Flame acceleration and transition to detonation in an array of square obstacles

We study flame acceleration and DDT in a two-dimensional staggered array of square obstacles by solving the compressible multidimensional reactive Navier–Stokes equations. The energy release rate for a stoichiometric H₂-air mixture is modeled by a one-step Arrhenius kinetics. The space between obstacles is filled with a stoichiometric H₂-air mixture at 1 atm and 298 K. Initially, the flow is at rest, and a flame is ignited at the center of the array. Computations show effects of the obstacles as a series of events leading to DDT. During the initial flame acceleration, the speed of the flame depends on the direction of flame propagation since some directions are more obstructed than others. This affects the macroscopic shape of the expanding burned region, which forms concave boundaries in more obstructed directions. As the flame accelerates, shocks form ahead of the flame, reflect from obstacles, and interact with the flame. There are more shock–flame interactions in more obstructed directions, and this leads to a greater flame acceleration and stronger leading shocks. When the shocks become strong enough, their collisions with obstacles ignite the gas mixture, and detonations form. The simulation shows four independent DDT events within a 90-degree sector, all in more obstructed directions. Resulting detonations spread in all directions. Some parts of detonation fronts are quenched by diffractions around obstacles, but they are reignited by collisions of decoupled shocks, or overtaken by other detonations. Thus detonations continue to spread and quickly burn all the material between the obstacles.     

Experimental investigation of fast flame propagation in stratified hydrogen–air mixtures in semi-confined flat layers

This paper presents results of an experimental investigation on fast flame propagation and the deflagration-to-detonation transition (DDT) and following detonation propagation in a semi-confined flat layer filled with stratified hydrogen–air mixtures. The experiments were performed in a transparent, rectangular channel open from below. The combustion channel has a width of 0.3 m and a length of 2.5 m. The effective layer thickness in the channel was varied by using different linear hydrogen concentration gradients. The method to create quasi-linear hydrogen concentration gradients that differ in the range and slope is also presented. The ignited mixtures were accelerated quickly to sonic flame speed in the first obstructed part of the channel. The interaction of the fast flame propagation with different obstacle set-ups was studied in the second part of the channel. The experimental results show an initiation of DDT by one additional metal grid in the obstructed semi-confined flat layer. Detonation propagation and failed detonation propagation were observed in obstructed and unobstructed parts of the channel.     

Nonlinear distribution characteristics of flame regions from methane–air explosions in coal tunnels

High temperature flame fronts generated in methane–air explosions are one of the major hazards in underground coal mines. However, the distribution laws of the flame region in explosions of this type and the factors influencing such explosions have rarely been studied. In this work, the commercial software package AutoReaGas, a finite-volume computational code for fluid dynamics suitable for gas explosion and blast problems, was used to carry out numerical simulations of a series of methane–air explosion processes for various initial premixed methane–air regions and cross-sectional areas in full-scale coal tunnels. Based on the simulated results and related experiments, the mechanism of flame propagation beyond the initial premixed methane–air region and the main factors influencing the flame region were analyzed. The precursor shock wave and turbulence disturb the initial unburned methane–air mixture and the pure air in front of the flame. The pure air and unburned mixture subsequently move backward along the axial direction and mix partially. The enlargement of the region containing methane induces that the range of the methane–air flame greatly exceeds the initial premixed methane–air region. The flame speed beyond the initial region is nonzero but appreciably lower than that in the original premixed methane–air region. The length of the initial premixed methane–air region has substantial influence on the size of the flame region, with the latter increasing exponentially as the former increases. For realistic coal tunnels, the cross-sectional tunnel area is not an important influencing factor in the flame region. These conclusions provide a theoretical framework in which to analyze accident causes and effectively mitigate loss arising from the repetition of similar accidents.     

An experimental study of flame acceleration and deflagration to detonation transition in representative process piping

The paper summarizes the results of experimental tests and accompanying analyses to investigate the factors that govern flame acceleration and potential transition to detonation in a relatively long unobstructed piping system. The overall aim of the work was to obtain sufficient experimental data so as to be able to develop and evaluate methodologies for classifying and predicting potential detonation flame acceleration and deflagration to detonation transition (DDT) hazard in industrial process pipes and mixtures. The present results show that the flame acceleration process in an unobstructed pipe exhibit three distinct phases: an initial establishment phase; a second rapid acceleration phase and a final transition to detonation phase. Test results with ethylene indicate that the acceleration process is not sensitive to initial pressure (all other parameters remaining constant) but can be sensitivity to initial pipe wall temperature or possibly mixture humidity. The presence of bends increases the local rate of turbulent combustion, an effect attributed to the additional turbulence generated downstream of the bend. For straight pipes, detonation was only observed to develop for hydrogen–air and ethylene–air mixtures. Detonation was not observed with methane, propane or acetone as fuel in the present piping apparatus.     

