Steel & Reinforcement
The desktop AI reads from the same knowledge base — this page is its English edition. When the AI cites 📖 guide · materials-steel in a conversation, it refers to this entry.
In One Sentence
Steel has only three real choices: Steel01 (bilinear, good enough), Steel02 (captures the Bauschinger effect—required for cyclic loading), and Steel04 (buckling and fracture—only needed for collapse analysis). Choosing Steel01 vs. Steel02 comes down to whether the structure will enter plastic cycling.
Choosing Among the Three Models
| Model | Constitutive features | When to use | When it must not be used |
|---|---|---|---|
| Steel01 | Bilinear + isotropic hardening | Static, monotonic loading, small deformation | Cyclic loading (overestimates hysteretic energy) |
| Steel02 | Menegotto-Pinto + kinematic hardening | Seismic time history, Pushover, any cyclic loading | When local buckling must be modeled |
| Steel04 | Multilinear + buckling + fracture | Member buckling, ultra-low-cycle fatigue, collapse | Routine design (too many parameters, unnecessary) |
Rule of thumb: loaded monotonically to failure once → Steel01; loaded back and forth → Steel02; need “what happens after buckling” → Steel04.
Under cyclic loading, Steel01 produces an overly fat hysteresis loop, overstating dissipated energy by 20–40%, and seismic response gets underestimated. This error never raises a message; it just makes the results look like “the structure performs well”—the most dangerous class of error.
Only Three Core Parameters
| Parameter | Meaning | Structural steel (Q235) | Rebar (HRB400) | Notes |
|---|---|---|---|---|
| fy | Yield strength | 235 | 400 | Under mm-kN-s, numerically equal to MPa |
| E | Elastic modulus | 206 | 200 | Not 206000 |
| b | Strain-hardening ratio | 0.01–0.02 | 0.01 | Dimensionless, not the hardening modulus |
b is defined as Esh/E (hardening-branch modulus ÷ elastic modulus), not the hardening modulus itself.
- Hot-rolled steel’s hardening-branch modulus is about 2–3 GPa; with E = 206 GPa → b ≈ 0.010–0.015
- b = 0 means elastic-perfectly-plastic: internal force redistribution in statically indeterminate structures cannot develop; capacity runs conservative
- b = 0.1 means a hardening modulus of 20 GPa, 7 times the real value, and ultimate capacity will be overestimated
This is the easiest steel parameter to get wrong: many people treat b as a “hardening percentage” and enter 2 or 5, making the hardening modulus 2–5 times E—stiffer than the elastic branch.
Steel02’s Extra Parameters Are Not Physical Parameters
Besides fy, E, and b, Steel02 has three parameters—R0, cR1, cR2—that control how smoothly the Menegotto-Pinto curve transitions from the elastic to the plastic branch:
- R0: initial curvature parameter; larger values make the transition sharper (closer to bilinear)
- cR1, cR2: control the decay of R during kinematic hardening
These three are numerical fitting parameters, not material properties. They don’t affect yield strength, elastic modulus, or ultimate strength—only the roundness of the hysteresis corners.
Practical guidance:
- Use the defaults for routine analysis (R0 = 10–20, cR1 = 0.925, cR2 = 0.15) and don’t touch them
- Only tune them when fitting a measured hysteresis curve, and the calculation report must state “calibrated against test XX”
- Tuning these three to “hit” a desired result is academic misconduct and the starting point of engineering failures
Criterion: if R0 ends up below 5 or above 30, you’re using it to compensate for some other error (usually a wrong b or wrong boundary conditions).
Which Model for Rebar
| Scenario | Model | Reason |
|---|---|---|
| Ordinary RC members | Steel02 | Rebar’s Bauschinger effect is pronounced under cyclic loading; Steel01 overestimates energy dissipation |
| Rebar buckling must be modeled (column plastic hinges) | ReinforcingSteel | Has buckling branch and low-cycle fatigue |
| Rebar fracture must be modeled | Steel04 or ReinforcingSteel | Has fracture criterion |
Use measured or characteristic values for rebar fy, not design values:
- HRB400’s fy,k = 400 MPa (characteristic); the design value fy = 360 MPa (already divided by γs = 1.1)
- The constitutive model’s yield strength corresponds to the actual material; entering 360 underestimates rebar stress by about 11%
- If a material test exists, use the test value—typically 5–10% above the characteristic value
Five High-Frequency Errors
Error 1: Entering the hardening modulus as b
b = 2000 (meaning to enter a 2 GPa hardening modulus). The hardening modulus becomes E × 2000 and the results are completely distorted. b is a ratio, on the order of 0.01.
Error 2: Entering E as 206000
That’s the MPa-system value. Under mm-kN-s, entering 206000 makes stiffness 1000 times too large and displacements 1000 times too small. It looks like “an extremely stiff structure.”
Error 3: Using Steel01 under cyclic loading
Overly fat hysteresis loop, energy dissipation overestimated, seismic displacement response underestimated. If any analysis step contains cyclic loading, Steel02 is mandatory.
Error 4: Tuning R0 to hit a result
R0 only affects the transition roundness. Using it to tune capacity or displacement means compensating for an error in the wrong place.
Error 5: Using one parameter set for steel and rebar
Rebar’s E is slightly lower (200 vs. 206), it lacks a marked yield plateau, and its fy is higher. Copying steel parameters directly underestimates rebar stress.
Criteria: Three Consistency Checks
- Yield strain = fy/E. Q235 should give 0.00114; HRB400 should give 0.002. Getting 0.0001 or 0.02 means the units are wrong
- Order of magnitude of b: between 0.005 and 0.03. Outside that means the concept is wrong
- Ultimate strength / yield strength: about 1.5 for structural steel (Q235’s fu/fy ≈ 1.57), about 1.25 for rebar (HRB400’s fu/fy = 540/400 = 1.35). If the model gives 5×, b is too large
The calculation report prints fy/E and b for every steel material—these two lines are what reviewers always read.