Designing a powertrain system to meet drivability, fuel economy and emissions performance requirements is a complicated task. There are many tradeoffs to be analyzed in terms of which components to use, such as lean burn technology versus classical components, characteristics of individual components, such as size or temperature operating range, and the control policies to be employed. In addition, there are performance tradeoffs to be analyzed, such as emissions level versus fuel economy. In the past, most of the powertrain design was made on the basis of hardware, that is, on the basis of laboriously assembling and testing many possible system configurations. Today, the time-line for vehicle design is constantly shrinking, the number of possible powertrain configurations is expanding, and the cost of doing hardware evaluations is growing. It is simply no longer feasible to make all (or even most) of the design decisions on the basis of hardware alone. More and more of the decisions must be made upon the basis of mathematical models and analysis.
This talk will describe the use of modeling and control techniques to assist in making powertrain design decisions on the basis of models and up-front control system analysis. The specific technology configuration analyzed here involves a gasoline direct injection (GDI) engine capable of stratified operation, in series with a three way catalyst (TWC) and a lean NOx trap (LNT). One of the truly novel features of a powertrain system that includes a lean NOx trap is that there does not exist a notion of a steady-state operating point. An LNT is fundamentally a dynamic device: it fills with NOx and must be emptied or ``purged'' periodically. Indeed, if the system is run continuously at a non-rich air-fuel ratio, the trap will saturate, its trapping efficiency will approach zero, and its NOx conversion efficiency will approach 20\%. This value is too low to meet emissions requirements, and thus the trap must be cycled.
The consequence for analysis and control is that a dynamic model of the emissions after treatment system is a necessity, and not a luxury. Why is this so different from a standard emission system based on a TWC? A three-way catalyst is, of course, a dynamic device. However, it has the property that as long as the feedgas remains at stoichiometry, and the TWC remains sufficiently warm, the conversion efficiencies do not significantly change with variations in mass air flow, engine speed, load, etc. As a result, many aspects of dynamic characteristics of a TWC-based emissions system can be ignored in the initial design stages of a powertrain, and most of the emphasis placed on the dynamics of the engine. For a system with an LNT, we will see that the situation is essentially reversed: the dominant dynamics are due to the emissions system.