

Scientists thought a smooth surface reduces drag best, by stabilising laminar flow. They now find that a little bit of random roughness is better. Apparently it significantly increases the stable range of laminar flow. In this range drag is reduced by 40 %. Also above this range drag is also lower.














I think the problem is the surface of ‘the ground’. In reality the ground stands still and the car moves relative to it. In the wind tunnel both would stand still and only the air is moving. I think this totally changes the situation of the experiment.
Also the ground would still introduced turbulence close to the studied object, which could also make it difficult to get good results.