The effect of earthquakes on a building can be investigated by creating a scaled down model of the building, and simulate an earthquake on a shake table.
A recording of a natural earthquake can be replayed on the shake table. How is the spacial scaling related to the progression in time?
A model in a different size has clearly different resonances, so replaying an earthquake in the original speed can not have the same effect.
Is it possible to scale the time in a way that would recreate an equivalent effect, assuming the mechanical properties of the model are representative of the original?
If yes, does that work robustly when the material and mechanical properties of the model can not exactly represent the original?
The problem seems difficult to me because minor differences in material properties can change resonance frequencies, which means small changes can have strong if not drastic effects.
I assume a model on a scale of 1:20 can not even approximate the material properties in a useful way, so I imagine a model scale of 1:4 or so. But it may already be hard to represent a single reinforced concrete beam on scale 1:4 with equivalent resonance frequencies.
