Fundamentals of the propagation characteristics of a detonation are reviewed. First, the propagation characteristics of a detonation are discussed from the one-dimensional point of view. In this discussion, the characteristics of a steadily-propagating detonation and how to determine its static parameters are elucidated. Second, the non-planarity and transverse-wave structure of an actual detonation front are discussed. The transverse waves propagating along the leading shock wave of a detonation front and colliding with each other depict a cellular pattern after the passage of the detonation. The importance of the characteristic size of this cellular pattern is emphasized introducing so-called dynamic parameters of detonations. Finally, the stability issue of the transverse-wave structure of a detonation front is touched on. This issue has been drawing the attention of many researchers for more than a decade.

Generally combustion of premixed mixtures is divided into two modes; deflagration and detonation. These two modes have completely different properties and can be easily distinguished from their travelling speed and propagation mechanism. If a mixture is ignited by a weak ignition source, a deflagration wave is obtained. Under appropriate conditions, a deflagration wave accelerates, resulting in an abrupt transition into a detonation wave. This type of detonation initiation is referred to as deflagration to detonation transition, DDT. In this paper, one-dimensional depiction of DDT phenomena is made, which is followed by the milestone visualization of onset of detonation by Urtiew and Oppenheim. Some reviews are focused on DDT in a smoothed tube and in an obstacle laden tube, including recent progress in these fields. Mechanism of detonation initiation is mentioned on the basis of new findings in numerical works.

First of all in this paper, I show a brief history of a pulse detonation engine research. Thermodynamic analyses for detonation combustion were described, and it was shown that the thermal efficiency of the detonation combustion cycle is higher than the Brayton one. The typical pulse detonation engine operation and gasdynamics models for the simplified pulse detonation engine were illustrated. I also presented recent researches regarding heat and friction losses on a PDE tube wall. The partially-filling effect, ejector effect and nozzle effect on thrust argumentation for PDEs were introduced. As important technical issues of PDEs, I show a deflagration-to-detonation transition process control and gas supplying systems in PDEs. As recent applied researches for PDEs, an air-breathing PDE, a pulse detonation rocket engine, a pulse detonation turbine engine, a PDE thermal spray, and a micro PDE were reported. Finally a rotating detonation engine was introduced.

Rotating or continuous detonation engine study is reviewed in this paper. First of all the history of rotating detonation engine study is presented and the mechanism of rotating detonation engine is discussed through the recent study. Especially how detonation front rotates in a coaxial cylindrical tube is shown through numerical analyses. Then the experimental system and application of rotating detonation engine are described from the experimental and numerical outcomes recently obtained. The future possibility of the real rotating detonation engine will be discussed at the last.

Small scale combustors have received keen attention as a power source for portable devices. The vortex combustion is one of the key techniques to stabilize the flame in a narrow channel; in fact, small scale vortex combustors have been successfully developed for gaseous and liquid fuels. The developments of those small scale vortex combustors triggered fundamental studies on the flame behavior in small scale vortex flows. In this paper, results of the fundamental studies were introduced. In meso-scale vortex flow, it was found that a flame could propagate rapidly within the meso-scale tube by the vortex bursting mechanism, even when the tube diameter was close to the quenching diameter. Up to now, the flame propagation limits due to the vortex bursting mechanism were found to be greatly affected by the Lewis number effect. In this paper, the development of the practical power system with the thermo-electric device was also introduced. Utilizing the vortex flow combustor, the electric output 1.1W was obtained with the system which was several times larger than that of the catalytic combustor power system.

