effect on March 1, 2024. It is technically aligned with ISO 6934-4 Steel for the Prestressing of Concrete — Part 4: Strands. The North American counterpart is ASTM A416 (product specification) tested per ASTM A1061 (test method), and the European test-method reference is ISO 15630-3. Knowing which product standard applies up front decides the acceptance thresholds — the test method is portable across standards, but the pass/fail numbers are not.
The strand itself is a helical construction: the dominant structural type is 1×7 — six outer wires wound around a central king wire — in nominal diameters from 9.50 mm up to 17.80 mm, with strength grades from 1720 MPa to 2160 MPa. The 1×19 construction (nineteen wires in two layers) is used for higher-capacity applications such as bridge stay cables and mining support. The product standard defines both the geometry and the mechanical minimums for each combination; the test methods confirm whether a production batch meets those minimums.
What Are the Mechanical Property Requirements for a 1×7 Strand?
The mechanical block of GB/T 5224-2023 is organised by diameter × strength grade, and the threshold table is the core of any strand test report. For the most common 1×7 construction:
| Nominal diameter (mm) | Strength grade (MPa) | Breaking force Fm (kN) | Proof force Fp0.2 (kN) | Total elongation at max force Agt |
|---|---|---|---|---|
| 9.50 | 1720 | ≥ 88.0 | ≥ 79.0 | ≥ 3.5 % |
| 11.10 | 1860 | ≥ 138.0 | ≥ 124.0 | ≥ 3.5 % |
| 12.70 | 1860 | ≥ 184.0 | ≥ 166.0 | ≥ 3.5 % |
| 15.20 | 1860 | ≥ 260.0 | ≥ 234.0 | ≥ 3.5 % |
| 15.70 | 1770 | ≥ 266.0 | ≥ 239.0 | ≥ 3.5 % |
| 17.80 | 1720 | ≥ 327.0 | ≥ 294.0 | ≥ 3.5 % |
Three numbers carry the diagnostic weight in this table:
Breaking force (Fm, 最大力) — the maximum tensile load the strand sustains, the headline strength number. For the 15.20 mm / 1860 MPa grade (the most common bridge and building strand), the floor is 260 kN. A strand that tests below its grade's Fm has either been under-drawn, over-annealed, or is simply mislabelled to a higher grade than it actually delivers.
Proof force (Fp0.2, 规定非比例延伸力) — the load at 0.2 % non-proportional elongation. The standard requires this to be no less than 88–90 % of Fm. The Fp0.2/Fm ratio is what separates a ductile high-strength strand from a brittle one: a strand that hits its Fm but shows an abnormally low Fp0.2 will yield early in service and lose prestress before the concrete has cured — the structural failure mode the test exists to prevent.
Total elongation at maximum force (Agt, 最大力总伸长率) — the strand's strain at the peak load, with a floor of 3.5 %. This is not the strain at fracture — it is the strain at maximum force, measured before necking begins. Agt below 3.5 % flags a brittle strand, typically from excessive cold-drawing or hydrogen pickup, that will snap rather than yield under overload. The measurement requires an extensometer that stays on through maximum force, not a crosshead-displacement reading.
How Is the Static Tensile Test Performed?
The static tensile test is the primary quality test for every strand batch, performed per GB/T 21839 (the Chinese test-method standard for prestressing steel) and ISO 15630-3 / ASTM A1061 internationally. The procedure looks simple on paper — pull to fracture and record force-elongation — but two operational details decide whether the result is valid.
Gauge length. The standard gauge length L0 is typically 500 mm or 610 mm, long relative to the strand diameter because the elongation at maximum force (Agt, ≥ 3.5 %) must be measured over a representative length of the helical structure. Short gauge lengths exaggerate local strain and overstate Agt. The 610 mm (24 inch) gauge is the ASTM A1061 default; the 500 mm is the common GB/T choice.
Grip failure. A strand is a helix of smooth round wires with tensile strength up to 2000 MPa. When gripped directly by serrated jaws, the outer wires notch at the tooth marks and fracture at the grip — a break that is invalid under every strand standard because it reflects clamp damage, not material strength. Only a fracture in the free span between grips counts. Valid strand testing therefore requires purpose-built grips: standard practice is a conical sleeve + counter-cone assembly into which each strand end is fed, supplemented by hydraulic parallel-closing jaws with controllable pressure. The cone assembly carries most of the gripping force and spreads it over the helix without notching; the hydraulic jaws supply the remainder and prevent slip. This is why strand tensile testing is frequently outsourced — the gripping hardware is specialised and costly.
