Quick start

This page walks through the two main ways to use GenSec: the command-line interface (YAML-driven) and the Python API.

Command-line workflow

1. Write a YAML input file describing materials, section geometry, and load demands (see YAML input reference for the full specification).

2. Run the analysis:

   uv run gensec examples/example_input.yaml --output-dir results
   

   Or equivalently:

   uv run python -m gensec examples/example_input.yaml --output-dir results
   

3. GenSec produces the following outputs in results/:

Python API workflow

Minimal example: rectangular 300×600 mm RC column, 3Φ20 top + 3Φ20 bottom.

import numpy as np
from gensec.materials import Concrete, Steel
from gensec.geometry import RebarLayer, RectSection
from gensec.solver import FiberSolver, NMDiagram, DemandChecker

# --- Materials ---
concrete = Concrete(fck=25.0, gamma_c=1.5, alpha_cc=0.85)
steel = Steel(fyk=450.0, gamma_s=1.15)

# --- Section geometry ---
A20 = np.pi / 4 * 20**2                     # area of one Φ20 bar
rebars = [
    RebarLayer(y=40,  As=3 * A20, material=steel, n_bars=3, diameter=20),
    RebarLayer(y=560, As=3 * A20, material=steel, n_bars=3, diameter=20),
]
section = RectSection(B=300, H=600, bulk_material=concrete, rebars=rebars)

# --- Solver ---
solver = FiberSolver(section)

# Generate N-M interaction diagram
nm = NMDiagram(solver).generate(n_points=400)

# Verify a single demand
checker = DemandChecker(nm)
eta = checker.utilization_ratio(N=-1500e3, Mx_or_M=200e6)
print(f"Utilization ratio η = {eta:.3f}")     # η < 1.0 → verified

# Per-fiber stress state at a specific demand
sol = solver.solve_equilibrium(N_target=-1500e3, Mx_target=200e6)
results = solver.get_fiber_results(sol["eps0"], sol["chi_x"])

Using EC2 properties from \(f_{ck}\)

GenSec can compute all EN 1992-1-1 Table 3.1 parameters automatically:

from gensec.materials import concrete_from_ec2, concrete_from_class

# From fck (French National Annex)
c30 = concrete_from_ec2(fck=30, ls='F', loadtype='slow', TypeConc='R')
print(c30.fcd)       # 17.0 MPa
print(c30.eps_cu2)   # -0.0035

# From class name
c50 = concrete_from_class('C50/60', ls='F')

# Access the full EC2 property object
print(c30.ec2.fcm)   # 38.0 MPa
print(c30.ec2.fctm)  # 2.9 MPa
print(c30.ec2.ecm)   # 32837 MPa

Structural steel (EN 10025-2)

from gensec.materials import steel_from_en10025

s355 = steel_from_en10025('S355', t=20)   # 20 mm plate
# s355.fyk = 345 (thickness-dependent)
# s355.k_hardening = 470/345 (fu/fy ratio)

Biaxial bending

For biaxial analysis, assign explicit x coordinates to rebars and use a 2-D fiber grid (n_fibers_x > 1):

rebars = [
    RebarLayer(y=40,  x=40,  As=A20, material=steel),
    RebarLayer(y=40,  x=260, As=A20, material=steel),
    RebarLayer(y=560, x=40,  As=A20, material=steel),
    RebarLayer(y=560, x=260, As=A20, material=steel),
]
section = RectSection(
    B=300, H=600, bulk_material=concrete,
    rebars=rebars, n_fibers_y=60, n_fibers_x=30,
)

solver = FiberSolver(section)
nm_3d = NMDiagram(solver).generate_biaxial(
    n_angles=36, n_points_per_angle=200,
)

checker = DemandChecker(nm_3d)
eta = checker.utilization_ratio(N=-1000e3, Mx_or_M=100e6, My=50e6)

Mixed materials (e.g. RC + CFRP)

Use TabulatedMaterial for any arbitrary \(\sigma\text{-}\varepsilon\) curve:

from gensec.materials import TabulatedMaterial

cfrp = TabulatedMaterial(
    strains=[0.0, 0.017],
    stresses=[0.0, 2800.0],
    name="CFRP",
)

rebars = [
    RebarLayer(y=40,  As=942, material=steel, embedded=True),
    RebarLayer(y=560, As=942, material=steel, embedded=True),
    RebarLayer(y=5,   As=150, material=cfrp,  embedded=False),  # external strip
]

Set embedded=False for elements external to the bulk material so that GenSec does not subtract the displaced concrete (see Sign conventions).