From the vault’s bluntest [reference] label: this material validates against an experiment, not a compiler. These projects are the cheapest available substitute — simulate it, then check the simulation against something physical you can measure.

Difficulty: 🟢 starter (hours–days) · 🟡 intermediate (a week or two) · 🔴 ambitious / portfolio-grade. ⭐ = highest signal.

Python + NumPy/SciPy + Matplotlib is the whole toolchain. Nothing here needs licensed software.

Projects

  • 🟢 Verify a beam by hand and by code — cantilever deflection analytically, then numerically. Done when: they agree to several digits, and you know which assumptions you made. Exercises: beams.

  • 🟡 ⭐ Write a 1D finite element solver — bar under axial load: shape functions, assembly, boundary conditions, solve. Done when: it matches the analytical solution, and refining the mesh converges. FEM stops being a black box the moment you assemble the stiffness matrix yourself. Exercises: FEM.

  • 🟡 2D heat or stress on a plate — extend the solver, or use FEniCS/scikit-fem, and validate against a known case. Done when: your answer matches a textbook benchmark.

  • 🔴 A CFD toy — lid-driven cavity flow by finite differences. Done when: your streamlines match the published benchmark at the same Reynolds number. Exercises: Navier-Stokes.

If you only do one

Write the 1D finite element solver. FEM stops being a black box the moment you assemble the stiffness matrix yourself.

↑ Back to Continuum Mechanics