多分支管道油气爆炸特性大涡模拟

为了研究油库常见的分支结构空间内发生油气爆炸时火焰和压力的传播特性,建立了基于WALE湍流模型及Zimont预混火焰模型的油气爆炸模型;模拟了6种不同分支管道结构空间内汽油/空气混合物爆炸发生发展过程;研究了分支管道数量及相对设置位置对爆炸超压的影响规律,以及分支管道对火焰传播形态和速度的影响规律;模拟结果与前人相关实验规律进行对比。研究结果表明:分支管道对汽油/空气混合气预混爆炸具有明显的强化激励作用;火焰锋面传播经过分支管道时,经历规则—褶皱—规则的变化过程;主管道内火焰传播速度,在分支管道对流场的突扩作用和湍流作用的共同影响下呈震荡变化的规律。     

Propagation Characteristics of Gas Explosion in Linked Vessels Based on DDT Criteria

To study the occurrence conditions and propagation characteristics of deflagration to detonation transition (DDT) in linked vessels, two typical linked vessels were investigated in this study. The DDT of the methane–air mixture under different pipe lengths and inner diameters was studied. Results showed that the CJ detonation pressure of the methane–air mixture was 1.86 MPa, and the CJ detonation velocity was 1987.4 m/s. Compared with a single pipe, the induced distance of DDT is relatively short in the linked vessels. With the increase in pipeline length, DDT is more likely to occur. Under the same pipe diameter, the DDT induction distance in the vessel–pipe–vessel structure is shorter than that in the vessel–pipe structure. With the increase in pipeline diameter, the length of the pipe required to form the DDT is reduced. For linked vessels in which detonation formed, four stages, namely, slow combustion, deflagration, deflagration to detonation, and stable detonation, occurred in the vessels. Moreover, for a pipe diameter of 60 mm and a length of 8 m, overdriven detonation occurred in the vessel–pipe–vessel structure.     

Numerical simulation of deflagration-to-detonation transition in large confined volumes

A methodology for the computationally efficient CFD simulation of hydrogen-air explosions (including transition to detonation) in large volumes is presented. The model is validated by means of the largest ever conducted indoor DDT experiments in the RUT facility. A combination of models is proposed with a particular focus on the influence of flame-instabilities, especially of thermal-diffusive nature, which are crucial for very lean mixtures. Excellent agreement is achieved in terms of flame acceleration. The quality of DDT predictions itself depends on the underlying mechanism. Whereas DDT by shock-focusing is successfully simulated on under-resolved meshes, DDT by local explosions in the vicinity of the turbulent flame brush remains a challenge. Adaptive mesh refinement therefore emerges as a key technique to resolve more of the essential phenomena at reasonable computational costs affordable by industry. Finally, a generic case demonstrates the influence of mixture inhomogeneity, which can promote flame acceleration and ultimately DDT.     

长径比对管道油气爆炸特性与火焰传播规律影响研究*

为了探究长径比对油气爆炸传播特性与火焰传播规律的影响，为复杂管道受限空间油气爆炸防控提供理论参考，结合油气爆炸与爆炸抑制工程实际需要，构建不同长径比管道油气爆炸模拟实验系统,在此基础上开展不同初始浓度的预混油气-空气混合气爆炸实验。研究结果表明:管道内部的预混油气爆炸超压信号呈先上升后下降的趋势，由于耗散以及憋压效应导致超压下降平稳后仍大于初始压力；同时长径比增加会导致达到最大爆炸超压的油气浓度增加，油气爆炸超压峰值随着长径比的增加呈现上升→下降→上升的规律，小长径比管道的油气爆炸超压峰值高于大长径比管道，但同为小长径比管道或大长径比管道工况的实验结果对比显示爆炸超压峰值随着长径比增加而提升；而超压上升速率则会随着长径比的增加而上升；长径比的增加同时也会促进火焰的加速传播并减小火焰持续时间。     

Numerical simulation of flame acceleration and deflagration-to-detonation transition of ethylene in channels

