Friction and the generation of heat from mechanical work

In Über die Erhaltung der Kraft (1847), Helmholtz placed friction both at the surfaces of two bodies moving over each other and inside a body as it changes form. He added that thermal and electrical changes always accompany it.¹ He disputed the usual mechanical treatment, which presents friction as a force against motion whose strength depends on velocity. He called this an incomplete expression of a complicated process.² On that view living force was lost absolutely, and the same was assumed for elastic impact.² Helmholtz argued that the heat gained, and usually the electricity produced, can themselves yield mechanical effects.

Bose (1902) used friction to shape response records in a torsional pendulum struck by a blow. He varied the friction and reported that less friction meant greater mobility, with the form of the curves changing while the stimulus stayed constant.³ With more friction the response height fell and recovery was much prolonged, which he called partial molecular arrest.⁴ He compared this curve to the one produced on living tissue or metals by small doses of chemical poisons.⁴ The comparison links a mechanical quantity to the behaviour of living and non-living matter.

Leduc (1911) noted that a normal sugar solution has an osmotic pressure of 22.35 atmospheres, and asked why such an enormous pressure is not more evident. He pointed to frictional resistance to diffusion, which grows as the solute divides into molecules and its surface increases.⁵ Bayliss (1915) addressed the claim that a catalyst can start a reaction held back by chemical friction. In his model, friction between the weight and the glass may be so great that no movement appears until oil is applied. He added that the weight may really move but be arrested again. He judged the question to be largely one of words.⁶ Hill (1926) found large irreversible heat when a muscle is stretched or released. He attributed it to frictional transformation of mechanical energy in a viscous system.⁷

Friston and Ao (2012) treat friction inside a formal account of dynamics. They state that friction is an emergent property of random fluctuations in velocity and nothing more.⁸ Kriegman, Cheney and Bongard (2018) use friction as a selective pressure in simulated soft robots. They report that natural selection can climb a path of increasing speed by growing legs that reduce the surface area touching the floor, and thus friction.⁹ In these works friction enters the models, though Friston and Ao call it emergent.