WINDLOADS.AI
Open Averroes
How Averroes works

The engine turns cited drawings into a connected model.

Averroes reconstructs the proposed specimen, checks the engineering ledger, and compares the result with physical laboratory evidence.

AIR ASTM E283/E283MWATER ASTM E331 + E547STRUCTURAL ASTM E330/E330MIMPACT ASTM E1886 + E1996HVHZ TAS 201 + 202 + 203
Window specimen undergoing a large-missile impact test
LABORATORY REFERENCE PHOTOGRAPHSpecimen evidence informs the reconstructed model.
READ + RECONSTRUCT

Reconstruct the geometry before running the checks.

Averroes links elevations, sections, profiles, schedules, hardware, anchors, glazing, and test procedures. It reconstructs coordinate-based geometry and connected assembly topology before calculated properties are allowed to influence the outlook.

  • Page-level citations and source provenance
  • Calculated Ix, Sx, J, Cw, rx, and ry
  • Connected members, panels, anchors, and boundary conditions
Finite element connection and thin-walled section model with deformation and fastener detail
RECONSTRUCTED SECTION

FLEXUREIx · Iy

TORSIONJ · Cw

LOAD OFFSETShear center

Calculated only from cited and dimensionally resolved profile evidence.
SECTION PROPERTIES

Compute the section properties the drawings can support.

Most reviews accept a published moment of inertia and stop. Averroes reconstructs the profile from drawing evidence and computes the properties that govern flexure, torsion, warping, and eccentric load application in thin-walled open and semi-closed extrusions.

  • Ix, Iy, Sx, Sy, and radii of gyration
  • Torsional constant J and warping constant Cw
  • Shear center located relative to the applied load
  • Composite action assessed under AAMA TIR-A8, never silently assumed
Instrumented glass specimen after an impact test
Laboratory reference photographObserved impact conditionComparable evidence remains scope filtered
COMPARE + FORECAST

Compare the proposal with what happened in the lab.

More than 500 laboratory reports, observed failure sequences, and sensor data are indexed for contextual comparison. Comparable cases inform engineering judgment but do not statistically calibrate the forecast.

  • Comparable evidence filtered by applicability
  • Observed outcomes and measured response
  • Confidence tied to evidence coverage
Modeling basis

Stated assumptions keep each result checkable.

Averroes distinguishes drawing facts, reconstructed geometry, calculation assumptions, and laboratory evidence. A solver result is not presented without its boundary conditions, material basis, units, and independent checks.

  1. 01Geometry Member centerlines, profiles, glazing, anchors, and connection topology remain tied to cited drawing pages.
  2. 02Stiffness Reinforced and thermally broken members are evaluated under the applicable AAMA TIR-A8 composite-action basis.
  3. 03Boundaries Head, sill, jamb, anchor, slip, and support assumptions are listed instead of replaced by ideal pins.
  4. 04Glass Panel load resistance and nonlinear response use the applicable ASTM E1300 basis when project evidence supports it.
  5. 05Verification Units, pressure-area force, reactions, equilibrium, and demand-to-capacity arithmetic are recomputed independently.
  6. 06Limits Missing dimensions, unresolved materials, and unsupported contact behavior remain explicit limitations.
From evidence to decision

Build the model before the specimen.

Give Averroes the testing program and drawings. Keep cited sources, calculations, and assumptions visible.

Start your test review