Engineering AI infrastructure SI units · self-checking REST · MCP · Python v0.5 · evidence status public
// Simulation infrastructure for engineering AI

Let agents design, simulate, verify and iterate.

Your agent can author geometry, materials, boundary conditions and loads—or import existing CAD and solver data. physicsbase meshes where needed, solves, checks the numerics and returns a model the agent can change and run again.

Run a live model → Read the API docs
ONE-TIME SOLVE · DESIGN VERIFICATION · OPTIMIZATION LOOP
  ONE INPUT PATH · STEP → MESH → SOLVE → RETURN
POST /v1/import

file       2020_corner_bracket.step
material   6061-T6 aluminium
load       100 N lateral
analysis   static

→ CAD imported
→ mesh             194 nodes · 501 tet4
→ peak stress       110 MPa
→ safety factor     2.18
→ equilibrium       balanced
→ mesh assessment   refinement advised
→ model bundle      ready
25
analysis types
25
element types
181
verified definitions · 987 runs
Every result
includes self-checks
New product · physicsbase Copilot

An expert FEA agent
that runs the solver.

Not a chatbot that guesses. Describe a structure and a question in plain language; Copilot plans the analysis, calls the real verified physicsbase solver, and answers in SI units — with the PASS/REVIEW verdict and the reference it was checked against. It returns the solved model itself: an interactive contour with the boundary conditions, loads and material properties drawn and labeled — switch result fields, toggle the mesh or deformed shape, hover the legend to isolate a stress band, and hover the model to read values at any node.

Try Copilot free → See subscription
SUBSCRIPTION · CLAUDE OR GPT · DRIVES THE SAME VERIFIED ENGINE AS THE API
PHYSICSBASE COPILOT
You
Cantilever plate, 2×0.4 m, 5 kN tip load, steel — show me the stress.
physicsbase Copilot
▶ ran static · PASS
Peak von Mises ≈ 5.3 MPa at the fixed root, as expected. Equilibrium residual passed. Here's the contour:
5.3M0.1MPa
von Mises · fixed edge + tip load · interactive in the app
02

An agent can propose a design.
physicsbase lets it test one.

Agents own the engineering intent and the next decision. physicsbase provides a consistent numerical execution layer, so each design can be solved, checked, changed and compared without rebuilding a solver.
Agent owns design intent

Create the model it actually needs.

Generate geometry or mesh connectivity, choose materials and sections, set supports, contacts and loads, then decide whether to stop, verify or iterate.

geometry · materials · boundary conditions · objectives
physicsbase owns numerical execution

Get a computed model back—not a dead-end report.

Meshing, assembly, solution, result recovery and scoped checks return through one contract, together with a reusable model for the agent's next decision.

computed fields · explicit checks · reusable model
01

Run a real model
on the live server.

Pick a problem and adjust an input; the API meshes and solves it on the server and returns stress or deformation, a self-check and the complete model bundle an agent can modify and re-submit.

Solve a real model on the live server.

Pick a problem, adjust the input, then run the real server solve and inspect the returned model. No key is required.

REAL API · NO KEY · MODEL BUNDLE RETURNED
  MESH · SOLVE · VERIFY — ready
Ready to run the complete server-side solve.
08

See the engine solve
the difficult cases.

Reproducible studies show the formulation, benchmark, result and limitations—not merely the final image.
Hertzian contact pressure ellipse

Contact without penalty tuning

Recover the Hertz pressure ellipse with an exact active-set solver.

→ read the case study
von Mises snap-through path

Past the limit point

Follow snap-through with Crisfield arc-length continuation.

→ read the case study
Moving heat-source temperature field

A moving heat-source field

Reproduce the melt-pool temperature field against the Rosenthal solution.

→ read the case study
More formulations, benchmarks and implementation notesEXPLORE ALL CASE STUDIES →
05

One contract for a single answer
or an entire design loop.

Use the same model contract for a one-time simulation, a requirement check, a comparison study or iterative optimization across geometry, material, section, load and boundary-condition choices.
01 · Structural

Loads, motion and stability.

Statics, modal, buckling, harmonic response, transient dynamics and prestressed modes.

02 · Nonlinear

When linear assumptions end.

Plasticity, contact, friction, hyperelasticity, large displacement and arc-length continuation.

03 · Thermal

Heat and transport.

Conduction, diffusion, moving heat sources, one- or two-way coupled thermo-mechanical response and acoustic cavity modes.

04 · Failure + design

Predict, grow and optimize.

Damage, XFEM, phase-field fracture, progressive composite failure, stress-life fatigue and topology optimization.

05 · Elements

From members to solids.

Bars, beams, frames, 2D and 3D solids, plates, shells and scalar-field elements.

06 · Model creation + import

Author it or bring it.

Submit agent-created nodes and elements directly, or import CAD, Abaqus, Nastran, Gmsh, VTK, STL and other engineering formats.

See all 40 analyses and 25 element typesOPEN FULL MATRIX →
03

Every result comes with
reasons to trust it.

Evidence is split by what it can prove: 987 numerical comparisons across 181 definitions, four solution benchmarks plus three experimental benchmarks (all passing), and no unearned physical-validation claim.
0.3% error

Hertzian contact

Contact half-width and peak pressure against Hertz's analytical solution.

→ case study
u = 0.3024

Scordelis-Lo roof

The standard shell benchmark matches the published reference displacement.

→ case study
residual < 10⁻¹⁵

Snap-through path

Arc-length continuation follows the equilibrium path through the limit point.

→ case study
987 / 181

Numerical Verification Guide

987 executed comparisons, correctly identified as 181 case definitions.