Extensometry. Because a strand fracture releases the stored elastic energy of six outer wires springing apart, a contact-arm extensometer is at physical risk at the moment of break. Optical (video) extensometers are the safer choice — they measure strain without touching the specimen, survive the fracture event, and track gauge marks through a field of view (typically 500–680 mm) large enough to cover the full gauge length. The extensometer must be capable of recording through maximum force to capture Agt.
What Is the Stress Relaxation Test and Why Does It Matter?
Stress relaxation is the property that makes a strand suitable for prestressed concrete rather than just for any high-strength application. When a strand is tensioned and held at constant length (the in-service condition inside concrete), its stress decays over time as the steel creeps. If the decay is excessive, the prestress that the strand was installed to deliver is lost, and the concrete loses its compression — the structural design assumption breaks down.
The test, per GB/T 5224 Clause 7 and GB/T 21839, loads the strand to a defined initial force (expressed as a percentage of the actual measured Fm) and holds the length constant while recording the force decay over 1000 hours. The acceptance thresholds for the low-relaxation grade:
| Initial force (% of Fm) | 1000 h relaxation (low-relaxation grade) |
|---|---|
| 60 % | ≤ 1.5 % |
| 70 % | ≤ 2.5 % |
| 80 % | ≤ 3.5 % |
The default test is at 70 % Fm, and the acceptance ceiling is 2.5 % loss over 1000 hours. A strand that loses more than 2.5 % at 70 % Fm is not low-relaxation grade — it is either a regular-relaxation product (stabilising-anneal omitted) or a defective batch. Because a full 1000-hour test is impractical for routine release, GB/T 5224 permits extrapolation from a 120-hour test to the 1000-hour value, using the log-time relaxation law. This is a pragmatic compromise — the 120-hour value is cheaper to obtain and correlates tightly with the 1000-hour result for properly stabilised strand, but it is an extrapolation, not a measurement, and the report must state which was used.
The relaxation test is the single longest-duration test in the strand panel, and it is the one most often omitted in cheap compliance reports. A strand can pass tensile and pass fatigue yet still be unsuitable for prestressed concrete if its relaxation is out of spec — the strength is there, but it does not persist.
What Does the High-Cycle Fatigue Test Prove?
The high-cycle fatigue test proves that the strand survives the dynamic loading it will see in service — traffic loads on a bridge, wind on a stay cable, crane cycles on a runway girder. Per ISO 15630-3 (and the GB/T framework that adopts its logic), the specimen must survive 2 million cycles at a maximum frequency of 20 Hz without fracture.
The fatigue test is where specimen gripping becomes most acute. The gripping force superimposes a clamping stress on the dynamic test load at the grip, and a strand that would survive 2 million cycles in the free span can fail at a few hundred thousand cycles at the grip if the clamp notches the wire. A fracture near or in the grip is an invalid test — it must be repeated, and because the test runs for several days, invalid tests are costly. Valid strand fatigue testing uses the same conical-sleeve-plus-hydraulic-jaw grip design as the static test, with the clamping pressure tuned to hold the strand without inducing fatigue initiation at the teeth.
The fatigue requirement is not universally applied — it depends on the application and the contract. A strand for a statically-loaded post-tensioning slab may not need fatigue clearance; a strand for a cable-stayed bridge or a railway bridge absolutely does. The test request should specify the stress range (the fatigue-driving parameter) and the number of cycles required, because these come from the structural design, not from the product standard.
How Do the Chinese, ISO, and ASTM Strand Frameworks Map?
The strand frameworks are more closely aligned than in many material families, because the product (a high-strength helical steel strand for prestressed concrete) is an international commodity. The practical correspondence:
| Scope | China (GB/T) | International (ISO) | North America (ASTM) |
|---|---|---|---|
| Product specification | GB/T 5224-2023 | ISO 6934-4 | ASTM A416 |
| Test methods | GB/T 21839, GB/T 228.1 | ISO 15630-3 | ASTM A1061 |
| High-cycle fatigue | GB/T 21839 (fatigue clause) | ISO 15630-3 | ASTM A1061 (annex) |
The mechanical thresholds are closely aligned: a 15.20 mm / 1860 MPa strand delivers Fm ≥ 260 kN under GB/T 5224, and the equivalent ASTM A416 Grade 1860 delivers the same. The differences are in scope and presentation rather than physics — GB/T 5224-2023 covers the 1×19 mining-support construction explicitly and adds specific clauses for it, while ASTM A416 focuses on the 1×7 construction used in building and bridge prestressing. A laboratory that understands both frameworks can run a combined programme, but the report must state each result against the specific standard.