Ethylene (C₂H₄) is a hydrocarbon fuel and widely used in chemical industry, however, ethylene is highly flammable and therefore presents a serious fire and explosion hazard. This work is initiated by addressing the hazard assessment of ethylene mixtures in different scale channels (d = 5 mm, 10 mm and 20 mm) from the aspect of flame acceleration (FA) and deflagration-to-detonation transition (DDT) by using large eddy simulation (LES) method coupled with the artificially thickened flame (ATF) approach. The fifth order local characteristics based weighted essentially non-oscillatory (WENO) conservative finite difference scheme is employed to solve the governing equations. The numerical results confirm that flame velocity increase rapidly at the beginning stage in three channels, and the flame acceleration rate is slower in the subsequent stage, afterwards, the flame velocity has an abrupt increase, and the onset of detonation occurs. Due to the fact that wall effect is significant in the narrow channel (e.g.,5 mm), especially in the ignition stage of the flame, flames have different shapes in wider channels (10 mm and 20 mm) and narrow channel (5 mm). Both the pressure and temperature profiles confirm DDT run-up distances are 0.251 m, 0.203 m and 0.161 m in 20 mm, 10 mm and 5 mm channels, respectively, which indicates that a shorter run-up distance is required in narrower channel. The cellular detonation structures for the ethylene-air mixture in different channels indicate that multi-headed detonation structures can be found in 20 mm channel, as the channel width decreases to 10 mm, detonation has a single-headed spinning structure, as the width is further reduced to 5 mm, only large longitudinal oscillation of the pressure can be observed.     

Efficient simulation of flame acceleration and deflagration-to-detonation transition in smooth pipes

Evaluation of accident scenarios including flame acceleration and deflagration-to-detonation transition (DDT) in chemical plant piping systems increases the need for an efficient numerical simulation tool capable of dealing with this phenomenon. In this work, a hybrid pressure-density-based solver including deflagrative flame propagation as well as detonation propagation is presented. The initial incompressible acceleration stage is covered by the pressure-based solver until the flame velocity reaches the fast flame regime and transition to the density-based solver is done. The deflagration source term is formulated in terms of a turbulent flame speed closure model incorporating various physical effects crucial for flame acceleration at low turbulence conditions (Katzy and Sattelmayer, 2018). Modelling of the detonation source term is based on a quadratic heat release function (Hasslberger, 2017). The presented numerical approach is validated in terms of DDT locations and pressure data from Schildberg (2015) as well as recently completed flame tip position measurements. For this purpose, H₂/O₂/N₂ mixtures ranging from 25.6 vol-% H₂ to 29.56 vol-% H₂ in two different pipe geometries are considered. The focus of the current work is on predicting the DDT location correctly and good agreement is observed for the investigated cases.     

Effects of premixed methane concentration on distribution of flame region and hazard effects in a tube and a tunnel gas explosion

Study of flame distribution laws and the hazard effects in a tunnel gas explosion accident is of great importance for safety issue. However, it has not yet been fully explored. The object of present work is mainly to study the effects of premixed gas concentration on the distribution law of the flame region and the hazard effects involving methane-air explosion in a tube and a tunnel based on experimental and numerical results. The experiments were conducted in a tube with one end closed and the other open. The tube was partially filled with premixed methane-air mixture with six different premixed methane concentrations. Major simulation works were performed in a full-scale tunnel with a length of 1000 m. The first 56 m of the tunnel were occupied by methane–air mixture. Results show that the flame region is always longer than the original gas region in any case. Concentration has significant effects on the flame region distribution and the explosion behaviors. In the tube, peak overpressures and maximum rates of overpressure rise (dp/dt)_max for mixtures with lower and higher concentrations are great lower than that for mixtures close to stoichiometric concentration. Due to the gas diffusion effect, not the stoichiometric mixture but the mixture with a slightly higher concentration of 11% gets the highest peak overpressure and the shock wave speed along the tube. In the full-scale tunnel, for fuel lean and stoichiometric mixture, the maximum peak combustion rates is achieved before arriving at the boundary of the original methane accumulation region, while for fuel rich mixture, the maximum value appears beyond the region. It is also found that the flame region for the case of stoichiometric mixture is the shortest as 72 m since the higher explosion intensity shortens the gas diffusion time. The case for concentration of 13% can reach up to a longest value of 128 m for longer diffusion time and the abundant fuel. The “serious injury and death” zone caused by shock wave may reach up to 3–8 times of the length of the original methane occupied region, which is the widest damage region.     

