Research On The Analysis And Identification Of Combustion Soot Generated By Typical Fuels

Researches on the combustion soot occupy significant status in the fields of environmental ecology, engines, combustion mechanisms, etc. and it is also an important issue involved in the field of fire investigation. Fire investigation is a complicated inter-discipline subject with strong comprehensiveness. The extraction, detection and identification of accelerants in fire scene is always an important research direction in fire investigation at home and abroad. The fire site is easily destroyed by many external factors, such as the damage intentionally by the arsonists or the fire-extinguishing action of firefighters and other human factors. These factors make it difficult to detect accelerants from the combustion residues. Nevertheless, the combustion soot existing in the fire scene can still well save the information of accelerants. To date, rare literature has been reported on the extraction and identification of combustibles soot in fire investigation and existing researches on this issue are not thorough enough. Therefore, to develop the study on the analysis, extraction and identification of typical combustion soot is very essential, which can not only provide scientific technical support for the judgment of whether there is accelerants and what kind the accelerants is in the fire scene, but also further enrich and deepen the research on the combustion soot in fire investigation.This work focuses on the extraction and identification of typical combustion soot on basis of controllable combustion experiments. Firstly, the physical properties and chemical components of typical combustion soot were analyzed. Through the comparison of the transmission electron microscopy (TEM) images of the combustion soot from the gasoline, the polystyrene (PS), the ABS, the diesel and the n-heptane three parameters namely the compactness, complexity and gray value of the images are analyzed in depth, which illustrate the physical diversities of the three different kinds of soot. Using the Solid Phase Microextraction (SPME)-GC-MS coupling technique, the chemical components of the first three kinds of soot are analyzed and compared. The aromatic hydrocarbons compounds in the combustion soot are conducted characteristic ions identification and through the deep comparison of the components of these three kinds of soot, major two-ring aromatics, three-ring aromatics and four-ring aromatics are found in the combustion soot. Further analysis on the influence factors such as the ventilation condition, the soot sampling position and the static placing time of soot, etc. on the formation of the combustion soot from some typical combustibles namely gasoline and diesel are conducted through the SPME method. With a series of experimental comparison of different extraction time and extraction temperature in SPME test, the optimal extraction condition is proposed and the diversities of the soot components, as well as the reasons causing the component differences are analyzed.Based on the SPME tests, this work introduces the relevant theories and approaches of pattern recognition into the field of combustion soot. Various approaches of pattern recognition are applied to identify the combustion soot from different sources. Firstly, SPME is applied to extract and analyze six kinds of accelerant soot including diesel soot, gasoline soot, and diesel-gasoline mixture soot under controlled combustion experiments. The diversities of the soot components are compared and the major component variables are obtained with principal component analysis method, based on which different kinds of soot are well classified. Then the hierarchical cluster analysis is applied for the dendrogram analysis of the six kinds of soot based on the major component variables, which also obtains good classification efficiency, demonstrating the effectiveness of this pattern recognition method in the identification of accelerants in imitated fire scene. Furthermore, the principal component analysis method and the hierarchical cluster analysis are respectively applied in the identification of four typical and common combustibles i.e. the gasoliner, the diesel, the polystyrene and the ABS in real fire scene, which also obtain a classification of100%accuracy of these kinds of soot.In order to compare the diversities of the SPME and traditional extraction method in the detection of soot components, the traditional liquid extraction is chosen in this dissertation. Petroleum ether is used as the extracting solvent to prepare soot samples of diesel, polystyrene and ABS. Gas chromatography and mass spectrometry (GC-MS) with automatic sampling is used to obtain the soot spectrograms. Compared with the spectrograms obtained by SPME sampling, it is found that the two sampling methods are complementary to some extent. With the liquid extraction, more PAHs with4to6rings are obtained, while with SPME extraction, more PAHs with1to3rings are obtained. Therefore, the information will be more comprehensive if the two extraction methods are applied together. Based on the spectrograms obtained by liquid extraction, the principal component analysis is used to recognize the three kinds of soot, which also proved to be efficient. Finally, the BP neural network is applied to classify and predict the sources of soot samples, which initiate new ideas of BP neural network in the information mining of combustion soot. Finally, to understand the influence of the soot formation process on the identification of soot sources, the mechanism of soot formation is discussed in depth. As the main substitute for gasoline, the toluene is chosen to investigate the mechanism of soot formation. With the infrared spectroscopic analysis, the diversities of soot components at different height of the toluene’s laminar flame are obtained and the soot formation mechanism of the toluene’s pool fire is studied. Besides, through the combustion experiments of the mixtures of toluene and methanol or ethanol which are conducted under the same conditions, it is found that the alcohol can noticeably reduce the extinction coefficient of the toluene soot, and meanwhile slow down the soot formation by decelerating the combustion. The analysis of the mechanism of soot formation provides theoretical basis for further understanding the soot components.