Air Intake Manifold Design 1D-3D fluid-dynamic multi-objective engine performances optimization

Magneti Marelli

CHALLENGE - The aim of this project is to define a methodology to optimize the performances of an intake manifold of a multi-cylinder internal combustion engine. To satisfy customer requests torque and power values must be maximized while pressure drop must be minimized considering runner to runner balance.

SOLUTION - The modeFRONTIER workflow combines a 1D analysis performed adopting GT-Power with a 3D CFD simulation done performing Ansys CFX. The same geometrical variables are used to modify the GT model and the parametric CAD geometry. To save computational effort, designs are first computed using the 1D solver, then, only the ones which respect a target value of power and torque are analyzed using the 3D CFD code.

BENEFITS - A multi-fidelity automatic optimization workflow has been implemented in modeFRONTIER using GT-Power and Ansys CFX. Results of a multi-objective optimization, run using the genetic MOGA II algorithm, show that the runner diameter for engine performance and the runner length for the fluid-dynamic performance are the most important parameters.