Prediction of gas explosion pressures: A machine learning algorithm based on KPCA and an optimized LSSVM

A gas explosion, as a common accident in public life and industry, poses a great threat to the safety of life and property. The determination and prediction of gas explosion pressures are greatly important for safety issues and emergency rescue after an accident occurs. Compared with traditional empirical and numerical models, machine learning models are definitely a superior approach. However, the application of machine learning in gas explosion pressure prediction has not reached its full potential. In this study, a hybrid gas explosion pressure prediction model based on kernel principal component analysis (KPCA), a least square support vector machine (LSSVM), and a gray wolf optimization (GWO) algorithm is proposed. A dataset consisting of 12 influencing factors of gas explosion pressures and 317 groups of data is constructed for developing and evaluating the KPCA-GWO-LSSVM model. The results show that the correlations among the 12 influencing factors are eliminated and dimensioned down by the KPCA method, and 5 composite indicators are obtained. The proposed KPCA-GWO-LSSVM hybrid model performs well in predicting gas explosion pressures, with coefficient of determination (R²), root mean square error (RMSE), and mean absolute error (MAE) values of 0.928, 26.234, and 12.494, respectively, for the training set; and 0.826, 25.951, and 13.964, respectively, for the test set. The proposed model outperforms the LSSVM, GWO-LSSVM, KPCA-LSSVM, beetle antennae search improved BP neural network (BAS-BPNN) models and reported empirical models. In addition, the sensitivity of influencing factors to the model is evaluated based on the constructed database, and the geometric parameters X1 and X2 of the confined structure are the most critical variables for gas explosion pressure prediction. The findings of this study can help expand the application of machine learning in gas explosion prediction and can truly benefit the treatment of gas explosion accidents.     

Effect of pyrolysis and oxidation characteristics on lauric acid and stearic acid dust explosion hazards

The effect of pyrolysis and oxidation characteristics on the explosion sensitivity and severity parameters, including the minimum ignition energy MIE, minimum ignition temperature MIT, minimum explosion concentration MEC, maximum explosion pressure P_max, maximum rate of pressure rise (dP/dt)_max and deflagration index K_st, of lauric acid and stearic acid dust clouds was experimentally investigated. A synchronous thermal analyser was used to test the particle thermal characteristics. The functional test apparatuses including the 1.2 L Hartmann-tube apparatus, modified Godbert-Greenwald furnace, and 20 L explosion apparatus were used to test the explosion parameters. The results indicated that the rapid and slow weight loss processes of lauric acid dust followed a one-dimensional diffusion model (D1 model) and a 1.5 order chemical reaction model (F1.5 model), respectively. In addition, the rapid and slow weight loss processes of stearic acid followed a 1.5 order chemical reaction model (F1.5 model) and a three-dimensional diffusion model (D3 model), respectively, and the corresponding average apparent activation energy E and pre-exponential factor A were larger than those of lauric acid. The stearic acid dust explosion had higher values of MIE and MIT, which were mainly dependent on the higher pyrolysis and oxidation temperatures and the larger apparent activation energy E determining the slower rate of chemical bond breakage during pyrolysis and oxidation. In contrast, the lauric acid dust explosion had a higher MEC related to a smaller pre-exponential factor A with a lower amount of released reaction heat and a lower heat release rate during pyrolysis and oxidation. Additionally, due to the competition regime of the higher oxidation reaction heat release and greater consumption of oxygen during explosion, the explosion pressure P_m of the stearic acid dust was larger in low concentration ranges and decayed to an even smaller pressure than with lauric acid when the concentration exceeded 500 g/m³. The rate of explosion pressure rise (dP/dt)_m of the stearic acid dust was always larger in the experimental concentration range. The stearic acid dust explosion possessed a higher P_max, (dP/dt)_max and K_st mainly because of a larger pre-exponential factor A related to more active sites participating in the pyrolysis and oxidation reaction. Consequently, the active chemical reaction occurred more violently, and the temperature and overpressure rose faster, indicating a higher explosion hazard class for stearic acid dust.     

