Your first simulation

The shortest path from opening the application to a result you can read. It uses a worked example, so there is nothing to import and nothing to draw.

Runs on the free tier.

Before you start

Two ideas make everything else make sense.

A case is one body and any number of studies of it. The geometry belongs to the case; the mesh, the physics and the run belong to a simulation inside it. That is why the setup tree has geometry once at the top and everything else repeated under each simulation — it is how you compare two meshes or two turbulence models on the same body without copying anything.

The application writes an ordinary OpenFOAM case. Nothing is hidden or proprietary. If you know OpenFOAM you can read every file it produces; if you do not, you never have to.

1. Open a worked example

From the home screen, choose Worked example. The case opens with its geometry already imported and its physics already chosen.

The setup tree on the left is the whole workflow, top to bottom:

Geometry
└── Simulation 1
    ├── Mesh
    ├── Refinements
    ├── Boundary layers
    ├── Operations
    ├── Quality
    ├── Models
    ├── Materials
    ├── Initial conditions
    ├── Boundary conditions
    ├── Advanced concepts
    ├── Numerics
    ├── Simulation control
    ├── Result control
    ├── Runs
    ├── Post-processing
    │   └── Boundary values
    └── Report

Advanced concepts only appears when the case has some — cell zones for MRF, AMI, overset, porous media or sources. It is not part of the free tier.

Work down it. Each step shows a status dot, and the dots are computed from the case rather than ticked off by hand — a step goes green because the rules that check it are satisfied, not because you visited it.

2. Generate the mesh

Select Mesh and press Generate mesh.

The mesh is built in the background and appears in the viewport when it finishes. Watch the log in the bottom dock if you want to see what it is doing.

Then look at Quality. It reports non-orthogonality, skewness and aspect ratio — the three numbers that decide whether a solver will struggle. A mesh that fails these does not usually announce itself by crashing; it announces itself by converging slowly to a slightly wrong answer, which is worse.

On the free tier the mesh is capped at 250,000 cells. The worked examples are sized to fit.

3. Look at what is already set

You do not need to change any of it, but look, because this is where the time goes on a real case:

• Models — the turbulence model. The default is k-omega SST, which is the reasonable default for external flow with adverse pressure gradients.
• Materials — the fluid, with its density and viscosity in real units.
• Boundary conditions — one entry per patch. Each knows which fields it needs, so you cannot leave a field unset without the Problems tab saying so.
• Simulation control — how long it runs and how often it writes.

The Problems tab in the bottom dock is worth trusting. It reports what is wrong before you run, not after twenty minutes of solving.

4. Run it

Select Runs and press Run.

The solver starts and the bottom dock fills with:

• the log, streamed live,
• residuals, plotted per equation,
• monitors, if the case defines any.

Residuals falling steadily by several orders of magnitude is convergence. Flat residuals mean it has stopped improving. Rising residuals mean it is diverging, and the run will not recover on its own — stop it and see Troubleshooting.

You can keep working while it solves. You cannot pause a running solver — stop it, and it ends at the next write.

5. Post-process

Select Post-processing. Everything is in the application; there is no ParaView round trip.

• Surfaces, slices and contours for the field distribution
• Streamlines and vector glyphs for direction
• Probes and plot-over-line for numbers you can put in a document
• Boundary values for what is happening on each patch

Start with a slice through the middle of the domain coloured by velocity. If that looks physically wrong — flow going the wrong way, a wake on the upstream side — the setup is wrong and no amount of further post-processing will fix it.

6. Report

Select Report to produce an A4 PDF: results in engineering units, the split of drag into pressure and friction, run cost, and a verification checklist that states the threshold behind each verdict.

On the free tier the report is watermarked. Everything in it is real.

What to do next

The single most useful next step is to run the same case on a finer mesh. Add a second simulation, raise the mesh density, and run it. If the answer moves a lot, the first answer was mesh-dependent and was not an answer.

The Comparison node exists for exactly this: it puts two or more simulations side by side and reports whether the result has converged on the mesh, and whether it depends on the turbulence model. Those are the two questions somebody will ask in a review, and a single run cannot answer either.

• Tutorials — full worked cases, start to finish
• Concepts — why y+ and mesh independence matter
• Reference — every panel, every setting