Incorporating Design Variation into a 1-D Analytical Model of a 4.6L-4V Ford Engine for Improving Performance Projections

One-dimensional simulation tools are used extensively in the automotive industry to improve and optimize engine design for WOT performance. They are useful in target setting and in assessing the effects of certain design changes (e.g. intake manifold, valve timing, exhaust manifold, etc.). Generally the inputs to these models are “nominal” values or curves from a particular set of data and, therefore, do not take into account design or assembly variations. Often times, performance expectations are not met due to these “real world” effects and may result in significant re-design and testing efforts.

The purpose of this paper is to assess the impact of typical model input variation on engine performance and to instill greater confidence in the use of these models in forecasting performance. The approach taken is to collect, analyze, and categorize actual build measurements from a 4.6L 4V Ford engine that are considered important inputs for a one-dimensional modeling. From these inputs Monte Carlo simulations are run through the one-dimensional model to produce a range of performance output. The output is then analyzed via statistical methods to assess the impact of each input on engine performance variance.