FPV — CAD and simulation
#fpv/cad · Note 3 of 7 · part of Drones
Two different things get called "simulation" in this hobby and it is worth separating them before choosing a tool:
- Flight simulation — learning to fly. That is FPV — Simulators, and the answer is a game, not a CAD package.
- Design simulation — will this part survive a crash. That is CAD and FEA, and it is what this note is about.
On AutoCAD specifically
Worth being straight about it: AutoCAD is not the right tool here. It is drafting software, strongest in 2D and in architectural/engineering documentation, and it is not free. It has a free one-year student licence, which is legitimate if you qualify, but even then it is an awkward fit for organic 3D parts like a camera mount.
If you want to model drone parts, these are better and genuinely free:
| Tool | Cost | Good for |
|---|---|---|
| FreeCAD | Free, open source | Parametric parts, and it has an FEM workbench for stress analysis |
| Onshape | Free for public documents | Browser-based parametric CAD, excellent constraints, nothing to install |
| Fusion 360 | Free personal licence | The most pleasant of the three; watch the personal-use restrictions |
| Tinkercad | Free | Ten minutes to a printable antenna mount. Not parametric, but fast |
| Blender | Free, open source | Organic shapes and renders; not a CAD tool, don't use it for tolerances |
Onshape is the one I would start with — no install, no licence anxiety, and the constraint solver teaches you to model properly.
What is actually worth modelling
Not the frame. Carbon frames are cheap, tested, and you cannot cut carbon at home safely without proper extraction — the dust is genuinely hazardous.
What you print:
- TPU camera mounts and guards — the part that dies most often
- Antenna mounts — geometry matters here, keep the antenna clear of carbon
- Battery pads and straps' anti-slip
- GoPro / naked-cam mounts at whatever angle you actually fly
- Motor soft-mounts and washers
Print TPU (95A) for anything that should absorb a crash, PETG for rigid brackets. PLA is fine for prototypes and useless in a hot car.
Stress analysis, honestly
You can run FEA on an arm in FreeCAD's FEM workbench or SimScale's free community tier, and it is a good way to learn where stress concentrates — usually at the motor mount holes and the arm root.
But: crash loads are impact loads, they are enormously variable, and modelling them properly is real engineering. Treat FEA here as insight, not proof. The empirical test is cheaper and more honest — fly it, crash it, see what broke.
CFD for prop wash is firmly in the "interesting, not useful" category for a hobby build. Skip it.
The workflow that works
- Measure the real part with calipers. Twice.
- Model it parametrically so the mount hole spacing is a variable, not a number.
- Print one. Fit it. It will be wrong.
- Change the variable, print again.
Keeping the model parametric is the whole benefit — see FPV — Claude Code in the loop for keeping the dimensions and the build log in one place.
← Previous: FPV — Simulators · → Next: FPV — What to buy · Series: Drones