We applied computational fluid dynamics (CFD) and AI-enabled simulation to optimise a Y-branch pipe, reducing head loss and limiting waste deposition.

The baseline geometry and several design variants were assessed across realistic flow rates and fluid properties. Variables included the branch angle, blend radius, crotch curvature and leg offsets.

We mapped energy-loss mechanisms, flow separation and secondary flows to understand how local design features affect pressure drop. This also helped identify low-shear areas where solid waste can accumulate.

Using the CFD design space, we built a fast-running surrogate model to guide the search towards high-performing designs. The model used multiple criteria, aiming to minimise the loss coefficient while reducing recirculation and wall-shear deficits.

Mesh and solver settings were refined to provide reliable results within a practical timeframe. Verification checks confirmed that comparative trends remained stable across the different options.

Outputs included loss-coefficient breakdowns, velocity and wall-shear maps, and indicators of deposition risk. Manufacturing constraints were also considered throughout the assessment.

The selected concept provided a clear and manufacturable route to reducing energy consumption and improving cleanliness. We also supplied implementation guidance and a repeatable workflow for future design iterations or changes in scale.