The cooling rate determines the amount of free volume in most glass-forming systems. In silicate glasses and in polymers (which I am more familiar with) due to the igh viscosity it is realtively easy to play with cooling rate as the non-equilibrium free volume can be stabilized. (Even though it is only a meta-stable state exhibiting slow relaxations - see phsical aging, volume relaxation, enthalpy relaxation). In the case of metallic glasses, due to the low viscosity one has to apply very fast cooling, neverthless the same phenomena can be observed. There is a huge amount of literature dealing with these problems.
Hi, The glass transition temperature (Tg) is strongly dependent on the cooling rate. When cooling process occurs quickly, the thermal motion of molecules are blocked, they couldn't be completed their motion and couldn't reach to close position as near as each other. Consequently, the free volume between the molecules will be higher than compared to the low cooling rate. Therefore, when this material exposed to heat, the molecules begin their heat-induced movement and decrease the free volume, due to their tendency to achieve close together. this behavior cause to decrease the specific volume of material and increase the density of that. Best Wish,
Free volumes are released to the surface or conglomeration to larger defects called pseudodyslociation dipoles. This is the normal behavior of free volumes related to thermal dependencies.