Search Results

Data decomposition techniques have become a standard approach for the analysis of 2D imaging data originating from optically accessible internal combustion engines. In particular, the method of Proper Orthogonal Decomposition (POD) has proven to be a valuable tool for the evaluation of cycle-to-cycle variability based on luminous combustion imaging and particle image velocimetry (PIV) measurements. POD basically permits to characterize the dominant structures of the process under consideration. Recently, an alternative procedure based on Independent Component Analysis (ICA) has been introduced in the engine field. Unlike POD, the method of ICA identifies the patterns corresponding to physical processes that are statistically independent. In this work, a Group-ICA approach is applied to 2D cycle-resolved images of the luminosity emitted by the combustion process. The analysis is meant to characterize cyclic variability of a port fuel injection spark ignition (PFI SI) engine.

This study aims at building an efficient and robust radial basis function (RBF) artificial neural network (ANN), to reconstruct the in-cylinder pressure of a diesel engine starting from the signal of a low-cost accelerometer placed on the engine block. The accelerometer is a perfect non-intrusive replacement for expensive probes and is prospectively suitable for production vehicles. The RBF network is trained using measurements from different engine operating conditions. Training data are composed of time series from the accelerometer and corresponding measured in-cylinder pressure signals. The RBF network is then validated using data not included in training and the results show good correspondence between measured and reconstructed pressure signal. Various network parameters are used to optimize the network quality.

This work reports on the application of spectroscopic measurements coupled with data processing techniques in order to study, in terms of spectral emissions, the dynamic of the HCCI (Homogeneous charge compression ignition) combustion that occurs inside the combustion chamber of an optically accessible direct injection Diesel engine. A pre-processing of the recorded spectra is required for a correct analysis. The procedure of pre-processing consists of two main steps, that is: noise filtering with a technique based on the POD (Proper Orthogonal Decomposition); estimate and subtraction of the baseline. The analysis of the dynamics of the recorded spectra was carried out by the estimates of the synchronous and asynchronous 2D correlation spectra.

Flame luminosity fields can nowadays be collected from optically accessible engines, with high spatial and temporal resolution, and constitute a very powerful investigation means for the transient combustion phenomena taking place in the engine chamber. Interpretation of the impressive amount of collected data can be quite challenging, mainly due to the variety of coupled phenomena involved. Application of Independent Component Analysis (ICA) aims here at separating spatial structures related to different combustion events, and is coupled with the analysis of the statistics of the coefficients of the independent components, and of the measured in-cylinder parameters. This paper reports on the comparison of the application of ICA to 2D images of combustion-related luminosity collected from two different optically accessible engines: Diesel and spark ignition.

This study aims at building efficient and robust artificial neural networks (ANN) able to reconstruct the in-cylinder pressure of Diesel engines and to identify engine conditions starting from the signal of a low-cost accelerometer placed on the engine block. The accelerometer is a perfect non-intrusive replacement for expensive probes and is prospectively suitable for production vehicles. In this view, the artificial neural network is meant to be efficient in terms of response time, i.e. fast enough for on-line use. In addition, robustness is sought in order to provide flexibility in terms of operation parameters. Here we consider a feed-forward neural network based on radial basis functions (RBF) for signal reconstruction, and a feed-forward multi-layer perceptron network with tan-sigmoid transfer function for signal classification. The networks are trained using measurements from a three-cylinder real engine for various operating conditions.

The growing concerns on the emission of particles smaller than 23 nm, which are harmful to human health, lead to the necessity of introducing a regulation for these particles not yet included in the current emission standards. Considering that measurements of concentration of sub-23 nm particles are particularly sensitive to the sampling conditions, it is important to identify an effective assessment procedure. Aim of this paper is the characterization of the effect of the sampling conditions on sub-23 nm particles, emitted by PFI (port fuel injection) and DI (direct injection) spark ignition engines fueled with gasoline, ethanol and a mixture of ethanol and gasoline (E30). The experimental activity was carried out on a 250 cm3 displacement four stroke GDI and PFI single cylinder engines. The tests were conducted at 2000 rpm and 4000 rpm full load, representative of the homologation urban driving cycle.