Abstract
The particulate matter samples of gasoline direction injection (GDI) gasoline engine obtained from road experiments are photographed to obtain the emission particulate matter images based on the transmission electron microscope under high altitude environment. The results show that the micro-morphology characteristics of the particulate matter emitted by GDI gasoline engine in plateau area are similar to those in low altitude area, and the main morphology is cloud-like, filamentous, flocculent, chain-like, cluster-like, like altitude contour line, etc. The images are processed by MATLAB and Python software, and the fractal dimension, average gray value and gray histogram are obtained. The fractal dimensions of the emission particulate data obtained are between 1.81 and 1.89, which has some deviation compared with the fractal dimension of the emission particulate data at low altitude of 1.58 and 1.80. With the increase of altitude, its fractal dimension increases more obviously. It can be seem from the characteristics of gray value and gray histogram of particle images that the average gray value of GDI gasoline engine particles is higher. The gray histogram distribution of 5–50 nm images is uniform, and the individual particles in the images are clear. The gray histogram of 100–500 nm images is scattered, so it is easy to extract and separate the image edge contour. 5–50 nm images are used to study the details and characteristics of individual particles, and 100–500 nm transmission electron microscope images are used to study the overall morphology of particles.
Introduction
In recent years, gasoline direction injection (GDI) engine has been gradually and widely used in urban motor vehicles because of its good fuel economy. Compared with traditional port fuel injection (PFI) gasoline vehicles and diesel vehicles equipped with Diesel Particle Filter (DPF), GDI gasoline vehicles emit more gasoline exhaust Particles (GEP), which is found to produce more ultra-fine particles (
Experiment and method of particulate matter acquisition
Experimental vehicles and road conditions
The test vehicle adopted is a light and compact model, and the engine is a direct injection engine in cylinder. The main parameters of the experimental vehicle engine are shown in Table 1 and the fuel used is shown in Table 2. The road test of particulate matter collection refers to real driving emission (RDE) experimental regulations, and the experiment took place in Kunming, China. The experimental mileage of urban, suburban and high-speed sections is 24 km, 25 km and 24 km respectively, accounting for one third of the total. The experimental time is about 56 min, 21 min and 17 min respectively, and the total time is 94 min. The lowest altitude is 1862.78 m and the highest altitude is 1952.96 m with a difference of 90.11 m.
Main parameters of experimental vehicle
Main parameters of experimental vehicle
Main physical and chemical indexes of experimental fuel
Dekati low pressure impactor (DLPI). The diameter range measured by Dekati low pressure impactor (DLPI) is 0.03–10 Other equipment. Equipment support frame, connecting pipe, sampling nozzle, sampling gun, heating package, temperature control device, vacuum pump, flowmeter, dryer, collecting membrane and filter membrane, balance, oven, etc.
A large number of samples were obtained in the experiment. In order to study the typical microscopic characteristics of GDI gasoline engine emission particles in plateau area, the samples with the most concentrated particles were selected and preliminarily processed, and the images of gasoline engine emission particles were obtained by field transmission electron microscope FEI TecnaiG2F20S-Twin. In order to further study the characteristics of particulate matter, its fractal dimension, gray scale and gray scale histogram are calculated and analyzed.
Results and discussion
Micro-morphology characteristics of particles
Typical image of particulate matter emitted by GDI gasoline engine.
Some typical images obtained from the experiment are shown in Fig. 1. Through analysis, we can get the following conclusions:
From the typical particle images in Fig. 1, it can see that there are certain typical shapes of particles emitted by GDI gasoline engine. Particulate matter with altitude contour can be seen from image (d), which is similar to the altitude contour in the map. Filamentous particles are shown in image (h), which is similar to silk. See image (l) for flocculent particles, which are similar in shape to cotton wool. There are also typical micro-morphology of gasoline engine emission particles such as cluster as shown in image (c) and chain as shown in image (i). The sorting structure of gasoline engine exhaust particulates is very irregular, and it can be seen that gasoline engine exhaust particulates are stacked and overlapped by basic carbon particles according to a certain level. Similar to the altitude contour of the map, it has a strong sense of hierarchy. Among them, many particles formed by the accumulation of basic carbon particles overlap and fuse with each other to destroy the original extra-nuclear structure, and finally combine into a particle with a relatively flat extra-nuclear structure. However, because too many particles overlap and fuse with each other, the original shell of particles accumulated by basic carbon particles is seriously damaged, and finally the particles are pasty, as shown in image (e). Most of the discharged particulate matter has crystallized, and only a few particulate matter exists in the form of non-carbonaceous, as shown in image (j). It can be seen from image (g) that the particulate matter is formed by the accumulation of a large number of basic carbon particles, and there are a barge number of chemical substances such as H and N between them. The gasoline engine emission particles with extremely irregular shell but chaotic smooth core are formed. In addition, it can be observed from image (c) that there are a large number of basic carbon particles with no fixed shape, and it can be found that their carbonization degree is very low, which leads to the absence of a clear carbon crystal structure in this particle.
