Diesel soot formation

Previous investigations show that soot particle volume fraction and number density were significantly reduced by exhaust gas recirculation (EGR) diluents CO 2 and H 2 O. However, these investigations were often convoluted by their experimental flame configurations and primarily focused on soot volume fraction rather than soot inception. To isolate the effects on soot inception and the corresponding chemistry, the current study measured the reactivity of CO 2 (up to 9.5% volume fraction) for both C 2 H 2 (1.00% volume fraction) and CH 4 (1.85% volume fraction) fuels in homogeneous mixtures. Computed effect of H 2 O on these and other fuels are also presented. Experiments were performed at high temperature (1640 K and 1770 K) and high equivalence ratios (U = 55 and 75) to understand the effect of CO 2 on polycyclic aromatic hydrocarbons (PAH) and formation of nascent soot particles with negligible oxygen influence. Experimental results show that CO 2 enhanced the soot inception rate when added to C 2 H 2 but had an undetectable affect on CH 4 . Gas chromatography confirmed that CO 2 increases CO mole fraction and reduces C 2 H 2 fuel concentration. Chemical kinetic simulations showed that the C 2 H 2 was being converted to soot precursors. CO 2 enhanced the soot inception rate for C 2 H 2 by producing OH radicals. Images of nascent soot particles produced in the presence of CO 2 were used to determine the size of PAH molecules in the particles and particle morphology. Both attributes were similar to particles formed without CO 2 . CO 2 had little impact on the long reaction pathway from CH 4 to PAH molecules because H and CH 3 radicals propagated these reactions more readily than OH radicals.

Structure and chemistry of soot extracted from the crankcase and cylinder wall and tribofilms on piston rings of a Mack T-12 engine test were examined. XANES spectra recorded on soot samples indicate the presence of thermo-oxidative and tribochemical decomposition products in engine oil indicative of interaction between piston ring and soot samples. HRTEM indicates the presence of nanoparticles of calcium phosphate compounds, ZnO and Fe 2 O 3 embedded at the surface of the turbostratic soot structure possibly responsible for third body wear. Abrasive grooves were found on the piston ring indicative of abrasive wear mechanism as a dominant wear phenomenon. In addition, significant differences in chemical composition of crankcase soot, cylinder soot and piston ring tribofilms were elucidated.

a b s t r a c t CeO 2 nanofibres have been prepared by means of the co-precipitation and ripening method. The morphology of the CeO 2 and the synthesis duration were influenced by changes in the mole ratio of the NaOH/citric acid. Three kinds of precipitates were obtained by increasing the NaOH/citric acid ratio: stuck-in-bundle, fibre and flake morphologies. The cobalt was impregnated on the prepared CeO 2 (0.8) nanofibres and the physico-chemical properties, soot combustion activities and deactivation were investigated. The Co/CeO 2 (0.8) nanofibres were evaluated for soot combustion in O 2 and NO/O 2 . The morphology of the fibrous structure of CeO 2 (0.8) and Co/CeO 2 (0.8) was observed through FE SEM images. The presence of cobalt in Co 3 O 4 form was confirmed by means of XRD and TPR analyses. The Co/CeO 2 (0.8) fibre catalysts showed improvements in soot combustion at low temperatures of around 400 • C with 10% O 2 and a decrease in the peak soot combustion temperature at about 40 • C was observed with the addition of 500 ppm NO. The XPS results of the Co 2p spectra of the Co/CeO 2 (0.8) catalysts confirmed the presence of Co in the Co 3+ , Co 2+ oxidation states. The thermal stability and the resistance to SO 2 poisoning of the 10% Co/CeO 2 (0.8) fibre catalyst were examined by ageing it in air and 1030 ppm SO 2 at 800 • C for 12 h.

A hybrid model for accurate and computationally efficient simulations of the particle trapping characteristics of automotive flow-through catalysts is suggested in this paper. The new model is validated against the performance of a more elaborate, but computationally far more expensive model. In this hybrid model, the trapping of the smallest particles is predicted using a computationally efficient submodel that can also be used for screening of new catalyst substrate designs. It is shown here that this screening model is very accurate for particles smaller than approximately 50 nm. A number of different catalyst designs are evaluated and compared using the screening model. In particular, the performance of a promising channel design with porous obstacles is evaluated. This design could potentially give over 70% reduction of small soot particles without a substantial increase in the pressure drop.

