Aging in nonequilibrium systems

Aging in nonequilibrium systems means that the measured response or correlation depends on how long the system has been evolving since preparation, not only on the interval between measurements. A two-time correlation $C(t_w+t,t_w)$ depends separately on waiting time $t_w$ when time-translation invariance has not been established.

One probing protocol quenches a disordered magnet, waits different times and then applies a small field. Comparing the resulting responses distinguishes preparation history from a single equilibrium relaxation time. Analytical spin-glass models exhibit such nonequilibrium correlation and response behavior; simple scaling with $t/t_w$ is a model-dependent form, not the definition of aging.^[ref]