Statistical Approach to Diesel Aftertreatment Accelerated Aging Performance Correlation to In-Use Population
Holmes Ahari, FCA
Accelerated aging of an aftertreatment system has been a common practice in gasoline application for decades. A well correlated aging behavior of Three Way Catalyst (TWC) to the Arrhenius relationship of time, temperature and activation energy has allowed OEMs to age catalysts at high temperatures for a much shorter period of time to mimic their behavior at full useful life (FUL) levels. These parts have historically been used for certification and OBD. For diesel applications however, this type of accelerated aging is not currently allowed by the government agencies. This is mainly due to the facts that not all components of diesel aftertreatment follow the Arrhenius behavior and because diesel accelerated aging suffers from lack of maturity and in-depth understanding for it to be allowed for use in certification. However, OEMs do use accelerated aging to create catalytic components used in developmental activities for calibration and OBD detection.
In cases where the accelerated aging procedure creates a system that is less aged at a certain mileage compared to a system aged on the road to that mileage, the system is considered under engineered and will create calibration miss match at high mileages. Ideally, it is better for an accelerated aged system used in development to be aged more severely than what is expected of a system aged on the road.
Due to infinite driving variation that could exist, one cannot proceed with high level of confidence that the accelerated aging methodology will mimic the in-use performance. In this study we set to establish a statistical approach where an aging methodology can be correlated to the behavior of a population of in-use vehicles with high level of confidence. It allows the OEMs to establish the level of coverage their accelerated aged systems will have over the in-use population and whether this coverage meets their internal coverage target. In cases where the coverage target is not met, modification to the aging procedure will allow improvement to population coverage.
Furthermore, the methodology is capable of establishing the coverage of any functionality from a limited test population to that of a general population. Thus, the approach was used to determine thermal behavior coverage from multiple in-house durability test vehicles running specific cycles, and correlating it to that of an in-use population. This enabled us to identify and overcome thermal behavior shortcomings by adjusting regeneration frequency and temperature and to better mimic that of customer usage and vehicle population.