Using computational fluid dynamics (CFD) for blast wave predictions

Explosions will, in most cases, generate blast waves. While simple models (e.g., Multi Energy Method) are useful for simple explosion geometries, most practical explosions are far from trivial and require detailed analyses. For a reliable estimate of the blast from a gas explosion it is necessary to know the explosion strength. The source explosion may not be symmetric; the pressure waves will be reflected or deflected when hitting objects, or even worse, the blast waves may propagate inside buildings or tunnels with a very low rate of decay. The use of computational fluid dynamics (CFD) explosion models for near and far field blast wave predictions has many advantages. These include more precise estimates of the energy and resulting pressure of the blast wave, as well as the ability to evaluate non-symmetrical effects caused by realistic geometries, gas cloud variations and ignition locations. This is essential when evaluating the likelihood of a given leak source as cause of an explosion or equally when evaluating the potential risk associated with a given leak source for a consequence analysis.In addition, unlike simple methods, CFD explosion models can also evaluate detailed dynamic effects in the near and far field, which include time dependent pressure loads as well as reflection and focusing of the blast waves. This is particularly valuable when assessing actual near-field blast damage during an explosion investigation or potential near-field damage during a risk analysis for a facility. One main challenge in applying CFD, however, is that these models require more information about the actual facility, including geometry details and process information. Collecting the necessary geometry and process data may be quite time consuming. This paper will show some blast prediction validation examples for the CFD model FLACS. It will also provide examples of how directional effects or interaction with objects can significantly influence the dynamics of the blast wave. Finally, the challenge of obtaining useful predictions with insufficient details regarding the geometry will also be addressed.     

Experimental study on unconfined methane explosion: Explosion characteristics and overpressure prediction method

Explosion accidents have become the main threat for the high-efficiency use of cleaner gas energy sources, such as natural gas. During an explosion, obstacle causing flame acceleration is the main reason for the increase of the explosion overpressure, which still remains to be fully understood. In this research, field experiments were conducted in a 1 m³ cubic frame apparatus to investigate the effect of built-in obstacles on unconfined methane explosion. Cage-like obstacles were constructed using square steel rods with different cross section size. The results demonstrated that the flame could get accelerated due to the hydrodynamic instability and obstacle-induced turbulence, which enhanced the explosion overpressure. In the near field, the overpressure wave travelled slower and the maximum overpressure could almost keep constant. Reducing the cross section size, or increasing the obstacle height or the obstacle number per layer could determine the rise of the maximum overpressure, the maximum pressure rising rate and the overpressure impulse. For uniformly constructed obstacles, self-similar theory was chosen to measure the influence of the hydrodynamic instability, and a parameter β was adopted to measure the flame acceleration caused by obstacle-induced turbulence, the value of which was 2 in this research. Based on the acoustic theory, an overpressure prediction model was proposed and the predicted results agreed with the measured values better than previous models, such as TNT equivalency model and TNO multi-energy model.     

Effects of vacuum levels,carbon dioxide,and structural sizes of linked vessels on dump vessel vented

The method of explosion venting is widely used in industrial explosion-proof design due to its simple operation, economical and practical features. A dump vessel vented platform was built. By changing the vacuum level and the gas in the dump vessels and the structural size of linked vessels, the pressure in the explosion vessel and the dump vessel was compared, and the influencing factors of explosion venting investigated. The main conclusions are as follows: In the explosion venting process, the higher the vacuum in the dump vessel, the smaller the pressure peak of the explosion vessel and the dump vessel, and the faster the explosion pressure is lowered. When the dump vessel is under the same vacuum level and the gas in the dump vessel is CO₂, the maximum pressure of the explosion vessel and the dump vessel is less than the maximum pressure when the containment medium is air. Under the same vacuum condition, the larger the volume ratio of the dump vessel and the explosion vessel, the smaller the pressure peak of the explosion vessel, the faster the explosion pressure drops, and the volume of the dump vessel reaches or exceeds the explosion vessel. Increasing the volume ratio of the containment vessel to the explosion vessel facilitates protection of the explosion vessel and the containment vessel. Under the same vacuum condition, when the gas explosion in 113 L vessel vents into 22 L vessel, the longer the length of the pipe, the greater the maximum pressure in the spherical vessel. When the gas explosion in 22 L vessel vents into 113 L dump vessel, as the pipeline grows, the maximum pressure in the two vessels decreases, but the reduction is not significant. In practical application, it is recommended to use a vacuum of 0.08Mpa or more for the dump vessel vented, and the containment medium is CO_2.In terms of the structural size of the container, it is recommended that the ratio of the receiving container to the explosion container be as large as possible, and the pipe length be as long.     

