Case archive · Verification & Validation

Every number can be recomputed

These are not customer stories — they are kernel verification records. Model files, analysis settings, key results and errors are public for every case. Download the model and recompute it yourself; if the numbers do not match, that is on us.

  • Models openEvery case ships with its full Tcl model file — elements, materials, analysis settings — rerunnable as-is in OpenSees 3.7.2.
  • Checked against analytical solutionsWherever an analytical solution exists, we list the analytical value, the computed value and the relative error — no "good agreement" hand-waving.
  • Failures recorded tooNon-converged and wrong runs stay on record — software that hides its failed cases does not deserve trust.
  • Version-boundEvery record states the solver version; after a kernel upgrade everything is rerun, and old results never pose as new ones.

CASE-A

Single-storey portal frame · nonlinear pushover

Verified

The most common structure type, used to see exactly what nonlinearity changes: after plastic hinges form at the column bases, capacity does not drop — it keeps rising. Elastic analysis cannot produce that part of the curve.

Three deformation states overlaid (30 / 60 / 100 mm), displacements scaled 4×
Three deformation states overlaid (30 / 60 / 100 mm), displacements scaled 4×
Base shear – top displacement curve from 100 real solution steps
Base shear – top displacement curve from 100 real solution steps

Model and settings

Geometry
Span 6.0 m · storey height 4.0 m
Sections
H400×200×8×13, fibre sections (web 10 fibres + flanges 4×2)
Materials
Steel01 · Fy = 345 MPa · E = 200 GPa · hardening ratio 0.015
Elements
forceBeamColumn × 4 · 5 integration points · columns include P-Δ
Analysis
Gravity (20 N/mm on the beam) + displacement-controlled pushover, 1 mm/step × 100 steps
Convergence
NormDispIncr 1e-6 · max 30 iterations · UmfPack

Key results

  • Peak base shear379 kN
  • Corresponding top displacement100 mm (2.5% drift)
  • Onset of yieldingApprox. 35 mm (clear kink in the curve slope)
  • Failure mechanismPlastic hinges at the column bases; beam ends unyielded
Download model filecaseA-portal-pushover.tcl

Runs as-is from the OpenSees 3.7.2 command line — no Kavvo needed to recompute it.

CASE-B

Four-storey two-bay steel frame · modal + pushover

Verified

Dynamic properties and the overall yield mechanism, read together. The fundamental period sets the seismic demand; the inter-storey drift distribution shows where the weak storey is — both numbers are below, with no room for vagueness.

First mode shape, annotated with storey drift shares of 16 / 47 / 77 / 100%
First mode shape, annotated with storey drift shares of 16 / 47 / 77 / 100%
Drift contour pushed to 7.5% drift
Drift contour pushed to 7.5% drift

Model and settings

Geometry
2 bays × 4 storeys · span 6.0 m · storey height 3.2 m
Sections
Columns H500×300×10×16 · beams H400×200×8×13 (all fibre sections)
Materials
Steel01 · Fy = 345 MPa · E = 200 GPa
Elements
forceBeamColumn × 20 · 5 integration points · with P-Δ
Mass
Lumped floor masses (equivalent to 245 kN on edge columns · 490 kN on interior columns)
Analysis
eigen(3) + displacement-controlled pushover under inverted triangular load, 1.5 mm/step × 160 steps

Key results

  • Fundamental period T₁0.460 s
  • 2nd / 3rd periods0.123 s / 0.055 s
  • Base shear (7.5% drift)628 kN
  • Yield mechanismBottom two storeys yield first, forming a global beam-hinge mechanism
Download model filecaseB-frame-modal-pushover.tcl

Runs as-is from the OpenSees 3.7.2 command line — no Kavvo needed to recompute it.

CASE-C

Cantilever beam deflection · analytical comparison

Verified

The most basic verification: a member you can compute by hand from a formula, checking that the kernel returns the same number. If this step fails, none of the later nonlinear results are worth discussing.

Model and settings

Geometry
Cantilever 3.0 m · rectangular 200×400 mm
Sections
Elastic · A = 80,000 mm² · I = 1.0667×10⁹ mm⁴
Materials
E = 30,000 MPa (concrete-scale, linear elastic)
Load
Tip point load 10 kN
Analytical
δ = PL³/(3EI)

Key results

  • ConclusionLinear-elastic deflection matches the analytical solution exactly
ItemComputedAnalyticalRelative error
Tip vertical displacement−2.8125 mm−2.8125 mm0.0000 %
Section moment of inertia I1.0667×10⁹ mm⁴1.0667×10⁹ mm⁴0.0000 %
Download model filecaseC-cantilever-verify.tcl

Runs as-is from the OpenSees 3.7.2 command line — no Kavvo needed to recompute it.

Want to see how your structure checks out?

The current records cover 3 member classes (frame pushover / multi-storey frame / fundamental member verification) and will grow with releases. Kavvo holds no certification from any design code body or third-party institution; these numbers only demonstrate that the kernel produces these values on our models, and do not constitute a compliance statement.