Search

The dynamics of a laminar premixed stagnating flame subjected to simultaneous velocity and equivalence ratio oscillations are investigated experimentally. A flame of premixed methane and air is formed in a wall-stagnating flow. The oscillations of velocity and equivalence ratio are imposed independently by two sets of dual cylinder–piston oscillators. To impose the equivalence ratio oscillation, methane/air mixtures with equivalence ratios of $$\phi =1.0$$ and $$0.4$$ are supplied to different cylinder–piston units. The pistons move in anti-phase and create a sinusoidal equivalence ratio oscillation while keeping the volume flow rate constant. To impose the velocity oscillation, a mixture with $$\phi =0.7$$ is supplied to the single unit of another oscillator. The phase difference between the two oscillations is set by changing the relative position of the pulley teeth of the two oscillators. The oscillator frequency is varied between 2 and 20 Hz, meaning that the oscillation wavelengths are much longer than the flame thickness. When frequencies of the velocity and equivalence ratio fluctuations are in a quasi-steady state condition, the flame motion is affected independently by velocity oscillation and equivalence ratio oscillation and the linear-superposition assumption can be applied. When the velocity and equivalence ratio oscillation frequencies are increased and exceed the transition frequency from the quasi-steady state to unsteady state, the flame motion gradually deviates from linear-superposition assumption. This is due to the variations in the unsteady molecular diffusion and the back-support effect.

The dynamic behavior of the near-field region in a coaxial variable property jet has been experimentally investigated under a swirling flow produced by rotating cylindrical inner and outer tubes, focusing on how the swirl of the outer jet affects the formation of a stagnation point in the swirling inner jet. The inner and outer jets rotate in the same direction. Air, CO2, or He is issued from the inner tube as a variable property jet, and air is issued from the outer tube in this work. In the case of a CO2 jet (a high-density, low-viscosity gas jet), a stagnation point flow is more easily formed than in the case of an air jet, and the stagnation point location is significantly lower than in that of the air jet. When the swirl of the outer jet is introduced, a stagnation point flow is more easily formed than in the case of a nonswirling outer jet, and the stagnation point location is much lower than in the case of a nonswirling outer jet. In the case of a He jet (a low-density and high-viscosity gas jet), the inner jet does not have a stagnation point flow, and its overall behavior remains nearly unchanged even under high swirl numbers of the inner and outer jets. These results clearly show that the density and viscosity differences between the inner and outer jets have a significant impact on the dynamic behavior of the near-field region in the coaxial swirling jet. The significant lowering of the stagnation point location can be physically explained by considering the theoretical equation obtained in this work.

The near-field instability of variable property jets of air, CO2, and He, issued into the ambient air, has been investigated experimentally within normal gravity and microgravity fields. The density ratio to the ambient air is unity for air jets, more than unity (1.53) for CO2 jets, and less than unity (0.14) for He jets, respectively. The ratio of kinematic viscosity to the ambient air is unity for air jets, less than unity for CO2 jets (0.53), and more than unity for He jets (7.75), respectively. The jet velocity is varied from 0.4 to 1.8 m/s and then the jet Reynolds number varies from 60 for Helium jet to 2,000 for CO2 jet, while the Richardson number varies from negative to positive values. The motion of the jet is visualized using a laser tomographic method and recorded by a high-speed digital video camera with 250 frames/s. The result shows that the instability of the jet is intensified when Re > 800 while it is weakened at Re

The combustion behavior of a methane hydrate sphere under normal gravity is experimentally investigated. The initial diameter of the sphere is 20 mm. Variation in temperature at the center of the sphere (T c) is measured with a K-type thermocouple at ignition temperatures (T c,i) from 193 to 253 K at 20 K intervals. Variation in the near-surface temperature of the sphere (T s) is measured at ignition temperatures (T s,i) from 233 to 263 K at 10 K intervals. Two combustion phases are observed. When the hydrate is ignited, a stable flame envelope is formed around the sphere (phase 1). In phase 1, the surface of the sphere is dry. After a few seconds, water formed by dissociation of the methane hydrate appears on the surface and methane bubbles are formed by methane ejected from inside the sphere (phase 2), thus destabilizing the flame and causing local extinction. Methane bubbles move down along the surface and merge into a large methane bubble at the bottom of the sphere. This bubble bursts, releasing methane to form a temporary flame, and the water drops from the hydrate sphere. Water on the surface is cooled by the hydrate inside, and an ice shell confines the methane gas that dissociated inside the sphere. Because the dissociation occurs continuously inside the hydrate, the inner pressure gradually increases and at some instant, the ice cracks and methane gas is ejected from the cracks, which results in a micro-explosion with a flame. In phase 1, the surface temperature is below the freezing point of water, and so the surface remains dry and a stable flame envelope is formed; in phase 2, the surface temperature is above the freezing point, and so water appears on the surface. When the temperature at the center of the sphere is lower (193, 213, or 233 K), some methane hydrate remains even after flame extinction because heat transfer from the flame decreases in phase 2 as a result of local extinction. The diameter of the sphere decreases during combustion in accordance with the d-square law, which indicates that the heat of dissociation is supplied by the flame and methane is supplied by the dissociation of methane hydrate in the sphere, as in single-droplet combustion.

