Echos & Kucher
Kucher
I was just reading about how the vibrations of a siege engine could carry through stone walls, how would you quantify that kind of resonance?
Echos
Sure, think of the wall as a big, heavy drum. Every part of it has a natural frequency, like the notes a tuning fork can hit. To quantify the resonance you’d first measure the wall’s modal frequencies and mode shapes with something like a laser vibrometer or a small impact hammer. Then you calculate the transfer function from the point of impact to the point of interest—basically how much vibration amplitude gets through at each frequency. The Q factor tells you how sharp the peak is; a high Q means the vibration will linger and potentially amplify. By comparing the engine’s vibration spectrum to the wall’s modal peaks, you can see where they line up and predict the resonance. Simple, but precise.
Kucher
That method is sound, but don’t forget the mortar’s damping effect; medieval walls were far from uniform, so the modal spectrum will shift with every joint and crack. In practice, you’ll find the resonance peaks are broader than a modern drum. Also, the siege engine’s vibration spectrum is usually broadband—just matching one peak isn’t enough to predict damage. You need to map the entire frequency response and then see where the energy piles up.
Echos
You’re right—those mortar joints are the wall’s own little dampers, so the peaks are more like smeared notes. That’s why we usually build a full‑frequency response model, either from measured data or a finite‑element simulation, and then run the engine’s broadband spectrum through it. The overlap gives us a total energy map, and the spikes in that map are what you’ll watch for to predict where the wall might give way.
Kucher
That’s the correct approach, but remember that the engine’s own harmonics can shift when the siege weapon is loaded. If you ignore that, you’ll miss a critical overlap that can trigger a failure. I’d double‑check the load‑induced frequency shift before finalizing the model.
Echos
Good point—adding that charge changes the engine’s own spring constants, so its harmonics drift. I’d run a quick vibration test with the load in place before locking the model. It’s the only way to make sure the overlap isn’t just an artifact of a lighter frame.
Kucher
Indeed the charge will alter the stiffness, so you must measure the engine with the load in place before you trust the model. It’s the only way to avoid a false overlap.
Echos
Exactly—measure it first, then model. No clever shortcuts, just data.
Kucher
Agreed. Numbers speak louder than theory. Let the data guide the model, and we’ll avoid the pitfalls of guesswork.