Development of a microcombustor with platinum porous catalyst layer and a cogeneration system using thermoelectric (TE) modules is introduced. In order to overcome the huge heat loss due to the small size of combustion field, catalyst combustion has been adopted because it has great advantage under the condition where the surface/volume ratio becomes large. To develop the monolithic porous structure of platinum, platinum powder is sintered in fuel-air mixture; the heat from the surface reaction melts the particles and makes agglomerates. The combustor is 5W class, but the heat release density of the combustor is 5GW/m³, which is comparable to that of conventional turbulent combustor. We couple the combustor with TE modules to convert the thermal energy to electricity. The TE modules are also preferable for small size generators because their characteristics that they use the heat flux from the hot zone to the surroundings match microcombustion. The output of the developed cogenerator is 165mW for thermal input of 4.7W; the final conversion ratio from fuel enthalpy to electricity reaches 3.5%. After integrating the micro-blower to supply air to the combustor, the cogeneration system shows the final efficiency of 1.7%.

Diffusion combustion (non-premixed combustion), which is formed by supplying fuel and oxidizer separately, is used in many practical combustors. However, we cannot control the flame temperature, because the diffusion flame is formed in the region at the stoichiometric condition. Hence, it is difficult to reduce the combustion products such as soot and NOx. So far, we have investigated combustion field in a triple port burner. In the triple port burner, since there are two boundaries of fuel and air, two flames are formed. Then, by changing the flow condition, four flame configurations are observed, which are, (i) attached flames, (ii) inner lifted/outer attached flames, (iii) inner attached/outer lifted flames, (iv) twin lifted flames. For further study, we investigated the slit burner which has air nozzles on both sides of a fuel nozzle, which is similar to the triple port burner. In experiments, we measured the flow field and flame structure by PIV/OH-PLIF simultaneous measurement. We also investigated hysteresis characteristic of the lifted flame in the slit burner.

Water mist is expected to have physical and chemical effects on the laminar flame speeds. In the present study, the effect of water mist on the flame speed and structure of propane-air premixed flames is numerically simulated by using PREMIX code of CHEMKIN package, modified to include the evaporation process of water mist. Evaporation process was assumed to follow the Arrhenius law. Chemical kinetic models for propane oxidation were evaluated by comparison of simulated flame speeds with experimental data for the case without water mist. The effects of water mist on flame speeds are separated into the dilution and chemical effects of water vapor, and the thermal effect which includes the heat of evaporation of water mist. The most effective is the sensible heat of water vapor, which is followed by the heat of evaporation. The chemical effect is relatively small but cannot be neglected. When the water mist is added, the flame temperature decreases due to thermal effect which reduces the rates of chemical reactions involving the radicals such as O and H, which have the positive sensitivity of flame speed. Furthermore, three-body reactions involving H₂O are enhanced. These reactions have high negative sensitivity of flame speed due to high chaperon efficiency of H₂O.

Assessment of a reacting flow solver with large detailed chemical kinetics is extensively performed in terms of its efficiency and capability. The present method solves the compressible Navier-Stokes equations with the chemical reaction source terms in the operator-splitting form, i.e., the chemical reaction and fluid parts are solved separately during one time step. For the chemical reaction, a dynamic multi-times scale (MTS) method is introduced for alleviating the stiffness. Several zero- to two-dimensional combustion problems with methane, n-butane, and n-heptane reaction mechanisms are used for the assessment of the present method. The ignition problems with the three reaction mechanisms demonstrate that the present method provides the higher efficiency with smaller time step size and larger number of chemical species, compared to a conventional implicit time integration method (VODE). The present method with MTS is 2∼30 times faster than the method with VODE for the ignition problems. The one-dimensional end-gas auto ignition problems with methane and n-butane reaction mechanisms also demonstrate the higher efficiency and the capability of the present method for capturing the interaction between combustion and compressibility, e.g., engine knocking-like behaviors. Further, the present assessment indicates that, if efficient time integration methods such as MTS were applied, the fluid part becomes the limiting factor for simulating reacting flows, because of the time-consuming calculation of the transport properties. The present problems show that the fluid part turns out to be more time-consuming than the chemical reaction part with more than 50 chemical species on the time step size of 1.e-8 s. As a result, the present method with MTS is 2∼5 times faster than the method with VODE for the one-dimensional problem. A detailed estimation of computational time for the transport properties is provided. Finally, the present method is successfully applied to the two-dimensional end-gas auto ignition phenomena of n-butane with available and reasonable computational resources.