Detta projekt behandlar metodutveckling för att modellera en trevägskatalysator som skulle kunna användas vidare för att uppskatta utsläppen från en Ottomotor. Modellen är byggd i AVL Cruise M för att bestämma omvandlingarna för de tre lagstiftade föroreningarna kolmonoxid, kväveoxider (NOx) och metan baserat på reaktioner som sker i trevägskatalysatorer. Projektet tar också upp olika experimentella metoder och specifika kalibreringsmetoder som används för att samla in data och bygga modellen. Projektet diskuterar två olika kalibreringsprocesser baserade på insamlade experimentella data och antalet reaktioner kalibrerade i varje steg. Dessutom diskuteras programvaran som används och ändringarna som gjorts i den fördefinierade modellen av programvaran. Resultatet efter kalibreringen visade att den andra kalibreringsprocessen gav bättre resultat, men vissa avvikelser observerades i den byggda modellen. Avvikelserna antas bero på följande tre anledningar. För det första, försummar ytreaktionsmodellen föroreningarnas diffusionshastigheter. Dessutom kan komplexiteten hos objektfunktionen som matas in i optimeraren samt de stationära data som används för kalibreringsändamål också ge avvikelser. Dessa tre argument testas i projektet och slutsatserna som dras för att kunna förbättra modellen är följande.  Objektfunktionen behöver förenklas så mycket som möjligt, diffusion av föroreningar genom washcoaten kan inkluderas i olika komplicerade steg och syrelagringsreaktioner blir viktiga när man tar hänsyn till transienta förhållanden. / This project discusses about development of a method to model the three-way catalyst which could be used further to estimate the emissions from an Otto engine. The model is built in the commercial software called AVL Cruise M to determine the conversions of three legislative pollutants namely, carbon monoxide, nitrogen oxide and methane based on reactions occurring within the three-way catalyst. The project also discusses about different experimental and calibration methods used for collecting the data and building the model. The project discusses two different calibration process based on the experimental data used and number of reactions calibrated on each step. Furthermore, the software used, and the modifications made in the predefined model of the software are also discussed. The result after the calibration showed that the second calibration process gave better results, but some deviations were observed in the model built. The deviations are assumed to be because of three arguments present. Firstly, considering the surface reaction model which neglects the diffusion rates of the species. Secondly, the complexity of the objective function fed into the optimiser. The optimiser is also a software by AVL named design explorer which helps to optimise the parameters. And thirdly, the use of steady state data only and not including transient conditions for the calibration purposes. These arguments are tested in the project, and it is concluded to improve the model, the objective function needs to be simplified as much as possible, diffusion of species through washcoat could be considered at advance stages and oxygen storage reactions become important when transient conditions are taken into consideration.

Development of a fuel economy model for light-duty and heavy-duty vehicles is part of the Texas Department of Transportation’s “Estimating Texas Motor Vehicle Operating Costs” project. A literature review for models that could be used to predict the fuel economy of light-duty and heavy-duty vehicles resulted in selection of coastdown coefficients to simulate the combined effects of aerodynamic drag and tire rolling resistance. For light-duty vehicles, advantage can be taken of the modeling data provided by the United States Environmental Protection Agency (EPA) for adjusting chassis dynamometers to allow accurate determination of emissions and fuel economy so that compliance with emissions standards and Corporate Average Fuel Economy (CAFE) regulations can be assessed. Initially, EPA provided vehicle-specific data that were relevant to a physics-based model of the forces at the tire-road interface. Due to some limitations of these model parameters, EPA now provides three vehicle-specific coefficients obtained from vehicle coastdown data. These coefficients can be related back to the original physics-based model of the forces at the tire-road interface, but not in a manner that allows the original modeling parameters to be extracted from the coastdown coefficients. Nevertheless, as long as the operation of a light-duty vehicle does not involve extreme acceleration or deceleration transients, the coefficients available from the EPA can be used to accurately predict fuel economy. Manufacturers of heavy-duty vehicles are not required to meet any sort of CAFE standards, and the engines used in heavy-duty vehicles, rather than the vehicles themselves, are tested (using an engine dynamometer) to determine compliance with emissions standards. Therefore, EPA provides no data that could be useful for predicting the fuel economy of heavy-duty vehicles. Therefore, it is necessary to perform heavyduty coastdown tests in order to predict fuel economy, and use these tests to develop vehicle-specific coefficients for the force at the tire-road interface. Given these coefficients, the fuel economy of a heavy-duty vehicle can be calculated for any driving schedule. The heavy-duty vehicle model developed for this project is limited to pre-2007 calendar year heavy-duty vehicles due to the adverse effects of emissions components that were necessary to comply with emissions standards that went into effect January 2007. / text