Experimental investigation of hydrogen–air deflagrations and detonations in semi-confined flat layers

This paper presents results of an experimental investigation on the deflagration and deflagration-to-detonation transition (DDT) in an obstructed (blockage ratio BR = 50%), semi-confined flat layer filled with uniform hydrogen–air mixtures. The effect of mixture reactivity depending on flat layer thickness and its width is studied to evaluate the critical conditions for sonic flame propagation and the possibility for detonation onset. The experiments were performed in a transparent, rectangular channel with a length of 2.5 m. The flat layer thickness was varied from 0.06 to 0.24 m and the experiments were performed for different channel widths of 0.3, 0.6 and 0.9 m. The experimental results show flame velocity vs. hydrogen concentration for different thicknesses and widths of the semi-confined flat layer. Three different flame propagation regimes were observed: slow subsonic flame (M << 1), sonic deflagration (M ～ 1) and detonation (M >> 1). It is shown that flame acceleration (FA) to sonic speed is independent of the width of the flat layer. The critical expansion ratio for effective flame acceleration to sonic speed was found to be linearly dependent on the reciprocal layer thickness.     

Combustion modeling in large scale volumes using EUROPLEXUS code

Most of the numerical benchmarks on combustion in large scale volumes for hydrogen safety, which were performed up until today have demonstrated, that current numerical codes and physical models experience poor predictive capabilities at the industrial scale, both due to under-resolution and deficiencies in combustion modeling. This paper describes a validation of the EUROPLEXUS code against several large scale experimental data sets in order to improve its hydrogen combustion modeling capabilities in industrial settings (e.g. reactor buildings). The code is based on the Euler equations and employs an algorithm for the propagation of reactive interfaces, RDEM, which includes a combustion wave, as an integrable part of the Reactive Riemann problem, propagating with a fundamental flame speed (being a function of initial mixture properties as well as gas dynamics parameters). Validation of the first combustion model implemented in the code is based on obstacle-laden channels, interconnected reactor-type compartments, vented enclosures and covers all major premixed flame combustion regimes (slow, fast and detonation) with an aim to obtain conservative results. An improvement of this model is found in a direction of transient interaction of flame fronts with regions of elevated integral length scales presented in the velocity gradient field due to e.g. interactions with geometrical non-uniformities and pressure waves.     

Scale effects on hydrogen–air fast deflagrations and detonations in small obstructed channels

An experimental study of flame propagation, acceleration and transition to detonation in hydrogen–air mixture in 2-m-long rectangular cross-section channel filled with obstacles located at the bottom wall was performed. The initial conditions of the hydrogen–air mixture were 0.1 MPa and 293 K and stoichiometric composition (29.6% H₂ in air). The channel width was 0.11 m and blockage ratio was 0.5 in all experiments. The effect of channel geometrical scale on flame propagation was studied by using four channel heights H of 0.01, 0.02, 0.04, and 0.08 m. In each case, the obstacle height was equal to H/2 and the obstacle spacing was 2H.

The propagation of flame and pressure waves was monitored by four pressure transducers and four ion probes. The pairs of transducers and probes were placed at various locations along the channel in order to get information about the progress of the phenomena along the channel.

As a result of the experiments, the deflagration and detonation regimes and velocities of flame propagation in the obstructed channel were established.     

大长径比管道汽油-空气混合气爆炸与抑制实验研究*

为探究狭长受限空间中油气爆炸失控时的发展状态,探索高效环保的油气爆炸抑制方法,利用长径比155的管道开展92号汽油-空气混合气爆炸发展规律和七氟丙烷主动抑爆技术研究。通过测量不同端部开口条件下油气爆炸超压、火焰传播速度、火焰强度等参数,对比研究空爆和抑爆工况下的油气爆炸变化规律,探讨长直管道中的油气爆炸特性,分析七氟丙烷抑爆效果。结果表明:大长径比管道中,端部开口泄爆对降低油气爆炸破坏能力的作用较小,开口与否对最大超压峰值的出现位置有影响;长直管道空爆时,油气爆炸由爆燃发展成爆轰,管道尾部的爆轰波速可达近2 000 m/s;密闭管道中,爆轰发生前火焰传播呈“已燃区-火焰锋面-待燃区-前驱激波-未燃区”的2波3区结构;主动抑爆方式下七氟丙烷抑爆效果良好,最大超压峰值降低幅度可达90%,火焰传播被及时阻断。     

