Coding classes in West Dunbartonshire

The ship and model are invented. The ship is 100 metres long and should cruise at 15 knots, about 7.7 metres a second. The model is built at a scale of 1 to 25, so it is 4 metres long. How fast should it be towed? The waves a hull makes depend on the Froude number, speed divided by the square root of gravity times length. For the model to make the same pattern of waves as the ship, the two Froude numbers must match, which means the model runs at the ship's speed divided by the square root of 25: about 1.54 metres a second. The learner checks it in Python: both Froude numbers come out at 0.2464.

The catch is friction. Friction along the hull depends on a different quantity, the Reynolds number, speed times length divided by the water's viscosity. The model's Reynolds number is 120 times smaller than the ship's, and matching it would mean towing the 4 metre model at about 185 metres a second, which is impossible and would wreck the wave pattern anyway. So the model has proportionally about twice as much friction as the ship. Froude's solution, still the basis of practice, is to split the measured drag: estimate the friction part from a formula, scale only the remainder by the cube of the scale, and add back the ship's own, smaller friction.

With an invented measured drag of 30 newtons, friction accounts for about 16 of them. Splitting and rebuilding gives a ship resistance of about 348 kilonewtons. Scaling the whole 30 newtons by the cube of 25 instead gives 481, 38 per cent too high, which would mean specifying engines far bigger than needed. The learner also sees how the tow speed changes with scale: a 1 to 10 model runs at 2.44 metres a second, a 1 to 50 model at just 1.09.

Knowing what a model can and cannot tell you is the reason a West Dunbartonshire teenager should still learn to code in 2026. The longer argument is in why teenagers should still learn to code in 2026.