Validation of a new formula for predicting the lower flammability limit of hybrid mixtures

A correlation of the lower flammability limit for hybrid mixtures was recently proposed by us. The experimental conditions including ignition energy and turbulence which play a primary role in a gas or dust explosion were at fixed values. The sensitivity of such experimental conditions to the accuracy of the proposed formula was not thoroughly discussed in the previous work. Therefore, this work studied the effect of varying the ignition energy and turbulence intensity to the formula proposed in our previous paper. For ignition energy effect, results from methane/niacin mixture demonstrated that the MEC and LFL will not be affected by changing ignition energy. There is no distinguishable difference among gas explosion index (K_G) and dust explosion index (K_St) derived from tests with every ignition energy (2.5 kJ, 5 kJ and 10 kJ) in a 36 L vessel. The proposed formula is independent of ignition energy. For turbulence effect, the proposed formula can have a good prediction of the explosion and non-explosion zone if the ignition delay time is within a certain range. The formula prediction is good as the ignition delay time increases up to 100 ms in this work. Propane/niacin and propane/cornstarch mixtures are also tested to validate the proposed formula. It has been confirmed that the proposed formula predicts the explosion and non-explosion zone boundary of such mixtures.     

Modelling of the effect of size on flocculent dust explosions

The effect of size on the severity of explosions involving flocculent materials has been simulated by means of a model previously developed for spherical particles and here extended to the cylindrical geometry of flock. The model consists of the identification of the regime (internal and external heating, pyrolysis/devolatilization reaction, and volatiles combustion) controlling the explosion by the evaluation of dimensionless numbers (Bi, Da, Th and Pc) and then of the estimation of the deflagration index as a function of flocculent size. The model has been validated by means of explosion data of polyamide 6.6 (nylon) at varying diameter and length. The comparison between model and experimental data show a fairly good agreement.     

Niacin,lycopodium and polyethylene powder explosibility in 20-L and 1-m3 test chambers

An experimental program has been undertaken to investigate the explosibility of selected organic dusts. The work is part of a larger research project aimed at examination of a category of combustible dusts known as marginally explosible. These are materials that appear to explode in laboratory-scale test chambers, but which may not produce appreciable overpressures and rates of pressure rise in intermediate-scale chambers. Recent work by other researchers has also demonstrated that for some materials, the reverse occurs – i.e., values of explosion parameters are higher in a 1-m³ chamber than one with a volume of 20 L. Uncertainties can therefore arise in the design of dust explosion risk reduction measures.The following materials were tested in the current work: niacin, lycopodium and polyethylene, all of which are well-known to be combustible and which cover a relatively wide range of explosion consequence severity. The concept of marginal explosibility was incorporated by testing both fine and coarse fractions of polyethylene. Experiments were conducted at Dalhousie University using the following equipment: (i) Siwek 20-L explosion chamber for determination of maximum explosion pressure (P_max), volume-normalized maximum rate of pressure rise (K_St), and minimum explosible concentration (MEC), (ii) MIKE 3 apparatus for determination of minimum ignition energy (MIE), and (iii) BAM oven for determination of minimum ignition temperature (MIT). Testing was also conducted at Fauske & Associates, LLC using a 1-m³ explosion chamber for determination of P_max, K_St and MEC. All equipment were calibrated against reference dusts, and relevant ASTM methodologies were followed in all tests.The explosion data followed known trends in accordance with relevant physical and chemical phenomena. For example, P_max and K_St values for the fine sample of polyethylene were higher than those for the coarse sample because of the decrease in particle size. MEC values for all samples were comparable in both the 20-L and 1-m³ chambers. P_max and K_St values compared favorably in the different size vessels except for the coarse polyethylene sample. In this case, K_St determined in a volume of 1 m³ was significantly higher than the value from 20-L testing. The fact that the 20-L K_St was low (23 bar m/s) does not indicate marginal explosibility of the coarse polyethylene. This sample is clearly explosible as evidenced by the measured values of MEC, MIE, MIT, and 1-m³ K_St (at both 550 and 600 ms ignition delay times).     