Numerical analysis of hydrogen deflagration mitigation by venting through a duct

This paper presents a model and simulation results for the mitigation of a hydrogen–air deflagration by venting through a duct. A large eddy simulation (LES) model, applied previously to study both closed-vessel, and open atmosphere hydrogen–air deflagrations, was developed further to model a hydrogen–air explosion vented through a duct. Sub-grid scale (SGS) flame wrinkling factors were introduced to model major phenomena which contribute to the increase of flame surface area in vented deflagrations. Simulations were conducted to validate the model against 20% hydrogen–air mixture deflagrations (vent diameters 25 and 45 cm) and 10% hydrogen–air mixture deflagration (vent diameter 25 cm). There was reasonable correlation between the simulations and the experimental data. The comparative importance of different physical phenomena contributing to the flame wrinkling is discussed.     

Experimental study on DDT for hydrogen–methane–air mixtures in tube with obstacles

An experimental study of flame propagation, acceleration and transition to detonation in stoichiometric hydrogen–methane–air mixtures in 6 m long tube filled with obstacles located at different configurations was performed. The initial conditions of the hydrogen–methane–air mixtures were 1 atm and 293 K. Four different cases of obstacle blockage ratio (BR) 0.7, 0.6, 0.5 and 0.4 and three cases of obstacle spacing were used. The wave propagation was monitored by piezoelectric pressure transducers PCB. Pressure transducers were located at different positions along the channel to collect data concerning DDT and detonation development. Tested mixtures were ignited by a weak electric spark at one end of the tube. Detonation cell sizes were measured using smoked foil technique and analyzed with Matlab image processing toolbox. As a result of the experiments the deflagration and detonation regimes and velocities of flame propagation in the obstructed tube were determined.     

Explosions and deflagration-to-detonation transitions in epoxy propane/air mixtures

The explosion and deflagration-to-detonation transition (DDT) in epoxy propane (E.P.) vapor/air mixture clouds under weak ignition conditions has been studied in an experimental tube of diameter 199 mm and length 29.6 m. E.P. vapor clouds were formed by injecting liquid E.P. into the experimental tube and evaporating of the fine E.P. droplets. The dimension and the evaporating process of the E.P. droplet were measured and analyzed. The E.P. vapor/air mixture clouds were ignited by an electric spark with an ignition energy of 40 J. The characteristics and the stages of the DDT process in the E.P. vapor/air mixtures have been studied and analyzed. A self-sustained detonation wave formed, as was evident from the existence of a transverse wave and a cellular structure. Moreover, a retonation wave formed during the DDT process in the E.P. vapor/air mixture. The influence of the E.P. vapor concentration on the DDT process has been studied. The minimum E.P. vapor concentration for the occurrence of the DDT in the E.P. vapor/air mixture has been evaluated and the variation of DDT distance with E.P. vapor concentration has been analyzed.     

Structure and flame speed of dilute and dense layered coal-dust explosions

Multidimensional time-dependent simulations were performed to study the interaction of a stoichiometric methane–air detonation with layers of coal dust. The simulations solved equations representing a Eulerian kinetic-theory-based granular multiphase model applicable to dense and dilute particle volume fractions. These equations were solved using a high-order Godunov-based method for compressible fluid dynamics. Two dust layer concentrations were considered: loose with an initial volume fraction of 1%, and dense with an initial volume fraction of 47%. Each layer was simulated with two types of dust: reactive coal and inert ash. Burning of the coal particles results in a coupled complex consisting of an accelerating shock leading a coal-dust flame. The overall structure of the shock–flame complex resembles that of a premixed fast flame with length scales on the order of several meters. The large length scales are direct results of time needed to lift, mix, heat, and autoignite the particle. The flame speeds are large and much larger than the gas-phase velocity. Large spikes of flame speed are characteristic of the 47% case. These spikes and high flame speed are caused by pockets of coal dust autoigniting ahead of the flame. The flame is choked in the 1% case due to the gas-phase products exceeding the sonic velocity with respect to the flame. The 47% case is choked due to attenuation of pressure waves as they propagate through particles. Inert layers of dust substantially reduce the overpressure, impulse, and speed produced by propagating blast wave. The results also show that loose layers of dust are far more dangerous than dense layers. The shock and flame are more strongly coupled for loose layers, propagate at higher velocity, and produce large overpressures and impulses.