Future CO2 emission legislations require the internal combustion engine to become even
more effcient. Of great signifi cance is the boosting system enabling downsizing and downspeeding.
The thermodynamic coupling of a reciprocating internal combustion engine and a turbocharger poses a great challenge to the turbine as pulsating admission conditions are imposed onto the turbocharger turbine. This holds especially true for Daimler's proprietary asymmetric twin scroll turbocharger turbine, where also requirements on exhaust
gas rerouting have to be considered.
This thesis proposes an approach to optimise the asymmetric twin scroll turbine in pulsating
engine environment. In order to do so, firstly, a thorough numerical loss audit is conducted assessing the current geometry and serving as a sensitivity study of the input parameters for the optimisation. Secondly, a novel approach to care for realistic boundary conditions is introduced. Thirdly, the current system is parametrised in a CAD environment,
and an automated parametrisation work flow and the optimisation process are introduced. Subsequently, the meta model based optimisation is executed, and improved optimised geometries are proposed. This study mainly focuses on numerical analysis, although experimental validation is provided wherever required. The optimisation process proposes two signi cantly improved geometries. One is optimised with the single-objective of cycle effciency, the other is multi-objectively optimised for cycle effciency and minimum radial interspace gap enabling an improved engine load
exchange. Compared to the baseline turbocharger turbine, effciency gains of 4.7% and
4.4% are predicted. Nonetheless, it is found, that the geometry changes are very subtle.
The effciency gains cannot be reduced to single parameters but prove to be a combination
of many changes. Although the changes in the geometric parameters are comprehensible,
the results indicate that optimising a turbine in a pulsating environment exceeds the means of an expert based design approach.