Materials & Mechanical Engineering | UC Santa Barbara
Our computational framework generates tens of thousands of physics-based simulations daily, transforming how we understand material behavior. By embracing statistical distributions rather than idealized cases, we reveal design spaces and processing windows invisible to traditional approaches. This massive data generation enables AI-driven materials development and fundamentally changes the questions we can ask.
Your material is statistical.
Your one test? It's just one pixel in this map.
Ti-6Al-4V lattice structures with manufacturing variability. della Ventura et al. (2024)
Processing targets are distributions, not outliers.
Defects aren't failures. They're distributions to control.
Thermal barrier coating systems with controlled porosity. Sehr et al. (2023)
Fast enough to be useful.
1000× faster in nonlinear regimes.
Ceramic matrix composite optimization with thermal cycling. Cox et al. (2014)