节点文献

压燃式发动机超细颗粒排放特性及其在大气环境中的污染特征研究

Study on Characteristics of Ultrafine Particles Emitted From Compression Ignition Engines and Its Pollution Features in Real World

【作者】 刘炜

【导师】 黄震;

【作者基本信息】 上海交通大学 , 环境工程, 2009, 博士

【摘要】 随着我国机动车保有量的迅速增长,机动车已经成为城市大气环境的主要污染源。机动车排放的超细颗粒能够附着大量有害组分,并通过呼吸道进入人体。这些有害组分的摄入会引起呼吸系统、心脑血管系统和消化系统等组织发生病变。因此,机动车超细颗粒污染已成为当今城市大气环境研究领域的热点。本文通过发动机台架实验、机动车实际大气环境排放现场测试以及CFD模拟计算等方法对影响发动机超细颗粒排放及其在大气环境中的稀释变化的主要因素进行了深入系统的研究。在自制的单级稀释系统上研究了燃料含氧量对压燃式发动机排气超细颗粒数浓度粒径分布及其化学组成的影响。研究表明,燃料含氧量对发动机超细颗粒排放特性有明显影响。随着燃料含氧量的增大,排气积聚模态微粒数浓度显著下降,而排气核模态微粒数浓度显著增大,排气总颗粒数浓度明显增大,排气总颗粒质量浓度则明显降低。随着燃料含氧量的增加,干碳烟含量明显降低,而可溶性有机物(SOF)含量显著增加,硫酸盐含量变化较小。燃用普通柴油和含氧混合燃料时,柴油机排气微粒中SOF主要成分为饱和烷烃。同时滤膜SOF分析得出,随着燃料含氧量的增加,烷烃和PAHs的含量降低,有机酸脂的含量增大。通过对比研究燃用普通柴油和超低硫柴油排气超细颗粒的特性,分析了燃料含硫量对压燃式发动机超细颗粒排放特性的影响。研究表明,燃料含硫量对压燃式发动机超细颗粒排放特性有显著影响。相对于燃用普通柴油,燃用超低硫柴油其排气积聚模态微粒数浓度变化不明显,而核模态微粒数浓度显著降低,排气总颗粒质量浓度略有降低,而总颗粒数浓度则显著降低,且降幅随负荷和转速增大。燃用普通柴油和超低硫柴油其排气微粒元素组成基本相似,主要由C、O、Cl、S、Si、Ca、Na、Al、K元素组成,但燃用超低硫柴油其排气微粒中没有发现S元素。排气微粒中SOF主要成分有饱和烷烃、有机酸酯和多环芳烃等。燃用超低硫柴油时微粒采样中未发现多环芳烃。在国产两缸压燃式发动机上研究了喷油参数(喷油提前角和燃油起喷压力)以及模拟EGR对柴油直喷燃烧时其超细颗粒排放特性的影响。研究表明,喷油提前角对发动机排气超细颗粒数浓度分布曲线形状影响较小,随着喷油提前角的增大,发动机排气积聚模态和核模态微粒数浓度均明显减小;随着燃油起喷压力的增大,排气积聚模态微粒数浓度明显减小,而核模态微粒数浓度则显著增大;随着进气道CO2浓度的增大,其排气积聚模态数浓度显著增大,而核模态微粒则明显减少。