A complete understanding of soot particle formation is critical for accurate combustion models. Limited details are known about the molecules and physical processes involved in the formation of nascent soot. The objective of this study was to discover molecular size distribution and particle morphology from freshly nucleated particles. To accomplish this, particles created in a homogeneous, high temperature (1600-2000 K) and pressure (10 atm) environment were collected at well-resolved times throughout the early stages of soot formation and were imaged with a high-resolution transmission electron microscope (HRTEM). Image post processing revealed that the molecules deposited on both nascent and aged soot particles had a constant size distribution, and that the molecules were likely to be polycyclic aromatic hydrocarbon (PAH) molecules. Additionally, the images revealed particle morphology. At low temperatures, PAH molecules coagulated to form fairly amorphous particle cores. At high temperatures the particle cores became comprised of agglomerated 2-8 nm diameter particles. Eventually all particle cores became surrounded by layers of molecules while the small particles no longer attached as growth species. The imaged transition from small particle agglomeration at the particle core to surface growth was replicated in a coagulation model. The model quantified the increase in small particle agglomeration that was attributed to an increase in the coagulating molecule's rate of production.

The activity and stability of Ba,K,Co/CeO 2 catalyst in the diesel soot oxidation reaction are studied. Different modes of preparation are analyzed, varying both the cobalt precursor and its order of impregnation. In order to study the stability in a diesel exhaust atmosphere, the catalysts were pretreated in streams of CO 2 , H 2 O, NO and SO 2 . The fresh and the treated catalysts were characterized by FTIR and XRD techniques. The catalytic activity was measured by TPO of soot-catalyst mixtures. The tight contact was used to analyze the intrinsic activity. It was found that the activity was higher for the catalysts prepared using Co(NO 3 ) 2 compared with the catalysts prepared using Co(AcO) 2 . This is because in the former, KNO 3 is present on the catalyst, being this compound very active for this reaction. The thermal stability is lower for the catalyst prepared with Co(NO 3 ) 2 . This catalyst displays a higher K lost when treated at high temperature. When K is present as K 2 CO 3 , as is the case of the catalysts prepared with Co(AcO) 2 , the thermal stability is higher since K 2 CO 3 is less volatile than KNO 3 . All the catalysts are stable in the presence of mixtures of (CO 2 + H 2 O + NO + O 2 ) having a composition similar to a real diesel exhaust. Under these conditions, the K mantains its original chemical state, either carbonate or nitrate, depending on the precursor used in the catalyst preparation. In presence of SO 2 , all the catalysts deactivate due to K 2 SO 4 formation, which is not active for soot combustion. However, the sensitivity to SO 2 depends on the precursors used to prepare the catalyst. The Ba,K,Co/CeO 2 catalyst prepared impregnating Co(NO 3 ) 2 on the Ba,K/CeO 2 catalyst has a higher resistance to the deactivation by SO 2 , since the following reaction occurs: Ba(NO 3 ) 2 + K 2 SO 4 → BaSO 4 + KNO 3 , which implies that active KNO 3 will dissapear slower from the catalytic surface, thus maintaining the activity. TPO experiments of the catalyst in loose contact with soot were also carried out. The Ba,K,Co/CeO 2 catalyst prepared impregnating Co(NO 3 ) 2 on the Ba,K/CeO 2 catalyst showed the higher activity in loose contact mode.

Previous investigations show that soot particle volume fraction and number density were significantly reduced by exhaust gas recirculation (EGR) diluents CO 2 and H 2 O. However, these investigations were often convoluted by their experimental flame configurations and primarily focused on soot volume fraction rather than soot inception. To isolate the effects on soot inception and the corresponding chemistry, the current study measured the reactivity of CO 2 (up to 9.5% volume fraction) for both C 2 H 2 (1.00% volume fraction) and CH 4 (1.85% volume fraction) fuels in homogeneous mixtures. Computed effect of H 2 O on these and other fuels are also presented. Experiments were performed at high temperature (1640 K and 1770 K) and high equivalence ratios (U = 55 and 75) to understand the effect of CO 2 on polycyclic aromatic hydrocarbons (PAH) and formation of nascent soot particles with negligible oxygen influence. Experimental results show that CO 2 enhanced the soot inception rate when added to C 2 H 2 but had an undetectable affect on CH 4 . Gas chromatography confirmed that CO 2 increases CO mole fraction and reduces C 2 H 2 fuel concentration. Chemical kinetic simulations showed that the C 2 H 2 was being converted to soot precursors. CO 2 enhanced the soot inception rate for C 2 H 2 by producing OH radicals. Images of nascent soot particles produced in the presence of CO 2 were used to determine the size of PAH molecules in the particles and particle morphology. Both attributes were similar to particles formed without CO 2 . CO 2 had little impact on the long reaction pathway from CH 4 to PAH molecules because H and CH 3 radicals propagated these reactions more readily than OH radicals.

