Founded by Prof. Luís Sá in 2023, the DO&GE Lab tackles multidisciplinary problems — structural, thermal and fluid flow — by combining parametric optimization, topology optimization and machine learning. The goal is simple to state and hard to reach: devices with maximum performance, efficiency and sustainability.
Our work has produced 33+ peer-reviewed papers and 440+ citations, in collaboration with Imperial College London, the Technical University of Denmark (DTU) and the National Laboratory for Scientific Computing (LNCC).
Approach
Algorithms redistribute material or tune parameters to reach highly efficient geometries under structural, thermal or fluid constraints — SIMP, TOBS (binary structures), topological derivatives and level-set methods.
Neural networks explore vast design spaces beyond human intuition. From boundary conditions and a few past designs, they propose novel solutions that balance competing goals — stiffness vs. weight, pressure drop vs. flow rate.
Surrogate models trained on historical designs predict performance almost instantly and accelerate convergence — especially for expensive problems like heat exchangers and complex flow devices.
Research lines
Our flagship line — from microfluidic channels to industrial turbomachinery, using integer linear programming, topological derivatives and continuous boundary propagation.
Redesigning flow channels of proton-exchange-membrane and solid-oxide fuel cells.
Pioneering formulations for subsonic compressible and rotating turbulent flows.
Rotors, Tesla-type pumps and labyrinth seals designed by topology optimization — and validated on the bench.
Optimization and generative design applied to medical devices.
People
Researchers and students currently supervised or co-supervised by Prof. Sá.
Since 2024
Topology optimization method applied to the design of compressor impellers for supercritical CO₂ considering resonance frequency and vorticity local constraints
Simulation and Optimization of Solid Oxide Fuel Cells
Experimental Study of Labyrinth Seal Designed by Topology Optimization
Thesis ↗Join the lab
Pick the path that fits you. If you're unsure, write anyway — we'll find the problem that fits.
For Mechanical, Mechatronics and related students who want hands-on research. No prior optimization experience needed — curiosity and commitment are what matter.
Positions in every research line. You'll develop original research, publish in international journals and present at conferences.
Openings arise periodically, usually tied to funded projects (FAPESP, Shell/RCGI).