CFD simulation of pressure piling

The explosion of flammable mixtures in interconnected compartments is commonly defined as “pressure piling”. Peak pressures much higher than the predictable thermodynamic values are likely to be generated in this geometry, yielding the phenomenon of major interest in industrial safety. In this paper, a CFD model was implemented, aiming at understanding the major factors affecting pressure piling in two cylindrical interconnected vessels, by varying the volume ratio between the two interconnected vessels and the ignition position. A combustion model was specifically developed to follow the flame propagation in any combustion regimes as a function of the local conditions: laminar, flamelet and distributed reaction zone.The model was validated by comparison with experimental results. The agreement between the experiments and the simulations has allowed the interpretation of the pressure piling phenomenon and the understanding of the mechanisms involved. More precisely, the results have showed that the pressure peak intensity is mainly affected by the coupling between the pre-compression of the mixture in the secondary vessel and the violence of explosion in the same vessel as related to the venting time, the latest quantified by the turbulent Bradley number, Br_t i.e. by the reaction time to the venting time ratio.     

球接管容器气体爆炸尺寸效应的试验研究与数值分析

为了解尺寸对球形容器连接管道甲烷-空气混合物爆炸的影响规律,利用Fluent软件,采用κ-ε湍流模型、涡耗散模型(简称EDC模型)、壁面热耗散、热辐射模型及SIMPLE算法,建立了球形容器连接管道内甲烷-空气混合物爆炸的数值模型,对容器与管道内甲烷-空气预混气体爆炸的尺寸效应进行了数值模拟。结果表明:随管道内径增大,球形容器内最大爆炸压力逐渐增大,管道末端最大爆炸压力变化无明显规律;而随管道长度增加,球形容器内最大爆炸压力逐渐减小;改变管道内径,较大体积球形容器内最大爆炸压力均大于较小体积球形容器内最大爆炸压力,最大爆炸压力上升速率的规律则相反,容器体积对管道末端最大爆炸压力的影响无明显规律。     

Venting of deflagrations: hydrocarbon–air and hydrogen–air systems

A set of 34 experiments on vented hydrocarbon–air and hydrogen–air deflagrations in unobstructed enclosures of volume up to 4000 m³ was processed with use of the advanced lumped parameter approach. Reasonable compliance between calculated pressure–time curves and experimental pressure traces is demonstrated for different explosion conditions, including high, moderate, low and extremely low reduced overpressures in enclosures of different shape (L_max:L_min up to 6:1) with different type and position of the ignition source relative to the vent, for near-stoichiometric air mixtures of acetone, methane, natural gas and propane, as well as for lean and stoichiometric hydrogen–air mixtures. New data were obtained on flame stretch for vented deflagrations.The fundamental Le Chatelier–Brown principle analog for vented deflagrations has been considered in detail and its universality has been confirmed. The importance of this principle for explosion safety engineering has been emphasized and proved by examples.A correlation for prediction of the deflagration–outflow interaction number, χ/μ, on enclosure scale, Bradley number and vent release pressure is suggested for unobstructed enclosures and a wide range of explosion conditions. Fractal theory has been employed to verify the universality of the dependence revealed of the deflagration–outflow interaction number on enclosure scale.In spite of differences between the thermodynamic and kinetic parameters of hydrocarbon–air and hydrogen–air systems, they both obey the same general regularities for vented deflagrations, including the Le Chatelier–Brown principle analog and the correlation for deflagration–outflow interaction number.     

Effective method of predicting confined pressure behavior under isotropic turbulence

Explosion pressure prediction is indispensable to ensure process safety against accidental gas explosions. This work is aimed at establishing a theoretical method for predicting confined methane-air explosion pressure under isotropic turbulence. The results indicated that the pressure rise rate becomes significantly increased by the existence of isotropic turbulence, which effect on peak value of explosion pressure is negligible. Among various models of turbulent burning velocity, the calculated pressure rise rate using Chiu model, Williams model and Liu model is relatively closer to experimental value. With the increase of turbulent integral length and RMS turbulent fluctuation velocity, the pressure rise rate becomes increased continuously. The influence of adiabatic compression and isothermal compression on pressure rise rate could be ignored. To predict explosion pressure in a more accurate way, the dynamic variation of turbulent integral length and RMS turbulent fluctuation velocity should be considered in the future.     

