Gough-Joule effect in rubber band When a stretched rubber band is heated, surprisingly, its tension will increase. What is the physics behind this?! 
Another thing. The Entropic spring model says, for a good specimen, that the internal energy shall not change if the material is stretched. Why is this so? I mean, what is the mechanism behind storing internal energy in such a material?   
 A: Rubber consists of many long-chain polymers. In an unstressed sample, these are randomly arranged. As a mental model think of them as anchor chains, where the angle between each link is entirely random - the overall polymer is in essence a random walk through the medium.  
Now, you pull on it.  The net result is to better align the backbone of the polymer with the direction of tension.  But, this alignment reduces the entropy of the system - the relative orientation from link-to-link is more defined, so you are reducing the available phase space for the polymer chain.  
Now you heat the material, sufficient for the chains to start trying to move around. To gain back their lost entropy, they will try to become more random, or curl back up again. It will do so the the extent that the increase in entropy (by returning to a more random configuration) can offset the increase in enthalpy from the increasing tension (kind of like PV work in a gas, but stress-strain work in a solid).
Overall, the free energy can remain the same, you are just trading enthalpy for entropy.
