As the society and economy in the world develops rapidly, the global demand for alternative energy sources increases with increasing depletion of fossil fuels and serious environmental pollution. Combustible gases are favored worldwide because of their potential high efficiency and low harmful emissions. From the point view of safety, however, combustible gases pose serious risk of accidental fire and explosions if they are not operated or used properly. In general, the combustible gaseous fuels are transported through pipelines or ducts in industries. When a gaseous fuel accumulates and mixes with air in pipelines or ducts during its production and use, explosion accident may occur. This is a safety concern worldwide. Therefore, understanding premixed flames propagating in ducts is of great importance for prevention and mitigation of gas explosions. This thesis aims to investigate the dynamics of premixed flame propagation in closed ducts and to explore effective suppression methods of premixed flame.

Firstly, a comparative study on the dynamics of typical premixed combustible gas fuels-air flame propagation in a closed duct was conducted. Four different combustible gases, including methane, natural gas, acetylene, and hydrogen were chosen. The high-speed schlieren photography system was applied to capture the flame shape changes and determine the flame tip speed, and the pressure transducer was adopted to record the pressure-time history. The results indicate that the characteristics of combustible gases influence the flame behaviors directly. Classical tulip flame forms in the equivalence ratio range of 0.79≤Φ≤1.30 for methane, 0.72≤Φ≤1.44 for natural gas, 0.40≤Φ≤1.70 for acetylene, and 0.60≤Φ≤5.56 for hydrogen. Distorted tulip flame forms in the equivalence ratio range of 1.00≤Φ≤2.38 for hydrogen, and at Φ=1.00 for acetylene. No distorted tulip flame was observed for methane and natural gas. Compared with pure methane, the small amount of ethane and propane that exists in natural gas helps accelerate flame propagation and increase pressure buildup. Meanwhile, acetylene shows a much faster flame propagation speed and higher pressure compared with natural gas and methane under similar conditions. Hydrogen has the fastest flame propagation speed and maximal pressure due to its high chemical reactivity. The theory proposed by Bychkov et al. for predicting the flame skirt motion coincides well with experimental data at equivalence ratios close to Φ=1.00. The results also shows that the Bychkov theory is more suitable for gases with high-chemical reactivity such as acetylene and hydrogen.

Methane and hydrogen were chosen as the representatives of low and high-chemical reactivity gases in the subsequent experiments of flame suppression using metal wire mesh. The effects of single-layer metal wire mesh and multi-layer metal wire mesh on premixed flame propagation were studied using the premixed flame suppression experimental system. The results demonstrate that metal wire mesh makes the tulip flame formation time advanced and makes the flame front inversion extent weaker compared with the case of no wire mesh. This can be attributed to the effects of metal wire mesh on advancing and weakening the interactions between flame front and pressure waves. In addition, the metal wire mesh also strengthens the disturbance of flow and generates a violent combustion after the flame propagates through the suppression zone. For the single-layer wire mesh, the quenching performance does not increase as mesh density increases. The reason is that the anti-destructive performance is also quite important in practical applications. No suppression effect on flame propagation in the upstream duct is obtained because the decrease of gas flow speed in unburned field is balanced by the increase of burning velocity of combustible gases. Multi-layer wire mesh can attenuate the maximum flame tip speed, maximum pressure, and maximum sound effectively, and the suppression effect increases as the mesh and layer increase.

Then, the factors affecting flame quenching under the effects of multi-layer metal wire mesh were analyzed systematically, including the characteristics of the combustible gases, initial temperature and pressure, dopants in combustible gases, equivalence ratio, volume of wire mesh, arrangement style of wire mesh, and ignition position. It was found that the characteristics of combustible gases, initial temperature and pressure, dopants in combustible gases, and equivalence ratio can influence the flame quenching through impacting on the laminar burning velocities. The relationship between critical quenching parameters and volume of wire mesh is revealed. The critical quenching speed increases almost linearly as the volume of metal wire mesh increases. However, the maximum critical quenching pressure maintains at a constant value of about 0.115 MPa. Metal wire mesh performs a better suppression effect on rich-fuel combustion than on lean-fuel combustion. Besides, it was found that the flame quenching performance of metal wire mesh can be changed through varying the spacing between the wire mesh. Whether the flame can be quenched in the suppression zone depends on the competition between the flame accelerating effects and the quenching effects of sidewalls. The ignition positions can affect the flame quenching significantly. Three different ignition positions were adopted to ensure the flame propagates into the suppression zone at three different stages, i.e., with tulip flame completely formed in ignition position 1, just before tulip flame formation in ignition position 2, and with finger shape flame in ignition position 3. It was found that the closer the ignition position is to the metal wire mesh, the easier the flame is quenched. When the flame is initiated at the closed end of the duct, tulip flame formation is essentially produced by the vortices generated by the interactions between pressure waves and flame front. While, for the cases of flame initiating from the center of the duct, tulip flame is mainly attributed to the reverse gas flow in burnt region enhanced by the great discrepancies of combustion intensity between the flame propagating toward upstream and downstream sections of the duct. Meanwhile, the results demonstrate that both of the flame propagating toward upstream and downstream sections are of great importance in determining the pressure dynamics in the cases of flame initiating from the center of the duct.

Finally, the coupling effects of inert gas and metal wire mesh on premixed flame propagation were discussed. It was found that CO₂ has a more effective suppression effect on premixed hydrogen-air flame than N₂. In rich-fuel combustion, CO₂-5% has little suppression effect on flame tip speed and pressure because it has almost no influence on flame surface areas. However, both of flame tip speed and pressure are reduced significantly under the effect of CO₂-5% in lean-fuel cases. As the concentration increases in the range between 0% and 25%, the suppression effect increases continuously, but becomes almost constant when the concentration changing from 25% to 30%. The increase of CO₂ concentration is able to suppress the hydrodynamic instability and reduce flame disturbance. As for thermal diffusion instability, the increase of CO₂ concentration in lean fuel combustion makes Lewis number close to unity and reduce the flame disturbance. On the contrary, thermal diffusion instability itself has a suppression effect on DL instability in rich-fuel cases. The addition of CO₂ only contributes to strengthening this effect and reducing flame disturbance further. The coupling effects of CO₂ dilution and metal wire mesh on premixed hydrogen-air flame is more effective than that using either of the two suppression agents separately. CO₂ dilution promotes the suppression effects of metal wire mesh on lean-fuel combustion, and metal wire mesh improves the suppression effects of CO₂ dilution on rich-fuel combustion. Moreover, the suppression effects of CO₂ dilution with a higher concentration on premixed hydrogen-air flame can be still enhanced by the addition of metal wire mesh.