VTAB: Vehicle-Level Simulations in the Development of Exhaust Aftertreatment Systems

Stricter global vehicle emissions regulations, such as Euro VI and upcoming Euro VII standards, are driving the development of increasingly efficient engines, in terms of both fuel consumption and emissions, and consequently of more and more advanced Exhaust Aftertreatment Systems (EATS) for heavy-duty vehicles. These systems are essential for reducing pollutants such as NOx and particulate matter, in order to meet legal limits and sustainability targets. However, EATS development and validation still require extensive testing campaigns in engine test cells and on-road vehicles, which is both time-consuming and costly. This raises the question: can simulation support the design and development process, making the testing phase more streamlined and efficient? This thesis, conducted at Scania CV AB, investigates the employment of VTAB (Virtual Truck And Bus), Scania’s in-house Software-in-the-Loop (SiL) platform based on the Functional Mock-up Unit (FMU) standard, to integrate and simulate an advanced EATS 1-D model, developed in AVL Cruise-M. VTAB enables virtual testing by connecting models of vehicle subsystems (engine, gearbox, ECUs, EATS…), allowing the reproduc-tion on a standard PC of realistic on-road driving and engine test bench scenarios. The main objectives were to assess the feasibility of using VTAB for exhaust system devel-opment and to establish a robust methodology for model integration, simulation execu-tion, post-processing, and data analysis. The methodology proposed was validated through the simulation of three representative scenarios: two homologation cycles performed in the test cell (WHTC and WHSC), and one on-road full-truck test. All simulations aimed to replicate physic tests executed by Scania’s engineers, using actual input datasets. Python scripts were developed in Visual Studio Code in order to automate the simulation process, ensuring repeatable and coher-ent runs, and enabling detailed analysis of exhaust temperature profiles, NOx emissions, and AdBlue dosing strategies. Additional stand-alone simulation in AVL Cruise-M was carried out to better isolate the EATS model from the rest of the virtual truck and assess its intrinsic behavior. Results show that VTAB, combined with the new EATS model, can reproduce real system dynamics with good accuracy when provided with correct input signals. However, dis-crepancies arose due to interactions with other truck components’ models. Overall, the work demonstrates that VTAB has strong potential to support EATS devel-opment, but its current maturity in terms of reliability and simulation accuracy is not yet sufficient for concrete engineering purposes. With further improvements of truck compo-nents’ models, including the EATS 1-D model utilized during the project, and simulation methodologies, VTAB could become a central tool in future EATS development work-flows.