[DoofusOfDeath]

Now I'm curious how, hundreds of years ago, engineers and architects determined what designs were viable.

Did they have quantitative data about materials' tensile strength, shear strength, etc.? If yes, what kind of math did they have for applying those data? If no, how did they safely design first-of-a-kind structures?

[skybrian]

This is at least partially survivor bias. Sometimes they miscalculated and buildings fell down.

[tssva]

Even the ones still standing often aren't because of fantastic original engineering but because they are the ones people have cared to carefully maintain and to perform renovations on to compensate for original engineering flaws.

[bdamm]

And sometimes simply by sheer mistake; the leaning tower of Pisa was supposed to be supported by the Corinthian columns and limestone internal blocks, but it turns out that its structure is actually supported significantly by the marble façade.

[geowwy]

An example is the Hagia Sophia which had a lot of reinforcements added after the original design was found to be unstable. The building you see today is not how it was originally designed to look, not just because of the minarets.

[philwelch]

So did we.

[sonthonax]

Usually very old buildings (at least the ones that have survived) are extremely overbuilt, with solid walls many times thicker than what we would use today.

[hprotagonist]

“safety factor: even more rock, please!”

[mjw1007]

As I understand it, they tried to to make sure pretty much everything was under compression, and tested using scale models (which work well in that case).

[vasco]

Cut some pieces of wood and test out smaller scale models

[jacquesm]

Wood and rock have such dissimilar structural properties that scale models out of wood don't say anything useful about full sized buildings out of rock.