Presets
Playback
Parameters
Camera
Layers
Quality
Model check
The science
What you see. A figure skater on a Blender-built rink. In Stroking mode she pushes off for 0.4 s every second and cruises round an oval, leaning into the bends; in Gliding mode she coasts from the start speed until friction and air drag stop her; in Spin mode she turns on the spot and the arm slider changes her moment of inertia. Ice temperature, blade width, mass and push force change the friction.
Why ice is slippery. Pressure melting is not the reason: the melting point falls by dT = T (vice − vwater) p / Lf, only about 0.0074 K per atmosphere, so a blade at 2 MPa shifts it by 0.15 K. What matters is frictional heating. The shear of a thin water film generates q = η v² / h per unit area; this melts ice at ρi Lf h v / lc and is partly conducted into the cold ice, qc = k (Tm − Tice) / √(α lc/v). The balance is a quadratic in the film thickness h, with the closed-form root used in the page. The friction force is then F = η v A / h and μ = η v / (h p): colder ice and slower sliding give a thinner film and more friction, a thin blade edge raises the pressure p.
Motion. m dv/dt = Fpush − μ(v) m g − ½ ρ CdA v². In a bend of radius R the skater leans by tan θ = v² / (g R) so that the resultant force passes through her centre of mass. Spinning with no external torque conserves L = I ω; the arms are modelled as two uniform rods of 4 kg and 0.68 m hinged 0.19 m from the axis, so pulling them in raises the rate and the rotational energy (the arms do work against the centrifugal load).
Honest simplifications. The contact patch (10 cm by the edge width), the dry-friction cap, the thick-film limit of 10 µm, the push-force profile, the drag area and the body masses are illustrative values, not measurements of a particular skater or rink. Real ice also has a quasi-liquid surface layer, ploughing and a rough surface that are not modelled; the skater is a single point on the path with a rigid lean; edge grip in a turn is assumed to be sufficient, and the spin friction is ignored.
Checks. (1) the cruising speed from a force-balance root against a Runge-Kutta integration; (2) the pressure-melting slope against the textbook 0.0074 K/atm; (3) the film thickness against bisection of the unsimplified melt balance, plus the heat ledger; (4) the work done pulling the arms in against the rotational energy change.