This short review article addresses the concept of scale modeling, philosophy, methods, and some specific examples. The article is written based on the authors' philosophy rather than collection of published articles on scale modeling which covers a very wide field of applications. The article is written not only for scale modeling researchers, but also for students, industry engineers and academic researchers who are interested in scale modeling.
One of the dominant scaling parameter in the fire safety research, Q*, represents a Froude number of natural convection induced by a fire. Q* has been used to model behavior of various fire-related phenomena including flame height of a turbulent diffusion flame, excess temperature and velocity of a fire plume and a ceiling jet, and even upper layer temperature of a fire room. In this article, examples of Froude modeling in reduced-scale experiments are overviewed. These include an investigation of the King's Cross Station Fire in 1987, a smoke flow experiment in the Himeji-jo castle, and an investigation of fire behavior of 96th floor of World Trade Center Tower 1 collapse in 2001.
From the perspective of energy security and environmental sustainability, IGCC (Integrated coal Gasification Combined Cycle) systems are expected as a coal-based power plant of the next generation. Gasification is a key technology for IGCC system. The entrained flow coal gasifier is operated at high temperature and elevated pressure, and very complicated phenomena occur in the gasifier. Clarification and the modeling of the individual phenomenon with the fundamental experiment equipment, the construction of the highly reliable numerical analysis technology based on the experimental results are important to the evaluation of the design and the examination of operating conditions of the large scale coal gasifier. In clarification and the modeling of the individual phenomenon with the fundamental experiment equipment, an important factor (e.g., including gasification reaction temperature and pressure, the properties of the molten slag, dimensions of equipment) is extracted precisely to simulate an actual coal gasifier. In consideration of them, it is necessary to establish the specifications of the fundamental experiment equipment, an experiment condition. Furthermore, the characteristic prediction and evaluation of the large scale coal gasifier which the influence of the scale-up considered is enabled by utilizing the numerical analysis technology that reflected results obtained in fundamental experiment equipment.
The residence time distribution (RTD) of coal particles in a gasifier of fluidized bed impacts its energy conversion efficiency. The RTD of coal in continuous bubbling fluidized bed is calculated by extended convection-diffusion model, in which segregation of coal is accounted for. The effect of bed length on the RTD of coal is investigated by this model, and the results are verified by experiments. When the bed length is increased, the dimensionless diffusion coefficient become smaller, and the flow patterns are also changed. Therefore, the bed length effects RTD and energy conversion efficiency of coal.
Combustion phenomena and model experiment in tuyere and raceway region in blast furnace were reviewed. In blast furnace, coke charged from furnace top and pulverized coal injected from tuyeres are main energy source, they are burned by blast blown from tuyeres. Combustion behavior in blast furnace raceway have the characteristic of high temperature blast, high temperature coke combustion, loose coke packed bed combustion, competitive reaction of coke and coal, mainly CO gas generation, so it is much different from another purverized coal combustion process such as a boiler for coal-fired power plant. Model experiment on raceway generation and combustion has been studied since long before. From the past coke-main operation period to recent high pulverized coal ratio operation period, research on combustion in tuyere and raceway region were explained.
The accidents induced by the combustion of combustible gases are explained. A gas fire will happen if an ignition occurs just after the leakage of combustible gas. A gas explosion will happen if an ignition occurs after some amount of combustible mixture has been formed. The risk of accident can be estimated by considering the probability of occurrence and the consequence of the accident. In this article, mainly the phenomena concerning the consequences of accidental gas explosions are explained to well mitigate their risk. The main consequences of accidental gas explosions, such as pressure increase, blast wave, and heat damage, are determined by the behavior of flame propagation. Flame front instabilities, which strongly affect the behavior of flame propagation, are described. The studies on body force instability (Rayleigh-Taylor instability) and hydrodynamic instability are explained. The phenomena of self-similar propagation were examined in largescale experiments, in which hydrodynamic instability becomes effective.
The unsteady 3-D numerical simulation was carried out in order to clarify the relative locations of flame and vortex filament soliton in the high-speed flame propagation phenomenon (the vortex bursting) along a line vortex. The position of vortex filament soliton was determined by a peak point of curvature distribution along the vortex line. The position of flame was determined by 3 kinds of physical properties, respectively; a gradient region of temperature distribution, a peak point of temperature gradient distribution, and a peak point of heat release rate distribution along the vortex line. In the 1-D analysis of flame propagation speed and in the 3-D visualization of helical vortex lines around the flame top, it was confirmed that the vortex bursting occurred in this numerical simulation. Furthermore, in the 1-D analyses of vortex line curvature, temperature, temperature gradient, and heat release rate along the vortex line, the peaks of vortex line curvature (the vortex filament solitons) were observed at the same locations as the peaks of temperature gradient and heat release rate (the flames), and the flames and the vortex filament solitons propagated in pairs along the vortex line. Due to these results, an evidence to support the vortex driving mechanism (the interaction between flame and vortex filament soliton) in the vortex bursting was obtained.
Gas explosion is a high-speed flame propagation phenomenon. The apparent flame speed during an explosion accelerates because of flame instability. Since the damage caused by an accidental explosion is significantly influenced by the flame speed, one must accurately evaluate such flame acceleration when assessing the risk level of an explosion hazard. An important risk-assessment tool of an accidental explosion will be direct numerical simulation (DNS) using a global reaction model. Then, the global reaction model must correctly reproduce flame instability behaviors. This study focuses on the Markstein number of H2/air mixture, an important parameter that describes the instability strength, and theoretically investigates if a one-step global reaction model can reproduce the correct Markstein number that was computed by a detailedchemistry simulation. It is first found that transport properties that depend on equivalence ratio must be considered when evaluating the Markstein number of a H2/air mixture. It is also found that rate parameters (such as the global activation energy) that depend on equivalence ratio must be used to correctly reproduce the Markstein number using a global reaction model. It is recommended to determine the global activation energy by fitting the Markstein number of the global reaction model with that computed by a detailed-chemistry simulation.
We studied the knock in the crevice of spark ignition engines experimentally by using a constant volume vessel. The vessel has an additional narrow channel which is simulated top land crevice. We find experimental fact that typical "crevice knock" is generated by the self-ignition of mixture in bottom of the channel when flame enters into the channel. We also find the amplitudes of measured of pressure waves in the crevice are stronger than these in the vessel. These experimental results are confirmed by mathematical knock model which is one-dimensional unsteady compressible flow with integration of Livengood-Wu.
We have developed a burner with high efficiency and low emission rate for the gas turbine combustor. The reductions of CO and NOx emissions are important because they influence the performance of the burner. We named this burner the “co-axial jet cluster burner” and, as the name indicates, it has multiple fuel nozzles and holes in a coaxial arrangement. To form lean premixed combustion, this burner mixes fuel and air in the multiple holes rapidly. For development of gas turbine burners, computational fluid dynamics (CFD) is a powerful tool to investigate the detailed distributions of various emissions and temperature. In recent years, the large eddy simulation (LES) model has been used for combustion analysis to analyze an unsteady combustion state in the combustor. Therefore, we have developed the hybrid turbulent combustion (HTC) model to calculate the form in which the non-premixed flame coexists with the premixed flame. Turbulent flow has been simulated using a LES with a dynamic sub-grid-scale (SGS) model. These two models were programmed to our simulation tool. However, there were unclear points about their applicability to an actual machine evaluation, especially prediction of CO concentration. In this study, we verified distributions of CO concentration by using the HTC model. In addition, we analyzed the CO generation mechanism for the lean premixed combustion in the burner. We found that CO is generated in the premixed flame zone and the flame quenching zone.