Fault Diagnosis in Internal Combustion Engines Using Nonlinear Multivariate Statistics

This paper presents a statistical-based fault diagnosis scheme for application to internal combustion engines. The scheme relies on an identified model that describes the relationships between a set of recorded engine variables using principal component analysis (PCA). Since combustion cycles are complex in nature and produce nonlinear relationships between the recorded engine variables, the paper proposes the use of nonlinear PCA (NLPCA). The paper further justifies the use of NLPCA by comparing the model accuracy of the NLPCA model with that of a linear PCA model. A new nonlinear variable reconstruction algorithm and bivariate scatter plots are proposed for fault isolation, following the application of NLPCA. The proposed technique allows the diagnosis of different fault types under steady-state operating conditions. More precisely, nonlinear variable reconstruction can remove the fault signature from the recorded engine data, which allows the identification and isolation of the root cause of abnormal engine behaviour. The paper shows that this can lead to (i) an enhanced identification of potential root causes of abnormal events and (ii) the masking of faulty sensor readings. The effectiveness of the enhanced NLPCA based monitoring scheme is illustrated by its application to a sensor fault and a process fault. The sensor fault relates to a drift in the fuel flow reading, whilst the process fault relates to a partial blockage of the intercooler. These faults are introduced to a Volkswagen TDI 1.9 Litre diesel engine mounted on an experimental engine test bench facility.

A note on the prediction of liquid hold-up with the stratified roll wave regime for gas/liquidco-current flow in horizontal pipes.

This note presents a simple model for prediction of liquid hold-up in two-phase horizontal pipe flow for the stratified roll wave (St+RW) flow regime. Liquid hold-up data for horizontal two-phase pipe flow [1, 2, 3, 4, 5 and 6] exhibit a steady increase with liquid velocity and a more dramatic fall with increasing gas rate as shown by Hand et al. [7 and 8] for example. In addition the liquid hold-up is reported to show an additional variation with pipe diameter. Generally, if the initial liquid rate for the no-gas flow condition gives a liquid height below the pipe centre line, the flow patterns pass successively through the stratified (St), stratified ripple (St+R), stratified roll wave, film plus droplet (F+D) and finally the annular (A+D, A+RW, A+BTS) regimes as the gas rate is increased. Hand et al. [7 and 8] have given a detailed description of this progression in flow regime development and definitions of the patterns involved. Despite the fact that there are over one hundred models which have been developed to predict liquid hold-up, none have been shown to be universally useful, while only a handful have proven to be applicable to specific flow regimes [9, 10, 11 and 12]. One of the most intractable regimes to predict has been the stratified roll wave pattern where the liquid hold-up shows the most dramatic change with gas flow rate. It has been suggested that the momentum balance-type models, which give both hold-up and pressure drop prediction, can predict universally for all flow regimes but particularly in the case of the difficult stratified roll wave pattern. Donnelly [1] recently demonstrated that the momentum balance models experienced some difficulties in the prediction of this regime. Without going into lengthy details, these models differ in the assumed friction factor or shear stress on the surfaces within the pipe particularly at the liquid–gas interface. The Baker–Jardine model [13] when tested against the 0.0454 m i.d. data of Nguyen [2] exhibited a wide scatter for both liquid hold-up and pressure drop as shown in Fig. 1. The Andritsos–Hanratty model [14] gave better prediction of pressure drop but a wide scatter for liquid hold-up estimation (cf. Fig. 2) when tested against the 0.0935 m i.d. data of Hand [5]. The Spedding–Hand model [15], shown in Fig. 3 against the data of Hand [5], gave improved performance but was still unsatisfactory with the prediction of hold-up for stratified-type flows. The MARS model of Grolman [6] gave better prediction of hold-up (cf. Fig. 4) but deterioration in the estimation of pressure drop when tested against the data of Nguyen [2]. Thus no method is available that will accurately predict liquid hold-up across the whole range of flow patterns but particularly for the stratified plus roll wavy regime. The position is particularly unfortunate since the stratified-type regimes are perhaps the most predominant pattern found in multiphase lines.

Qualitative methods in the study of children's racial attitudes and identities

This article examines the role that qualitative methods can play in the study of children's racial attitudes and behaviour. It does this by discussing a number of examples taken from a qualitative, ethnographic study of five- and six-year-old children in an English multi-ethnic, inner-city primary school. The examples are used to highlight the limitations of research that relies solely on quantitative methods and the potential that qualitative methods have for addressing these limitations. Within this context the article contrasts the strengths and weaknesses of qualitative and quantitative methods in the study of children's racial attitudes and identities. The article concludes by arguing that a much more integrated multi-method approach is needed in this area and sets out some of the most effective ways this could be achieved.

Influence of individual predisposition, maternal experience and lactation environment on the responses of pigs to weaning at two different ages

This study assessed the effect of predisposition to perform harmful social behaviour, maternal rearing environment, and lactation environment on the responses of pigs to weaning at 3 or 5 weeks of age. Predisposed and non-predisposed gilts were selected as dams for this study at 7 weeks of age. Selection was based on behaviour in a

'PERMIT' ion migration test for measuring the chloride ion transport of concrete on site

The ingress of chlorides into concrete is predominantly by the mechanism of diffusion and the resistance of concrete to the transport of chlorides is generally represented by its coefficient of diffusion. The determination of this coefficient normally requires long test duration (many months). Therefore, rapid test methods based on the electrical migration of ions have widely been used. The current procedure of chloride ion migration tests involves placing a concrete disc between an ion source solution and a neutral solution and accelerating the transport of ions from the source solution to the neutral solution by the application of a potential difference across the concrete disc. This means that, in order to determine the chloride transport resistance of concrete cover, cores should be extracted from the structure and tested in laboratories. In an attempt to facilitate testing of the concrete cover on site, an in situ ion migration test (hereafter referred to as PERMIT ion migration test for the unique identification of the new test) was developed. The PERMIT ion migration test was validated in the lab by carrying out a comparative investigation and correlating the results with the migration coefficient from the one-dimensional chloride migration test, the effective diffusion coefficient from the normal diffusion test and the apparent diffusion coefficient determined from chloride profiles. A range of concrete mixes made with ordinary Portland cement was used for this purpose. In addition, the effects of preferential flow of ions close to the concrete surface and the proximity of reinforcement within the test area on the in situ migration coefficients were investigated. It was observed that the in situ migration index, found in one working day, correlated well with the chloride diffusion coefficients from other tests. The quality of the surface layer of the cover concrete and the location of the reinforcement within the test area were found to affect the flow of ions through the concrete during the test. Based on the data, a procedure to carry out the PERMIT ion migration test was standardised.