In 2020, Guangju et al. used diesel engine state experiment to collect particulate matter emitted by diesel engine [10]. They found that the diameter of the particles was related to the aromatic hydrocarbon content after processing the particles collected in the experiment. There is a big gap between the GDI particulate image and diesel particulate image. As shown in images (a) and (b) in Fig. 1, because the experiment is conducted in a high-altitude area, the low oxygen content in the cylinder of the gasoline engine causes insufficient combustion in the cylinder when the gasoline engine is working, so that the temperature in the cylinder does not reach the temperature where aromatic hydrocarbons are generated. The gasoline engine generates fewer aromatic hydrocarbons in high altitude areas than in low altitude areas, which leads to a lower degree of ordering of the particulate carbon layer obtained in this experiment. In 2020, Hu et al. used gasoline bench test machine to study the microscopic characteristics of particulate matter emitted by gasoline engines [7]. The results show that the cumulative total number of particulate matter emitted by gasoline engine is different in different running periods. As shown in the image (j) of Fig. 1, this will lead to different levels of particulate matter accumulation when the gasoline engine runs in different time periods. Particulate matter is accumulated from a large number of basic carbon particles. The particulate matter samples obtained in this experiment are collected after driving on the road for a period of time, which will result in the accumulation of basic carbon particles of the collected particulate matter samples being greater than those obtained on the bench test machine. Mohit Raj Saxena, a research scholar in India, used off-road gasoline engines for research in 2017 [11]. The results show that using different fuels and mixing different fuels in a certain proportion have great influence on the particle size distribution and concentration of gasoline engine emissions. Especially, the fuel with higher premixed ratio of gasoline and methanol will lead to a sharp increase in the concentration of particulate matter emitted. Compared with the concentration of particulate matter samples obtained in this experiment, the concentration of particulate matter samples obtained in Mohit Raj Saxena experiment is higher and the microscopic morphology presented is more intuitive. Oil injection technology is the key factor affecting the emission level of particulate matter quantity. The study showed that the same fuel injected into the cylinder in different ways would affect the concentration of emission particulate matter and the accumulation degree of basic carbon particles [12]. As shown in image (b) of Fig. 1, compared with PFI gasoline vehicles, GDI gasoline vehicles have a higher concentration of exhaust particles and a higher degree of accumulation of basic carbon particles.
Fractal dimension of image
Fractal dimension of image
Through the analysis, the conclusions are as follows:
Some studies have shown that the fractal dimension of particulate matter will affect the filtration efficiency of GPF, and the filtration efficiency of particulate matter with different fractal dimensions is different [13]. The fractal dimension of GDI gasoline engine emission particulate data obtained by software is between 1.81 and 1.89. Compared with the research results of Tianjin University, the fractal dimension is correct, but there is a certain deviation compared with the fractal dimension of GDI gasoline engine emission particulate data in low altitude areas which is between 1.58 and 1.8 [14, 15]. The experiment was carried out at high altitude. The oxygen content in the air was lower than that at low altitude, which affected the reduction of the combustion efficiency of the fuel in the cylinder. As a result, more unoxidized H and C were adsorbed by the particulates during the exhaust and expansion process of the gasoline engine, which finally led to the increase of the number and concentration of nuclear membrane particles, and the increase of the concentration of the sample particulates obtained in the experiment. Different experimental conditions will lead to a large deviation of fractal dimension, and different ignition conditions will lead to a large range of fractal dimension of the particulate matter samples. The experiment includes three road tests, namely, urban road, rural road and expressway. The running state of gasoline engine is different in these three experimental states. Moreover, the different running time of gasoline engine will lead to a higher degree of accumulation of basic carbon particles of emission and emission particles [16]. As a result, the concentration of the emission particulate matter samples obtained in the experiment increased, and the fractal dimension was relatively large.
(1) Gray value
The average gray value of image processing is obtained by programming according to the principle of average gray value acquisition, as shown in Table 4.
Average gray value of image
Average gray value of image
(2) Gray histogram
The gray histogram of image processing is obtained by programming according to the principle of gray histogram acquisition, as shown in Fig. 2.
Gray histogram of typical particle images.
(3) Analysis and discussion
The range of gray value of images with different multiples is found to be quite different by processing the emission particulate data images. Compared with port injection (PFI), GDI has a higher degree of accumulation of basic carbon particles, resulting in gray values higher than the average [14]. Through comparison, it is found that the gray values of images with typical particle micro-morphology are quite different from those with the same multiple. This is because the images with typical particle micro-morphology are already a mature emission particle cluster, and the adsorption saturation of mature particle clusters has reached the maximum compared with those of immature particle clusters. The gray histogram can visually express the gray frequency in the image. The gray histogram distribution of 5–50 nm images is uniform, and the individual particles in the images are clear, which is convenient to study the size and shape of individual particles, and provides a basis for the design of filter aperture and shape. Combining with the gray histograms obtained by 50 nm, 100 nm, 200 nm and 500 nm image processing, and the typical images of gasoline engine emission particles, it can be seen that their distribution is scattered, and it is easy to extract and separate the edge contour of the images. On this basis, based on the overall accumulation shape characteristics of particles, it provides theoretical support for the structural design of GPF.
The microscopic images of particulate matter emitted by gasoline engines in plateau areas are similar to those in low altitude areas, and they all contain some typical particulate matter characteristics, such as filamentous particulate matter, flocculent particulate matter, chain particulate matter and cluster particulate matter. The fractal dimension of the particulate lens image of an experimental gasoline engine with direct injection in the plateau environment is larger than the fractal dimension measured in the low altitude area, and the range of the fractal dimension is different from that measured in the gasoline engine with direct injection in different ignition periods. With the increase of altitude, the fractal dimension increases more obviously, and the filtration characteristics of GPF will also change. It can be seen from the gray value and gray histogram characteristics of the particle image that 100–500 nm emission particle transmission electron microscope image is easy to use for studying the overall shape of particles, and 5–50 nm image is easy to use for studying particle details and individual particle characteristics.
In the future work, more experiments will be completed to support our research on the characteristics of particulate matter. According to the different gray values of different particle sizes and the frequency of their appearance, the amount of particles with different particle sizes discharged under different working conditions can be estimated, and the collection efficiency of GPF can be improved by combining the microscopic characteristics of particles.
Footnotes
Acknowledgments
The authors acknowledge the National Natural Science Foundation of China (51968065).