Search Results

Internal combustion engines continue to grow more complex every day out of necessity. Legislation and increasing customer demand means that advanced technologies like variable valve actuation (VVA), multi-path exhaust gas recirculation (EGR), advanced boosting, and aftertreatment systems continue to drive ever-expanding requirements for engine control to improve performance, fuel economy, and reduce emissions. Therefore, controller development and implementation are becoming more costly, both in terms of time and the monetary investment in engine hardware. To help reduce these costs, a sophisticated tool chain has been created which allows a real-time, physical, crank-angle resolved one-dimensional (1D) engine model to be implemented on a rapid prototyping engine control unit (ECU) which is then used in the control strategy of a running engine. Model-based controllers have been developed and validated to perform as well as or better than controllers using traditional sensors.

Acoustics of automotive intake and exhaust systems have been modelled very successfully for many years using 1D gas dynamic simulations. These use pseudo 3D models to allow complex components to be constructed from simple building blocks. In recent years, tools have appeared that automate the construction of network models from 3D geometries of intake and exhaust components. Using these tools, concurrent noise and performance predictions are a core part of most engine development programmes. However, there is still much interest in the more traditional field of linear acoustics: analysing the acoustic behaviour of isolated components or predicting radiated noise using a linear source. Existing approaches break the intake and exhaust system down into a set of components, each with known acoustic properties. They are then connected together to create a network that replicates the donor non-linear model.