Influence of gasoline - Ethanol blends on engine torque variation

In this paper the possibility to use the instantaneous torque measurement to estimate the injected fuel mixture is explored. The analysis focuses on a four stroke SI engine equipped with a low pressure common rail type multi-fuel injection system. First, the injection system is simulated by means of a comprehensive lumped and distributed parameter numerical model, in order to evaluate the dynamic behavior of the fuel rail in terms of injection pressure profiles, instantaneous mass flow rate delivered to each cylinder and engine heat of combustion power. The accuracy of the model is addressed by comparing the predicted results with the measured data. By using both the experimental in-cylinder pressure profiles as well as the engine total efficiency, and the calculated injection profiles, the instantaneous torque is determined for different engine speeds and ethanol/gasoline blends. In particular the torque variation during the transition from one fuel mixture to a different one is investigated, accounting for the influence of the blend composition rate of change. Finally, the magnitude of instantaneous torque variation is compared with the overall accuracy of a torque sensor currently adopted in production series engines to detect combustion misfires, in order to evidence its capability to monitor the fuel blend transition.

In this paper the possibility to use the instantaneous engine torque measurement to estimate the injected fuel mixture is explored. The analysis focuses on a four stroke SI engine equipped with a low pressure common rail type multi-fuel injection system. First, the injection system is simulated by means of a comprehensive lumped and distributed parameters numerical model, in order to evaluate the dynamic behavior of the fuel rail in terms of injection pressure profiles, instantaneous mass flow rate delivered to each cylinder and engine heat of combustion power. The accuracy of the model is addressed by comparing the predicted results with the measured data. Afterward, the 1D model of the whole engine is constructed and validated against experimental measurements. By using one dimensional engine simulation the previously calculated injection profiles are used to determine the instantaneous torque for different engine speeds and ethanol/gasoline blends. Finally, the engine torque variation is assessed under vehicle accelerating conditions during the transition from one fuel mixture to a different one. Furthermore, the fuel blend composition rate of change for each injector is preliminarily calculated by means of a CFD analysis of the fuel switch over within the low pressure fuel rail during the acceleration. An open source computational fluid dynamics code is used in the modeling. The magnitude of instantaneous engine torque variation is compared with the overall accuracy of a standard torque sensor, in order to evidence its capability to monitor the fuel blend transition. Copyright © 2009 SAE International.