Inspiration

This project started from our love for software and physics, and how they can be used to solve real-world problems. We’ve always been fascinated by how equations describe motion and forces, but we wanted to see how that knowledge could actually save lives. After reading about the 1985 and 2017 earthquakes in Mexico, we were shocked at how many buildings collapsed even though the technology exists to predict their behavior. That’s what inspired us, the idea that we could use simulation and data to prevent that kind of destruction before it happens.

What We Learned

Before this, we didn’t fully understand how buildings react during earthquakes beyond just “they shake.” We learned how engineers actually model that shaking using equations of motion.

Each part of this equation describes something real, mass for each floor, damping for energy loss, and nonlinear springs that show how the building bends or breaks. Through this, we learned how stiffness, damping, and ductility all interact. We also learned how to translate physical ideas into code, which was honestly the hardest but most rewarding part.

How We Built It

We started by pulling earthquake data from the USGS API, which gave us information like magnitude, depth, and coordinates. We used that data to generate how the ground would actually move during an earthquake. Then we created a building model made up of floors connected by nonlinear springs and dampers. Each floor reacts differently depending on the material, steel bends more, masonry cracks earlier. We also added Rayleigh damping, which makes the shaking fade naturally instead of going on forever. Finally, we used the Newmark-β method to solve how each floor moves through time and then passed those results into an AI model that gives suggestions to make the building safer, for example, adding damping or changing materials.

Challenges

The biggest challenge was making the simulation realistic while keeping it stable. If one floor collapsed, we had to make sure the whole building responded logically — not explode into errors. Another challenge was tuning damping and stiffness, so it didn’t behave randomly or unrealistically. Also, connecting all the physics, math, and code together took a lot of trial and error. But that process taught us more than anything else, especially how theoretical equations can actually drive real engineering tools.

Reflection

This project made us realize how powerful applied math and coding can be when they’re used for something meaningful. We started this because we love problem-solving, but we finished it feeling hopeful, hopeful that one day, better simulation tools can help create safer buildings and save lives in earthquakes around the world.

Built With

Share this project:

Updates