Satheesh Makkapati

Modeling and control of a heated air intake homogeneous charge compression ignition (HCCI) engine

In this paper a statistical physics based mean value model (MVM) is introduced for a heated intake homogeneous charge compression ignition (HCCI) engine based on a previously developed crank angle resolved model and validated. In this HCCI engine, regulation of combustion timing is based on manipulating the cylinder charge temperature through the cold and hot throttle actuators. This two-input and single-output (TISO) system offers an input redundancy which is utilized to separately account for the fast fuel loading changes and the slower variations in the hot inlet (heated air) temperatures via appropriate actuator authority allocation. All the available experimental data sets suggest that the combustion duration Δθcomb between the crank angle of 10% and 90% fuel burned (θCA10 and θCA90) can be used as the combustion timing regulation variable. This choice is shown to guarantee both combustion stability and fuel efficiency in terms of covariance of indicated mean effective pressure (IMEP) and indicated specific fuel consumption (ISFC) respectively. Simulations of the linear controller with the nonlinear plant demonstrate the closed loop performance.

A Novel Way of Manipulating and Optimizing CAM Profiles for Internal Combustion Engines

Generating an optimum cam profile for an internal combustion engine application is usually a very involved process. This is because of the challenge faced by the engine engineers to perform several trade-offs between fuel economy, performance, durability and emissions. Typically, the trade offs are worked out using several computer codes and with the involvement of a cross-functional team of engineers. Hence, there is a need to develop a tool that can encapsulate the various computer codes and can manipulate the cam profile with ease. With the existence of such an automated tool, optimization of the cam profile can be achieved with a specified trade-off between the several metrics identified above. This paper describes the development of such a tool, and discusses the ingredients that make it flexible and computationally efficient. Results from the use of this tool are documented in this paper as well.Copyright