ElsevierSerrano Cruz, JR.; Arnau Martínez, FJ.; García-Cuevas González, LM.; Dombrovsky, A.; Tartoussi, H. (2016). Development and validation of a radial turbine efficiency and mass flow model at design and off-design conditions. Energy Abstract Turbine performance at extreme off-design conditions is growing in importance for properly computing turbocharged reciprocating internal combustion engines behaviour during urban driving conditions at current and future homologation cycles. In these cases, the turbine operates at very low flow rates and power outputs and at very high blade to jet speed ratios during transitory periods due to turbocharger wheel inertia and the high pulsation level of engine exhaust flow. This paper presents a physically based method that is able to extrapolate radial turbines reduced mass flow and adiabatic efficiency in blade speed ratio, turbine rotational speed and stator vanes position. The model uses a very narrow range of experimental data from turbine maps to fit the necessary coefficients. By using a special experimental turbocharger gas stand, experimental data have been obtained for extremely low turbine power outputs for the sake of model validation. Even if the data used for fitting only covers the turbine normal operation zone, the extrapolation model provides very good agreement with the experiments at very high blade speed ratio points; producing also good results when extrapolating in rotational speed and stator vanes position.
RPerfect gas constant (Jkg −1 K −1 ) r Rotor radius (m) sd Standard deviation (−) 12 provided by manufacturers as a standard practice. Turbine maps are necessary 13 when using 1D or 0D modelling tools to predict the whole engine behaviour. In 14 1D modelling approach the one-dimensional unsteady non-homentropic mass, 15 momentum and energy conservation laws (Euler equations) are solved. Specific 16 source terms are used to simulate the friction and heat exchange in the pipes. 17 The 1D simulation codes make possible the calculation of gas dynamics engine 18 behaviour at low computational costs. Some engine components are modelled 19 4 with a 0D approach, using specific lumped parameter models or performance 20 maps. That is the case of cylinders, injectors, valves, compressors and turbines 21 which are coupled to the 1D computational domain as it is described in [4]. 22 For that reason, turbocharged ICE designers must rely on map extrapolation 23 tools when predicting engine performance outside of turbine design operative 24 conditions [5]. It is typical in pulsating flow conditions, requiring different mod-25 elling approaches similar to the proposed in [6], where meanline one-dimensional 26 models are discussed and in [7], where non-adiabatic pressure loss boundary con-27 dition is discussed. One-dimensional tools are also used in design process for 28 fast evaluation of turbine capabilities as in [8]. In [9] a Taylor series expansion 29 is used to develop a model able of predicting mass flow parameter of radial 30 turbines. 31 CFD models for turb...