The use of color-ratio pyrometry (CRP) methods, with variable or prescribed soot content (KL) to image flame–wall interactions was examined, with results compared with that obtained using the more mature two-color pyrometry (TCP) technique. The CRP and TCP methods were applied to flame–wall impingement images recorded in a optically-accessible constant volume combustion chamber (CVCC) under compression-ignition (CI) engine conditions. Good correlation in the result trends were observed for the CRP method with fixed KL output and that generated using TCP. Slight discrepancies in the predicted absolute temperature values were observed, which were linked to the difference in the KL value prescribed to the CRP method, and the KL value predicted using TCP. No useful output was obtained with CRP method with variable soot output because of channel noise. A simplified flame transparency modeling was performed to assess the inherent errors associated with the pyrometry methods. The results indicated that the uncertainties arising from the fixing of the KL output appeared acceptable.

The aim of this study is to investigate the impact of walls on soot processes of a post-injection strategy at different dwell times. The experiments were performed in an optically accessible constant-volume combustion chamber simulating compression ignition engine conditions with moderate exhaust gas recirculation. The experiments with various injection strategies were performed under ambient conditions with gas density, pressure, and temperature of 20.8 kg/m3, 6 MPa, and 1000 K, respectively, and 15 vol % O2 concentration. The main and post injections had a quantity ratio of 8:2 (main/post) totaling 10 mg, and a flat wall was placed 35 mm axially from the injector. The dwell time between the main and post injections was also varied to induce different levels of interaction between the injections. High-speed flame natural luminosity imaging and two-color pyrometry techniques were applied to observe flame characteristics and to obtain soot temperature and KL factor information, respectively. By comparing the wall jet and free jet cases with no direct jet interaction, it was found that the wall affected the post jet flame structure similarly to a single jet or the main jet. However, the post jet with a greater extent of interaction with the main jet induced by shorter dwell time can achieve better mixing for the wall jet case. Increased interaction between the main and post jets also appeared to induce a soot oxidation phase, which was otherwise not observed when the injections were more temporally separated.

Previous investigations show that soot particle volume fraction and number density were significantly reduced by exhaust gas recirculation (EGR) diluents CO 2 and H 2 O. However, these investigations were often convoluted by their experimental flame configurations and primarily focused on soot volume fraction rather than soot inception. To isolate the effects on soot inception and the corresponding chemistry, the current study measured the reactivity of CO 2 (up to 9.5% volume fraction) for both C 2 H 2 (1.00% volume fraction) and CH 4 (1.85% volume fraction) fuels in homogeneous mixtures. Computed effect of H 2 O on these and other fuels are also presented. Experiments were performed at high temperature (1640 K and 1770 K) and high equivalence ratios (U = 55 and 75) to understand the effect of CO 2 on polycyclic aromatic hydrocarbons (PAH) and formation of nascent soot particles with negligible oxygen influence. Experimental results show that CO 2 enhanced the soot inception rate when added to C 2 H 2 but had an undetectable affect on CH 4. Gas chromatography confirmed that CO 2 increases CO mole fraction and reduces C 2 H 2 fuel concentration. Chemical kinetic simulations showed that the C 2 H 2 was being converted to soot precursors. CO 2 enhanced the soot inception rate for C 2 H 2 by producing OH radicals. Images of nascent soot particles produced in the presence of CO 2 were used to determine the size of PAH molecules in the particles and particle morphology. Both attributes were similar to particles formed without CO 2. CO 2 had little impact on the long reaction pathway from CH 4 to PAH molecules because H and CH 3 radicals propagated these reactions more readily than OH radicals.

This work aims to assess the effects of flame-wall impingement on the combustion and soot processes of diesel flames. For this work, experimental measurements were performed in a constant-volume combustion chamber (CVCC) at ambient conditions that are representative of compression-ignition engines. The characteristics of
impinging and free flames were compared at two identical ambient and injector conditions (20.8 kg/m3 ambient density, 6 MPa ambient pressure, 1000 K bulk temperature, 15 and 10 vol% ambient O2 concentration, and 100 MPa injection pressure). To simulate flame-wall impingement, a flat plane steel wall, normal to the injector axis, was initially placed at 53 mm from nozzle, but was varied from 53 to 35 mm during the experiments. Under the test conditions of this work, it was found that wall impingement resulted in lower soot temperature and soot content, in addition to a loss of momentum for the wall jet. The results also revealed that decreasing impingement distance from the nozzle resulted in reduced soot temperature and soot level for the wall jet. The reduced soot content observed for the wall jet appeared to be mainly driven by enhanced mixing. Flame transparency modeling was also performed to assess the uncertainties of two-color measurements for flameplane wall impingement. The analysis indicated that the derived soot temperature and concentration values would be affected by the actual temperature profiles, rendering the technique useful to reveal trends, but not reliable for absolute soot concentration measurements.

