Frequently Asked Questions

1. Do I need to know OpenFOAM to use this?

No. Describe your problem in plain language — the agent selects the solver, builds the mesh, sets boundary conditions, and runs the case. You only need a CAD file and the physical conditions (velocity, fluid, etc.).

2. What file formats can I upload?

Geometry must be uploaded as STP (STEP) format. Make sure the model is a clean, watertight solid — assemblies should be exported as a single STP file.

3. What types of simulation are supported?

Currently the platform supports incompressible flow — both internal (pipes, ducts, manifolds) and external (bluff bodies, airfoils, vehicles). Compressible, multiphase, and heat-transfer cases are on the roadmap but not yet available.

4. How accurate are the results?

The platform runs real OpenFOAM simulations — the same solver used across industry and academia. Accuracy depends on mesh quality and boundary conditions, both tuned automatically by the agent but reviewable and overridable by you.

5. How long does a typical case take?

It depends on case complexity and on your plan's compute allocation — higher-tier plans run on more CPU cores and finish faster. As a rough guide, simple external aerodynamics or pipe flow typically completes in 10–30 minutes on a standard plan; larger or finer-mesh cases can take longer.

6. How does pricing work?

Pricing has two parts:

1. Tokens, consumed by agent reasoning (setup, meshing logic, post-processing).
2. Compute, consumed by the OpenFOAM solver while it runs.

Different plans bundle different token allowances and compute tiers. New accounts receive free tokens to try the platform.

7. Is my geometry data kept private?

Yes. Cases are isolated per user — other users cannot access your files. For on-premise or air-gapped environments, the platform can be self-hosted via Docker Compose or Kubernetes.