→ inspect every comparison
not established

Physical Validation Status

Zero qualifying experiments. See the evidence required before that claim can change.

→ inspect the admission gate
4 pass / 0 fail

Benchmark Guide

Inspect each published target, mesh history, fixed tolerance and convergence history.

→ inspect the evidence
04

From design intent
to the agent's next decision.

The loop begins with agent-authored intent or imported data and ends with fields, checks, provenance and a complete model ready to change and run again.
01 / AUTHOR

Create or import.

Generate geometry or finite elements directly, or start from STEP, IGES, BREP, Abaqus, Nastran, Gmsh, VTK or STL data.

02 / DEFINE

Express the intent.

Set materials, sections, supports, contacts, loads, analysis settings, requirements and design variables.

03 / COMPUTE

Mesh, solve and check.

physicsbase builds the numerical model, runs the analysis and evaluates the checks supported for that analysis.

04 / ITERATE

Use the model again.

Inspect fields, reactions, provenance and evidence; then change the returned bundle, compare it or re-submit it.

06

Use it once—or make it
the agent's design loop.

The API publishes capabilities, schemas and discovery files. An agent can author or import a model, solve it, interpret the evidence and use the returned state as the starting point for its next design.

One-time simulation. Verification gate. Optimization engine.

Connect an engineering copilot, autonomous design workflow or simulation platform. The same editable model, computed fields and evidence travel through REST, MCP, OpenAPI and Python.

Open the integration guide →
GET/llms.txtcompact discovery for language models
GET/v1/capabilitiessupported analyses, elements and limits
GET/openapi.jsontool schemas for agent frameworks
MCP/mcpnative tools for compatible agents
POST/v1/importexisting CAD, mesh or solver data to simulation
POST/v1/solveagent-authored model to checked, reusable result
07

What agents cannot ask
physicsbase to do yet.

This is the current capability boundary, not a promise disguised as a feature list. Partial areas and known failures remain visible until implementation and evidence are complete.
Contact · not yet

General contact interfaces.

Self-contact; large-sliding surface-to-surface and mortar contact; frictional dynamic contact; cohesive-zone and delamination interfaces.

Materials · partial

Broader nonlinear constitutive models.

Shipping: J2 plasticity (temperature-dependent yield), finite-strain J2, Perzyna viscoplasticity, Norton creep (2D plane-strain + 3D), Prony viscoelasticity, Ogden/Mooney-Rivlin/HGO/GOH hyperelasticity, Lemaitre damage-plasticity (trusses + continuum). Still open: kinematic/combined (Chaboche) hardening; Hill anisotropic yield; plane-stress creep.

Shells + solids · partial

Advanced formulations and laminate recovery.

Curved/degenerated and solid-shell elements; unsymmetric laminate B-coupling and per-ply recovery; an analytic consistent tangent for the continuum finite-strain plasticity solver.

Dynamics · partial

Nonlinear dynamics beyond trusses.

Shipping: quad4 continuum nonlinear dynamics (implicit Newmark + Newton) and complex damping-coupled modes (damped modal). Still open: beam and shell nonlinear dynamics; explicit impact/crash/wave propagation; contact coupled to dynamics; rotordynamics (gyroscopic/Campbell).

Fracture + fatigue · partial

Mixed-mode growth and life prediction.

Shipping: stress-life (Basquin S-N, Miner, Goodman/Soderberg/Gerber, rainflow), Paris-law crack growth, and strain-life (Coffin-Manson-Morrow) fatigue. Still open: mixed-mode KI/KII separation with crack-tip enrichment; crack-path tracking as a service.

Fluids · partial

Viscous CFD and FSI.

Shipping: viscous incompressible Stokes, full steady Navier-Stokes with convection, and transient (unsteady) Navier-Stokes (stabilized Q1-Q1), verified vs Poiseuille, Kovasznay and the Taylor-Green vortex; plus inviscid potential flow and scalar transport. Still open: turbulence models; free-surface flow; fluid-structure interaction.

Composites + multiphysics · partial

Beyond CLT and staggered coupling.

Shipping: two-way thermo-mechanical coupling (strain-coupled conductivity, staggered fixed point), piezoelectricity (1D electro-mechanical bar), acoustics (cavity/duct modes + driven Helmholtz response), vibro-acoustic structure↔fluid coupling and 2D magnetostatics (magnetic vector potential, B=curl A). Still open: 3D woven RVEs, full WWFE envelopes, interlaminar/free-edge response, fully monolithic coupling, time-harmonic/eddy-current electromagnetics, 2D/3D piezo/vibro-acoustic continua, and hygrothermal response.

Meshing + import · partial

Production geometry automation.

Feature-aware sizing and quality controls for agent-authored geometry; full solver-deck assembly fidelity; p-refinement; contact-aware and boundary-layer meshes; remeshing surface-only 3D files.

Solver scale · partial

Large nonlinear and distributed solves.

Iterative routing beyond static analysis; parallel, distributed and GPU assembly/solve; model-order reduction, substructuring and superelements.

Evidence · partial

Decision-grade uncertainty and validation.

Nonlinear spatial and nonlinear-dynamic time error estimators; multi-level and 3D contact convergence beyond today's opt-in two-grid check; goal-oriented error; uncertainty quantification; stable targeted refinement; a qualified experimental campaign. Physical validation is not established today.

Source-grounded status, modelling limits and suggested build orderOPEN FULL ROADMAP →
STATUS LANGUAGE: “partial” means a narrower implementation exists today. “not yet” means the named capability is outside the current solver contract.

Give your engineering agent
a real simulation loop.

Start with one solve. Keep going through verification, comparison and optimization with the complete model returned after every run. The live example needs no key.

Start building free → Run the live model