在改造的发动机上研究了正庚烷和二甲醚缸内直喷压燃、进气道喷射均质压燃(HCCI)以及进气道—气缸喷射复合燃烧3种燃烧模式发动机排气超细颗粒特性。正庚烷燃烧试验研究表明,排气超细颗粒数浓度粒径分布曲线随预混合率变化显著。正庚烷直喷燃烧时,排气超细颗粒以积聚模态微粒为主。随着预混合率的增大,排气积聚模态微粒数浓度明显降低;而核模态数浓度则显著增大。正庚烷HCCI燃烧时发动机排气超细颗粒以核模态微粒为主,积聚模态微粒数浓度很低。正庚烷复合燃烧和HCCI燃烧相对于直喷燃烧发动机排气总颗粒数浓度明显较高。排气核模态数浓度与HC排放明显相关。二甲醚燃烧研究表明,发动机排气超细数浓度分布特征随着预混合率变化较小,均以核模态为主,积聚模态数浓度很低。DME复合燃烧其排气总颗粒数浓度相对DME HCCI燃烧时略小,然而相对DME直喷燃烧时明显较大。排气核模态数浓度和总颗粒数浓度均与HC排放显著相关。实地测试了上海市3种典型道路(城市主干道、城市支干道和高速公路)旁机动车超细颗粒实时排放特性。结果表明,实际大气环境中,车流量、车流种类构成对测试结果有明显影响。超细颗粒数浓度和质量浓度随时间变化明显,超细颗粒数浓度随车流量的增大而增大;相对下午车流高峰期,上午车流高峰期超细颗粒数浓度和质量浓度明显较大。3种道路超细微粒数浓度粒径分布曲线形状差异较小,数浓度分布几何平均粒径约为22~35nm;质量浓度分布几何平均直径差异较大。总颗粒平均数浓度、CO平均浓度以及PM1平均浓度之间明显线性相关。计算得到机动车超细颗粒平均数量排放因子为4.52×1014 ~ 8.50×1014 particles/(km·Vehicle)。现场测试了高速公路下风向超细颗粒横向分布特性,并通过风向频率加权模型结合CFD模型对超细颗粒的横向分布特性进行了模拟研究。实测试验发现,随着横向距离的增加,超细颗粒数浓度和质量浓度浓度峰值均显著下降,而其峰值粒径变化较小。对于数浓度分布,不同距离处积聚模态的峰值和峰值粒径变化较小。同时可以得出,随着距离的增加,超细颗粒数浓度分布平均粒径显著增大。106m处相对于17m处,其平均颗粒数浓度粒径分布平均粒径增大了约12nm。NTotal与CO平均浓度有较好的相关性(R2=0.78)。幂函数能较好地反映高速公路超细颗粒横向CFD模拟值和实际测量值衰减率。由实际测量的结果与风向加权结合CFD模拟计算结果比较可知,相对于CFD模拟,实际测量超细颗粒数浓度和质量浓度横向衰减率均明显较大。同时对实际测量数据的分析得出,总颗粒数浓度的衰减主要是由核模态微粒的衰减所引起。

