3 ms·

The article does address that somewhat: > There won’t be a single perfect solution to the problem of concrete. One potential advancement is green concrete, whi

The article does address that somewhat: > There won’t be a single perfect solution to the problem of concrete. One potential advancement is green concrete, which is decarbonized by changing the recipe, production process or longevity. Another is to capture emitted carbon at cement plants and store or reuse it. Scroll down to the bit about ETH Zürich's research and there's some interesting stuff beyond the long opening about history. > Block’s early research was focused on heritage architecture — vaulted cathedrals and other historic structures. “Our modern engineering tools are quite inadequate in explaining how safe these buildings are,” he said. Sometimes, he went on, digital engineering models of old buildings concluded that there should be no way they could still be standing, that they should’ve collapsed centuries ago. But there they were, solid, unmoving. Maybe, Block thought, there was something wrong with the models.

adrian_b 4y ago
Any kind of physical model that is computable must make huge simplifications by ignoring most characteristics of the modeled system. Therefore whoever creates the model must choose a priori which features should be included in the model, because it is believed that they influence the results. and which features should be ignored, because it is believed that they have negligible influence. Any kind of physical model must be validated by comparison with experiments. It is extremely frequent to discover that the model does not match the experiments, which means that the a priori judgments about what can be neglected have been wrong. In such cases, the assumptions on which the model is based must be reviewed and some of the decisions about what features should be included in the model should be reversed. This is the normal way of working with any models, so the fact that the first attempt of modeling the ancient vaults has failed, does not say much, except that those who had made that initial attempt either did not understand well the behavior of certain rocks under compressive stresses, or they did not understand well how the building techniques used by the old masters redistributed the stresses in the rocks composing the vaults. The article does not explain which was the problem with those wrong models, but then it mentions that the team of Block has applied the traditional method of stone masonry, of making structures which are under compressive stresses, instead of bending stresses, like in a flat floor or ceiling, to design floors that need much less concrete than the conventional floors. This seems like a very obvious method, which should have already been used if good, so I wonder if the comparison about the concrete consumption was made at equal usable interior spaces. The main reason why flat floors and ceilings have been preferred in modern multi-level buildings was to provide interior spaces that can be completely used, without being partially blocked by structural elements (e.g. if the ceiling is a vault, you cannot put at a wall a cabinet raising up to the ceiling, but only up to the height where the curved part of the vault joins the vertical walls; if the vault has ribs, those impose additional constraints).

adrian_b 4y ago

Any kind of physical model that is computable must make huge simplifications by ignoring most characteristics of the modeled system. Therefore whoever creates the model must choose a priori which features should be included in the model, because it is believed that they influence the results. and which features should be ignored, because it is believed that they have negligible influence. Any kind of physical model must be validated by comparison with experiments. It is extremely frequent to discover that the model does not match the experiments, which means that the a priori judgments about what can be neglected have been wrong. In such cases, the assumptions on which the model is based must be reviewed and some of the decisions about what features should be included in the model should be reversed. This is the normal way of working with any models, so the fact that the first attempt of modeling the ancient vaults has failed, does not say much, except that those who had made that initial attempt either did not understand well the behavior of certain rocks under compressive stresses, or they did not understand well how the building techniques used by the old masters redistributed the stresses in the rocks composing the vaults. The article does not explain which was the problem with those wrong models, but then it mentions that the team of Block has applied the traditional method of stone masonry, of making structures which are under compressive stresses, instead of bending stresses, like in a flat floor or ceiling, to design floors that need much less concrete than the conventional floors. This seems like a very obvious method, which should have already been used if good, so I wonder if the comparison about the concrete consumption was made at equal usable interior spaces. The main reason why flat floors and ceilings have been preferred in modern multi-level buildings was to provide interior spaces that can be completely used, without being partially blocked by structural elements (e.g. if the ceiling is a vault, you cannot put at a wall a cabinet raising up to the ceiling, but only up to the height where the curved part of the vault joins the vertical walls; if the vault has ribs, those impose additional constraints).