This structure of turbulent flow in an annulus with strong inner cylinder wall heating has been studied in terms of velocity and temperature with wall temperatures up to 707 °C and a Reynolds number of 48,000. With increase in wall heating, the turbulence very close to the wall was suppressed due to an increase in the kinematic viscosity. In the inner region, the intermittent mixing became intensive and the turbulent intensity increased whereas, in the outer region, the turbulence was suppressed since intermittent mixing was no longer effective. The results show that the thermal structure can be considered in terms of a passive scalar for wall temperatures less then around 200 °C, except in the leading region of the heated area.

Velocity statistics along the stagnation line of an axi-symmetric wall stagnating turbulent flow are studied experimentally. A low turbulence, uniform air flow from a nozzle type air supply with an exit diameter of 50 mm stagnates at a wall located 50 mm downstream. A flow velocity is set to 3 m/s, 10 mm downstream from the exit of the air supply. Instantaneous values of streamwise and radial velocities are measured by laser-Doppler velocimetry. The turbulence level in the air flow is changed by use of turbulence generator. When the turbulence generator is not installed in the air supply, the mean velocity profile in the streamwise direction fits well with that of a laminar viscous flow with the rms value of velocity fluctuations low near the wall. With the turbulence generator installed, a significant turbulence structure appears near the wall. When the wall is approached, the rms value of velocity fluctuations in the streamwise direction decreases monotonically while the profile of the rms value in the radial direction reaches a maximum near the wall. The increase in the rms value of velocity fluctuations in the radial direction near the wall is attributed to the bi-modal histogram of the fluctuating velocity in the radial direction. Near the wall, the instantaneous stagnation streamline fluctuates and the probability of the mean location of the stagnation point reaches a maximum not at the stagnation line but on a circle around the stagnation line, resulting in the bi-modal histogram. Turbulence statistics, the rms value of velocity fluctuation and the turbulent kinetic energy, can be normalized successfully by similarity parameters based on the strain rate and the reference turbulent kinetic energy introduced by Champion and Libby.

Velocity-temperature correlations in a strongly heated channel flow were investigated experimentally by a LDV and a resistance thermometer. The wall heat flux is varied up to 50,000 W/m2 with reference mean-velocity of 15 m/s, and then, the wall temperature reaches up to 1,000 K. The results show that the ejection fluid motion is intensified by the strong heating near the wall increasing the turbulent heat flux from the wall. The intensification of ejection motion balances the destruction of turbulent heat flux. Then, the overall turbulent heat transfer does not change clearly.

Turbulence statistics of a channel flow with strong wall heating at Reynolds number of 14, 000 were investigated experimentally. The statistical quantities up to triple correlations were estimated, using data measured by a single component LDV. The wall temperature varies up to 1000 K. Rms values of streamwise and cross-stream velocities and their correlation decrease in the thermal boundary layer except near the wall. Near the wall, the ejection of low-speed fluid outward from the wall is intensified due to thermal expansion, and it cancels the decrease in statistical quantities. As a result, the statistical quantities do not vary apparently.

A laser Doppler velocimeter and a resistance thermometer were used to study velocity and temperature statistics in a strongly heated turbulent two-dimensional channel flow, with the wall temperature up to 700 °C and a Reynolds number of 14,000. Normalized mean velocity and mean temperature profiles were not significantly affected by the wall heating. Turbulent intensities of temperature fluctuation were also insensitive to the heat flux. However, turbulent intensities of velocity fluctuation were suppressed in the region away from the wall, whereas those near the wall were not changed noticeably by the wall heating. This phenomenon was explained by the balance of three parameters: turbulent production, viscous dissipation and intermittency.