Improvements to 1D engine modeling accuracy and computational speed have led to greater reliance on this simulation technology during the engine development process. The benefits of modeling show up in many ways: increased simulation iterations for better optimization, reduction in prototype hardware iterations, reduction in program timing and overall cost.
In this study a 1D GT-Power model of a turbocharged engine system was used to assist in the initial design phase and throughout the program. The model was developed using Chrysler Group LLC proprietary modeling features for predictive combustion and knock event prediction. In all stages of this project the model's accuracy was improved through regular correlation with dynamometer data.
This paper mainly focuses on engine compression ratio selection, turbocharger selection, and cycle-to-cycle variation/cylinder-to-cylinder variation reduction through the combination of 1D GT-Power model optimization and dynamometer tests. In the case of cycle-to-cycle variation/cylinder-to-cylinder variation it is found that mixture of RGF, CEGR, air and fuel (PFI only) back-flowing from intake and exhaust valves is re-distributed into each cylinder at current cycle and the following cycle. Consequently the back flow through intake valve and exhaust valve is one of the main factors generating these variations in each cylinder. Based upon this understanding, a new set of valve lift profiles is designed through 1D GT-Power model optimization, together with input from valve design engineer, which shows that the approach used yields benefits in fuel consumption, EGR tolerance, and engine stability and engine performance enhancement.