Diesel soot overloads the ELPI-impactor rapidly if it is equipped with the standard at-surface impactors. This non-ideal behaviour was studied recently (J. Aerosol Sci. 32 ). It was found that rapid overloading, or surface build-up, is a result of a u y bed of soot particles that covers the impactor surfaces and starts to ÿlter airborne soot particles. This paper reports additional results with oil-soaked sintered impactors for the ELPI. It is demonstrated that (rapid) overloading is eliminated with oil-soaked sintered impactors. The maximum allowed mass load for the ELPI impactor is increased 50-fold. ?

The catalytic behavior of Ag-loaded fumed SiO 2 has been investigated for diesel soot oxidation. The diesel soot generated by burning pure Mexican diesel in laboratory was oxidized under air flow in presence of the catalyst inside a tubular quartz reactor in between 25 and 600 • C. UV-vis optical spectroscopy was utilized to study the electronic state of silver in Ag/SiO 2 catalysts. The soot oxidation was seen to be strongly enhanced by the presence of metallic silver on the catalyst surface, probably due to the formation of superoxide ion O 2 − . During the catalytic process, the Ag 2 O species on the catalyst acts as a strong oxidizing agent which gets reduced readily by the soot carbon to metallic Ag. The Ag/SiO 2 catalysts exhibited strong performance for diesel soot oxidation below 300 • C. Mechanical mixing of extra SiO 2 with 3% Ag loaded SiO 2 resulted a further increase of soot oxidation rate, probably due to the increased interactions of O 2 − and soot particles through the highly porous surface of SiO 2 .

The catalytic behavior of Ag-loaded fumed SiO 2 has been investigated for diesel soot oxidation. The diesel soot generated by burning pure Mexican diesel in laboratory was oxidized under air flow in presence of the catalyst inside a tubular quartz reactor in between 25 and 600 • C. UV-vis optical spectroscopy was utilized to study the electronic state of silver in Ag/SiO 2 catalysts. The soot oxidation was seen to be strongly enhanced by the presence of metallic silver on the catalyst surface, probably due to the formation of superoxide ion O 2 − . During the catalytic process, the Ag 2 O species on the catalyst acts as a strong oxidizing agent which gets reduced readily by the soot carbon to metallic Ag. The Ag/SiO 2 catalysts exhibited strong performance for diesel soot oxidation below 300 • C. Mechanical mixing of extra SiO 2 with 3% Ag loaded SiO 2 resulted a further increase of soot oxidation rate, probably due to the increased interactions of O 2 − and soot particles through the highly porous surface of SiO 2 .

Simultaneous laser-induced incandescence (LII) and laser-induced scattering (LIS) were applied to investigate soot formation and distribution in a single cylinder rapid compression machine. The fuel used was a low sulfur reference diesel fuel with 0.04% volume 2-ethylhexyl nitrate. LII images were acquired at time intervals of 1 CA throughout the soot formation period, for a range of injection pressures up to 160 MPa, and in-cylinder pressures (ICP) up to 9 MPa. The data collected shows that although cycle-to-cycle variations in soot production were observed, the LII signal intensities converged to a constant value when sufficient cycles were averaged. The amount of soot produced was not significantly affected by changes in in-cylinder pressure. Soot was observed distributed in definite clusters, which were linked to slugs of fuel caused by oscillations in the injector needle. The highest injection pressure exhibited lower soot productions and more homogeneous soot distributions within the flame. Despite diffusion flames lasting longer with lower injection pressure, it appeared that the extended oxidation time was insufficient to oxidize the excess production of soot. In addition, soot particles were detected closer to the nozzle tip with higher injection pressures. The recording of LII sequences at high temporal resolutions has shown that three distinct phases in soot formation can be observed. First, high soot formation rates are observed before the establishment of the diffusion flame. Second, a reduced soot formation rate is apparent from the start of diffusion flame until the end of injection. Finally, high soot oxidation rates occur after the end of injection and for the duration of the flame.

A complete understanding of soot particle formation is critical for accurate combustion models. Limited details are known about the molecules and physical processes involved in the formation of nascent soot. The objective of this study was to discover molecular size distribution and particle morphology from freshly nucleated particles. To accomplish this, particles created in a homogeneous, high temperature (1600-2000 K) and pressure (10 atm) environment were collected at well-resolved times throughout the early stages of soot formation and were imaged with a high-resolution transmission electron microscope (HRTEM). Image post processing revealed that the molecules deposited on both nascent and aged soot particles had a constant size distribution, and that the molecules were likely to be polycyclic aromatic hydrocarbon (PAH) molecules. Additionally, the images revealed particle morphology. At low temperatures, PAH molecules coagulated to form fairly amorphous particle cores. At high temperatures the particle cores became comprised of agglomerated 2-8 nm diameter particles. Eventually all particle cores became surrounded by layers of molecules while the small particles no longer attached as growth species. The imaged transition from small particle agglomeration at the particle core to surface growth was replicated in a coagulation model. The model quantified the increase in small particle agglomeration that was attributed to an increase in the coagulating molecule's rate of production.