The one substantive divergence to watch: the relaxation test's default initial force. GB/T 5224 defaults to 70 % Fm; some ASTM and ISO contracts specify 80 % Fm, which is the tighter test (≤ 3.5 % at 80 % is harder than ≤ 2.5 % at 70 %). A strand can pass one and fail the other, so the initial-force percentage must be stated in the test request.
Our Testing Capabilities
Beijing ZKGX Research provides prestressing steel strand testing against the GB/T 5224-2023 product standard, with test methods per GB/T 21839 and GB/T 228.1, and cross-reference to ISO 15630-3 and ASTM A1061.
Static mechanical properties:
- Breaking force Fm (up to 600 kN capacity)
- Proof force Fp0.2 (0.2 % non-proportional elongation)
- Total elongation at maximum force Agt (≥ 3.5 %)
- Yield strength Rp0.2 per GB/T 228.1
- Optical extensometry through maximum force
Stress relaxation:
- 1000-hour relaxation at 60 % / 70 % / 80 % initial force
- 120-hour extrapolation to 1000-hour value (low-relaxation grade)
- Low-relaxation (≤ 2.5 % at 70 % Fm) vs regular-relaxation classification
Dynamic testing:
- High-cycle fatigue, 2 million cycles at up to 20 Hz
- Servohydraulic loading with conical-sleeve strand grips
- Invalid-grip-fracture detection and re-test protocol
Surface and geometry: wire diameter, pitch of the helical lay, surface defect inspection (nicks, scratches, rust), straightness. Dimensional verification per the GB/T 5224 geometry tables.
Sample constructions: 1×7 (9.50 – 17.80 mm), 1×19 (bridge stay cable, mining support), in strength grades from 1720 to 2160 MPa.
If you need a GB/T 5224 strand report — for a prestressed-concrete project acceptance, a market-supervision sampling response, a supplier batch release, or a bridge-cable fatigue qualification — contact our laboratory with the strand's designated structure (e.g. 1×7-15.20-1860), strength grade, and the applicable standard, and we will scope the test plan.
FAQ
What is the difference between Fm and Fp0.2 in a strand test report?
Fm (最大力) is the maximum breaking force — the highest load the strand sustains. Fp0.2 (规定非比例延伸力) is the load at 0.2 % non-proportional elongation, the strand's proof strength. GB/T 5224 requires Fp0.2 to be at least 88–90 % of Fm. The ratio matters: a strand with high Fm but low Fp0.2/Fm will yield prematurely in service and lose prestress, even though its ultimate strength is on grade.
Why is the elongation measured at maximum force (Agt) and not at fracture?
Because fracture elongation includes the localised necking that happens after maximum force, which is unstable and varies with where the break initiates in the helix. Agt — the strain at maximum force — is measured before necking, is reproducible, and is what the structural designer uses to confirm the strand will yield (not snap) under overload. The 3.5 % floor in GB/T 5224 is on Agt, not on fracture strain.
Can a 120-hour relaxation test replace a 1000-hour test?
For routine release, yes — GB/T 5224 permits extrapolation from 120-hour data to the 1000-hour value using the log-time relaxation law, and the correlation for properly stabilised (low-relaxation) strand is tight. But it is an extrapolation, not a measurement. For a qualification test on a new product, or for a dispute, the full 1000-hour test is the definitive result. The report must state which was used.
Does every prestressing strand need a fatigue test?
No. Fatigue testing is required for strands in dynamically-loaded structures — cable-stayed bridges, railway bridges, crane runways, structures subject to significant cyclic live load. A strand in a statically-loaded post-tensioning floor slab typically does not require fatigue clearance. The requirement comes from the structural design code, not from GB/T 5224 itself. Specify the stress range and cycle count from the design when requesting the test.
What makes a strand test result invalid?
A fracture at or near the grip, caused by clamp notching rather than material strength, is invalid under GB/T 21839, ISO 15630-3, and ASTM A1061 — the result must be discarded and the test repeated. Other invalidity conditions: fracture outside the gauge length, equipment malfunction, and specimen preparation damage (bending, nicking, heating during cutting). This is why strand testing uses conical-sleeve grips and long gauge lengths — to force valid breaks in the free span.