Where does the heat come from when rubber is stretched adiabatically? As I understand when a rubber band is stretched adiabatically (which I am assuming means no change in entropy and so no heat flow from surroundings) its polymers naturally are straightened and thus their entropy would be reduced, so in order to keep entropy constant, the rubber band's temperature must increase. So how exactly does the work done on the rubber band become manifested as randomised thermal energy?
My initial guess is that the polymers 'lining up better' equates to them being in a lower potential energy state and so kinetic energy (randomised into thermal energy) is released in this process but I am not sure. Or possibly the work done just transfers kinetic energy directly to the particles in the rubber?
 A: Rubber is pretty damn complex!
On a macroscopic level, you do work to the rubber and it heats. When the rubber does work back, it cools. Just like when you compress a gas and then the gas expands.
The microscopic situation with the gas is simple (assuming you just play with a piston in a cylinder of gas) : the molecules bounce from a moving wall and either loose some energy (when the piston goes away) or gain energy (when the piston compresses the gas).
Rubber has long entangled and mostly spiral molecules and their thermal motion keeps them in some equilibrium shape. When you stretch the rubber, you tend to unwind some molecular spirals and impose more tension over the chemical bonds. This increases their vibrational energy and this energy is quickly "thermalized" over the whole material.
A (very rough) analogy: a rope, hanged more or less horizontally between two points. You quickly move one of these points to straighten the rope and it starts to oscilate.
A: When the rubber band is stretched it heats up because of the viscous friction of the chain molecules slides on each other. On the other hand, it is not because of adiabatic expansion because there is no expansion in the rubber band and it is an incompressible material (it deforms on stretching but volume change is 0).
