Analysis of ion current signal during negative valve overlap of HCCI combustion with high compression ratio

Abstract

Homogenous charge compression ignition (HCCI) combustion is a low temperature combustion process which combines high combustion efficiency with ultra-low [Formula: see text] raw emissions. Steep increases of the in-cylinder pressure and unstable combustion sequences at the limits of the operating range can damage the engine and limit the use of HCCI to part load operation. This can be done using closed loop combustion control based on combustion parameters like the indicated mean effective pressure and the combustion phasing. Since in-cylinder pressure sensors are expensive components and therefore not suitable for series application, ion current sensors can be used as an additional source of information about the combustion. Combustion analysis using methods similar to those used in pressure based measurements can be implemented using an online analysis of the ion current signal. In this study, the ion current sensor will be examined for its suitability for combustion control under HCCI conditions with lean air/fuel ratios and high compression ratios. Research has found that the ion current signal is strongly depended on the boundary conditions. Especially the air/fuel ratio which plays an important role for signal strength during the combustion process. When using valve timings with negative valve overlap in combination with a fuel pre-injection, a further peak of the ion current signal close to the gas exchange top dead center can be found in addition to the one during combustion. At the same time, it is hard to extract information from the cylinder pressure signal during NVO. Under lean conditions this peak even exceeds the signal during combustion. This study analyzes the ion current signal during NVO and its potential to be used for future combustion control concepts. The ion current signal shows potential to stabilize HCCI combustion at high loads. However, the prediction of late combustion cycles is still challenging.

Physics-based modeling of homogeneous charge compression ignition engines with exhaust throttling

“…The sampling period is set as the engine cycle time 0.08 s at 1500 rpm. The resulting state space representation of the discrete-time second order linear model is shown in (33) and (34).…”

“…Figure 24 shows the comparison between the 2 nd order discrete-time linear model in (33) and (34) and the 11 th order nonlinear continuous-time model in Section 3 during step changes in fuel and exhaust throttle commands. As the fuel steps down, the combustion timing CA50 is retarted.…”

“…Firstly, the Kalman filter based estimator is designed to estimate the states based on the linear model in (33) and (34). The implementation of Kalman filter is as follows:…”

“…For instance, re-induction can be achieved by reopening the exhaust valve during the intake stroke, hence some of the exhaust gas is reinducted into the cylinder along with the fresh charge. 24,26,27 On the other hand, early closing of the exhaust valve and late opening of the intake valve, also known as negative valve overlap (NVO) [28][29][30][31][32][33] leads to retention of burned charge inside the cylinder. Stability analysis is conducted based on an HCCI engine with high dilution reinducted by a second opening of the exhaust valve during the intake stroke.…”

“…The proposed exhaust throttle mechanism, which is little explored, is cost effective and can easily be installed even on two wheelers. The crank angle resolved model gives the full trace of the cylinder pressure, temperature, and burned gas fraction, hence provides more insight about the gas exchange and combustion process compared to the mean value model in Souder et al 35 On the other hand, compared to the multi-zone models in Shingne et al, 31 Garcia-Guendulain et al, 32 Wick et al, 33 which provide more detailed thermodynamic conditions, the zero-dimensional single zone model is capable of predicting the HCCI combustion timings with reduced computation cost in the real-time hardware-in-the-loop (HIL) simulation environment when uniform distributions of the mixture and temperature are considered. 38 The dynamics of the engine model are exhibited collectively by the evolution of states in the intake, exhaust runners, and the cylinder.…”