Engineers can now design microfluidic devices faster, no matter the shape

A new AI method predicts forces inside tiny fluid channels. It works for any channel shape, which means engineers no longer need to train a new model for each design. This speeds up the design of devices that sort or move tiny particles, like cells, for medical tests or industrial uses.

Designing microfluidic devices has been slow. Every time an engineer changed the channel shape, they had to build a new simulation model from scratch. This paper shows how to use one model for many shapes. This means engineers can test new designs much faster. It could accelerate the development of new medical diagnostic tools or industrial filters that sort cells or other tiny particles.