Journal of Mechanical Science and Technology 28 (8) (2014) 3311~3323 www.springerlink.com/content/1738-494x DOI 10.1007/s12206-014-0742-x Model-based optimization of injection strategies for SI engine gas injectors † Stefano Beccari, Emiliano Pipitone * , Marco Cammalleri and Giuseppe Genchi Department of Chemical, Management, Computer and Mechanical Engineering, University of Palermo, Viale delle Scienze, Palermo, Italy (Manuscript Received July 15, 2013; Revised April 1, 2014; Accepted April 17, 2014) ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Abstract A mathematical model for the prediction of the mass injected by a gaseous fuel solenoid injector for spark ignition (SI) engines has been realized and validated through experimental data by the authors in a recent work [1]. The gas injector has been studied with particu- lar reference to the complex needle motion during the opening and closing phases. Such motion may significantly affect the amount of injected fuel. When the injector nozzle is fully open, the mass flow depends only on the upstream fluid pressure and temperature. This phenomenon creates a linear relationship between the injected fuel mass and the injection time (i.e. the duration of the injection pulse), thus enabling efficient control of the injected fuel mass by simply acting on the injection time. However, a part of the injector flow chart characterized by strong nonlinearities has been experimentally observed by the authors [1]. Such nonlinearities may seriously compro- mise the air-fuel mixture quality control and thus increase both fuel consumption and pollutant emissions (SI engine catalytic conversion systems have very low efficiency for non-stoichiometric mixtures). These nonlinearities arise by the injector outflow area variation caused by needle impacts and bounces during the transient phenomena, which occur in the opening and closing phases of the injector. In this work, the mathematical model previously developed by the authors has been employed to study and optimize two appropriate injec- tion strategies to linearize the injector flow chart to the greatest extent. The first strategy relies on injection pulse interruption and has been originally developed by the authors, whereas the second strategy is known in the automotive engine industry as the peak and hold injection. Both injection strategies have been optimized through minimum injection energy considerations and have been compared in terms of linearization effectiveness. Efficient linearization of the injector flow chart has been achieved with both injection strategies, and a similar increase in injector operating range has been observed. The main advantage of the pulse interruption strategy lies on its ease of implementation on existing injection systems because it only requires a simple engine electronic control unit software update. Meanwhile, the peak and hold strategy reveals a substantial lack of robustness and requires expressly designed injectors and electronic components to perform the necessary voltage commutation. Keywords: Gas injector; Injection strategy; Spark ignition engine ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 1. Introduction Experimental observation by the authors in a previous work [1] revealed the existence of strong nonlinearities in the lower part of a typical spark ignition (SI) engine gas injector flow chart (i.e., the diagram that reports the injected mass as a func- tion of injection time). These nonlinearities can cause unstable control of engine air-fuel ratio and may thus compromise both engine efficiency and pollutant emissions. The literature re- search conducted by the authors showed the existence of nu- merous studies on the simulation and modeling of internal combustion engine injection systems. Compression ignition (CI) engines are usually equipped with high-pressure (1600 bar-2000 bar) common rail injectors [2, 3], which can be acti- vated by either a solenoid or a piezoelectric element. These injectors use the high pressure of fuel to move the needle and open the nozzle [4]. SI engines may be port injected or direct injected. The for- mer usually employ low-pressure (3 bar-10 bar, depending on fuel type) injectors [2, 3, 5], whereas the latter may require higher injection pressure (100 bar-500 bar) [3, 6]. Although considerable research on injection system simula- tion is available in the literature, only a few works cover the dynamic modeling of injector needle motion. Such motion significantly influences the injector diagram and is the focus of this paper. In relation to common rail injection systems, needle motion has been well discussed in the literature. For example, the fluid-dynamic model presented in a previous work [7] predicts the injection pressure variations to derive proper injection control laws. The model developed in Ref. [8] also predicts the needle lift and injection rate for different injection pressures. The common rail piezoelectric injector model realized in Ref. [9] considers both the hydraulic part (fluid flow, discharge coefficients) and the mechanical part * Corresponding author. Tel.: +39 091 23897280, Fax.: +39 091 23860840 E-mail address: emiliano.pipitone@unipa.it † Recommended by Associate Editor © KSME & Springer 2014