【Abstract】 With the rapid increase of the vehicle population in China, the vehicle emission pollution have been becoming as the major pollution source in the urban atmosphere. Many toxicological studies reported that ultrafine particles (Dp<100 nm) contain considerable trace amounts of toxic substances and organic pollutants, which can penetrate deep into the lung and enter interstitial tissues, causing severe respiratory inflammation and acute pulmonary toxicity. Therefore, the ultrafine particles from vehicle have attracted extensive attention.The aim of this study is to evaluate the effects of engine factors on ultrafine particles from engines and atmospheric factors on dispersion of ultrafine particles during exhuast dilution. Using a single-stage dilution tunnel and SMPS particle size analyse instrument, author conducted engine rig test, roadside measurement and CFD simulation.Effect of oxygen content of fuel on the exhuast particle number concentration and size distribution was studied. The results indicate that, with the increase of oxygen content of fuel, number concentration in the accumulation mode decreases obviously, number concentration in the nucleation mode increases significantly. Meanwhile, with the increase of oxygen content of fuel, number concentration of total particles (10~487nm) increases, mass concentration of total particles decreases obviously. Also, with the increase of oxygen content of fuel, the mass content of DS decreases significantly. However, the mass contents of SOF and ester increase evidently with oxygen content of oxygenated diesel.Number size distributions (NSDs, 10-487nm) and composition of nanoparticle emitted from an engine fuelled with ordinary diesel (OD) and low sulfur diesel (LSD) fuel were comparatively studied. The results indicate that, compared with the OD, the LSD is found to slightly reduce the mass concentration, and significantly reduce the number concentration of the total particles (10-487 nm), and the reduction of number increases with the speed and load of engine. Under the same engine conditions, the nucleation mode for LSD fuel is significantly lower than those of ordinary diesel. However, the accumulation mode for the two fuels shows little difference. The elements composition of exhaust particles include C, O, Cl, S, Si, Ca, Na, Al and K. The S element is not detected in LSD fuel case. The main components of soluble organic fraction (SOF) of exhaust particles for the two fuels include saturated alkane (C15-C26), ester and polycyclic aromatic hydrocarbons (PAHs). However PAHs are not found in LSD fuel case.Effects of fuel supply advance angle , injection pressure , and simulated exhaust gas recirculation (EGR) on NSDs of exhaust particles emitted from a DI diesel engine were experimentally investigated. The results show that increase fuel supply advance angle slightly reduces the total exhasust particle. Increasing injection pressure is found to reduce the accumulation mode particle number and favour the formation of nucleation mode particles. With the increase of CO2 concention of the engine intake manifold, number concentration in the accumulation mode increases, but number concentration in the nucleation mode decreases evidently.Effect of combustion modes (including compression ignition direct injection (CIDI), homogeneous charge compression ignition (HCCI) and compound charge compression ignition combustion (CCCI)) on NSDs of exhaust particles from engines fuelled with n-heptane and dimethyl ether (DME) was studied. The results indicate that the structure of NSDs for n-heptane changes with rp obviously. Accumulation mode dominantes in CIDI case. With the increase of premixed ratio(rp), the accumulation mode decreases, the nucleation mode increases significantly. In HCCI case, nucleation mode occupies the dominant part of the total particles, and the number concentration of accumulation mode is very low. And the total exhaust particles number concentrations for CCCI and HCCI are obviously higher than those of CIDI. In addition, the number concentration of nucleation mode is highly correlation with HC concentrations. The results of combustion mode study for DME indicate that the structure of NSDs for DME shows less change with rp at all test engine conditions. In all cases, nucleation mode occupies the dominant part of the total particles, and the number concentration of accumulation mode is very low. Under the same engine condition, the total exhaust particles number concentrations for CCCI are slightly lower than those of HCCI, but higher than those of CIDI significantly. The number concentrations of nucleation mode particles and total particles have a significantly positive correlation with HC concentrations.Field measurements of particle number and mass concentrations, CO concentration, and PM1 mass concentration were performed at 3 typical roads in Shanghai at daytime of 7:30~19:30. Mean diurnal variations show that the ultrafine particles concentrations, PM1 and CO concentrations change obviously with traffic situation. The total number and mass concentrations of particles increase with the traffic volume. Also, the mean particles concentrations in morning rush hour are evindently higher than those of afternoon rush hour for three cases. The CO average concentration and PM1 average concentration are highly correlation with total particle number average concentration (Ntotal). Moreover, the mean emission factor per vehicle was calculated, which is about 4.52×1014~8.50×1014 particles Km-1Vehicle-1.Transverse distribution characteristic of ultrafine particles at the downwind of a major road (Highway A4)were investigated. CFD numerical simulation coupling with the wind direction frequency weighted (WDFW) method was carried out based on the actual conditions. The results indicate that the peak concentrations of NSDs and MSDs decrease significantly with the increase of the distance. However, the diameters of peak concentration are in the same range. For NSDs, the peak concentration and the diameters of peak concentration of accumulation mode show less change. Meantime, it is found that the mean diameter of NSDs increases with the increase of the distance. The mean diameter for NSDs at 106m sampling site is 12nm larger than that of 17m sampling site. The CO average concentration is highly correlation with Ntotal(R2=0.78). From the results of CFD simulation, it was concluded that the decay rates of Ntotal and M total in field measurement case are larger than those of simulation case. And the number concentration decay mainly reflect the dacay of nucleation mode. The decay rates of simulated result and experimetal data can be described by a power function.

节点文献中: