A Methodology for Measuring Turbocharger Adiabatic Maps in a Gas-Stand and its Usage for Calibrating Control Oriented and 1D Models at Early ICE Design Stages

Abstract

Abstract Turbocharged engines are the standard architecture for designing efficient spark ignition and compression ignition reciprocating internal combustion engines (ICE). Turbochargers characterization and modeling are basic tasks for the analysis and prediction of the whole engine system performance and this information is needed in quite early stages of the engine design. Turbocharger characteristics (efficiency, pressure ratio, mass flow rates...) traditionally rely in maps of pseudo non-dimensional variables called reduced variables. These maps must be used by reciprocating ICE designer and modeler not only for benchmarking of the turbocharger, but for a multiplicity of purposes, i.e: assessing engine back-pressure, boost pressure, load transient response, after-treatment inlet temperature, intercooler inlet temperature, low pressure EGR temperature, ... Maps of reduced variables are measured in gas-stands with steady flow but non-standardized fluids conditioning; neither temperatures nor flows. In concrete: turbine inlet gas temperature; lubrication-oil flow and temperature; water-cooling flow and turbo-machinery external heat transfer are non-standardized variables which have a big impact in assessing said multiplicity of purposes. Moreover, adiabatic efficiency, heat losses and friction losses are important data, hidden in the maps of reduced variables, which depend on the testing conditions as much as on the auxiliary fluids temperature and flow rate. In this work it is proposed a methodology to standardize turbochargers testing based in measuring the maps twice: in close to adiabatic and in diathermal conditions. Along the paper it is discussed with special detail the impact of the procedure followed to achieve said quasi-adiabatic conditions in both the energy balance of the turbocharger and the testing complexity. As a conclusion, the paper proposes a methodology which combines quasi-adiabatic tests (cold and hot gas flow) with diathermal tests (hot gas flow) in order to extract from a turbocharger gas-stand all information needed by engine designers interested in controlling or 1D-modelling the ICE. The methodology is completed with a guide for calibrating said control-oriented turbocharger models in order to separate aerodynamic efficiency (adiabatic) from heat transfer losses and from friction losses in the analysis of the turbocharger performance. The outsourced calibration of the turbocharger model allows avoiding uncertainties in the global ICE model calibration, what is very interesting for turbochargers benchmarking at early ICE-turbo matching stages or for global system analysis at early control design stages.