Topology optimization of an engine piston to reduce particulate emission during cold start operation 2019-01-0835
The vast majority of engine out particulate emissions are released in the first several minutes of cold start operation, in large part due to low piston surface temperatures which fall well below the boiling point of the injected fuel. Unfortunately while topology optimization based on heat transfer analysis is a promising approach to help solve this challenge, existing works have focused largely on basic small-scale canonical scenarios in the steady-state. In this work an algorithm was developed and demonstrated computationally which is aimed at optimizing the internal structure of an engine piston to increase piston surface temperatures during the early phases of engine cold start, while simultaneously limiting overall peak temperatures during hot steady-state conditions. A finite difference heat transfer model of a generic engine piston geometry was created and a genetic optimization algorithm in conjunction with a Lagrange Multiplier Method (LMM) were used to develop optimal piston topologies. Various optimized designs were generated and common geometric traits were identified, providing to critical insight for future piston designs. The relationship between average piston surface temperature after one minute of operation from cold start v. peak piston temperature during hot steady-state operation was quantified, illustrating a clear trade-off between rapid warming and material limits. Various search algorithms were also evaluated, revealing the most effective and computationally efficient approaches. Next steps to this work will include integration of mechanical stress into both the model and the constrained optimization algorithm to further enable autonomous physics-based engine piston design optimization.