Draft. This manual is new and still growing. If something here does not match what you see in the software, the software is right — tell us and we will fix the page.
Tutorials
Start to finish, one geometry, every click. A tutorial is for somebody who does not yet know what questions to ask, so it makes every decision for them.
Every tutorial must run verbatim, or say in its first paragraph which tier it needs. One that hits the 250,000-cell free cap halfway through is worse than no tutorial at all.
Rated like SimScale rates theirs, because it is genuinely useful and costs nothing: Easy needs no CFD background beyond knowing what you are looking at; Moderate assumes you have meshed something before; Advanced assumes you can read a residual plot and argue about turbulence models.
| Tutorial | Level | Teaches | Tier |
|---|---|---|---|
| Flow over a cylinder | Easy | The whole workflow on a case that meshes in seconds. Vortex shedding is visible, so a wrong answer is obvious | Free |
| Flow in a pipe | Easy | Internal flow, wall functions, y+ | Free |
| External aerodynamics on a body | Moderate | Refinement regions, boundary layers, drag decomposition | Free to follow; a mesh worth defending wants more cells than the free cap |
| Heat sink, conjugate heat transfer | Moderate | Multi-region, solid and fluid together | Free, within the 250,000-cell cap |
| Dam break, free surface | Moderate | VoF, transient control, why gravity is not optional | Free |
| Mesh independence study | Moderate | Three simulations, one case, the Comparison node | Free |
| Rotating machinery with MRF | Advanced | Cell zones, rotating frames | Free, within the 250,000-cell cap |
On the tier column
The free tier is limited by scale, not by physics. Every analysis type, every turbulence model, every boundary condition and every cell zone is available without a licence. What free does not do is run big, run wide, or run somewhere else: the mesh is capped at 250,000 cells, the solver gets one core, remote and headless execution are off, and the report carries a watermark.
So all seven tutorials can be built and solved on the free tier. What differs is how much mesh is left over afterwards, which is what the Tier column above is recording.
The cell cap is reported two ways: as a row on the Problems tab before a run, and as a banner on the Mesh pane directly under the estimate that is too big. Before a mesh exists the row is a warning against the estimate rather than an error, because the estimate is a projection and snappyHexMesh routinely lands under it; once the mesh exists it is judged on the real count.
The core limit is not a warning at all. The Cores list on Simulation control is shortened to what the licence allows, so on free it is one entry long and the hint on it reads "This licence solves on one core. The run still completes; it takes longer."
Reading order
Start with the cylinder. It is the shortest case that produces an answer somebody can check by eye — if the wake is not shedding, something is wrong, and no plot is needed to see it. That property is worth more in a first tutorial than coverage.
Then the mesh independence study, which reuses the cylinder case and turns its answer into a result. Everything after that can be read in any order.