Comparative assessment of the explosion characteristics of alcohol–air mixtures

Explosion characteristics of five alcohol–air (ethanol, 1-butanol, 1-pentanol, 2-pentanol and 3-pentanol) mixtures were experimentally conducted in an isochoric chamber over wide ranges of initial temperature and pressure. The effect of temperature and pressure on the different explosion behaviors among these alcohols with various structures were investigated. Results show that the peak explosion pressure is increased with the decrease of temperature and increase of pressure. Maximum rate of pressure rise is insensitive to the temperature variation while it significantly increases with the increase of initial pressure. Among the 1-, 2-, and 3-pentanol–air mixtures, 1-pentanol has the highest values in peak explosion pressure and maximum rate of pressure rise and 2-pentanol gives the lowest values at the initial pressure of 0.1 MPa. These differences tend to be decreased with the increase of initial pressure. Among the three primary alcohol–air (ethanol, 1-butanol and 1-pentanol) mixtures, a similar explosion behavior is presented at the lean mixture side because of the combined effect of adiabtic temperature and flame propagation speed. At the rich mixture side, 1-pentanol gives the highest values in peak explosion pressure and maximum rate of pressure rise and ethanol gives the lowest values. This phenomenon can be interpretated from the combining influence of heat release and heat loss, since the flame speeds of ethanol-, 1-butanol-, 1-pentanol–air mixtures are close at rich mixture side.     

PAN dust explosion inhibition mechanisms of NaHCO3 and Al(OH)3

We investigate the PAN dust explosion inhibition behaviors of NaHCO₃ and Al(OH)₃ in a 20 L spherical explosion system and a transparent pipe explosion propagation test system. The results show that, in the standard 20 L spherical explosion system, the highest PAN dust explosion concentration is 500 g/m³, the maximum explosion pressure is 0.661 MPa, and the maximum explosion pressure increase rate is 31.64 MPa/s; adding 50% NaHCO₃ and 60% Al(OH)₃ can totally inhibit PAN dust explosion. In the DN0.15 m transparent pipe explosion propagation test system, for 500 g/m³ PAN dust, the initial explosion flame velocity is 102 m/s, the initial pressure is 0.46 MPa, and the initial temperature is 967 °C; adding 60% NaHCO₃ and 70% Al(OH)₃ can totally inhibit PAN dust explosion flames. Through FTIR and TG analyses, we obtain the explosion products and pyrolysis patterns of the explosion products of PAN dust, NaHCO₃, and Al(OH)₃. On this basis, we also summarize the PAN dust explosion inhibition mechanisms of NaHCO₃ and Al(OH)₃.     

Effects of a carbon monoxide-dominant gas mixture on the explosion and flame propagation behaviors of methane in air

This work aimed to experimentally evaluate the effects of a carbon monoxide-dominant gas mixture on the explosion characteristics of methane in air and report the results of an experimental study on explosion pressure measurement in closed vessel deflagration for a carbon monoxide-dominant gas mixture over its entire flammable range. Experiments were performed in a 20-L spherical explosion tank with a quartz glass window 110 mm in diameter using an electric spark (1 J) as the ignition source. All experiments were conducted at room temperature and at ambient pressure, with a relative humidity ranging from 52 to 73%. The peak explosion pressure (P_max), maximum pressure rise rate ((dp/dt)_max), and gas deflagration index (K_G) were observed and analyzed. The flame propagation behavior in the initial stage was recorded using a high-speed camera. The spherical outward flame front was determined on the basis of a canny method, from which the maximum flame propagation speed (S_n) was calculated. The results indicated that the existence of the mixture had a significant effect on the flame propagation of CH₄-air and increased its explosion risk. As the volume fraction of the mixed gas increases, the P_max, (dp/dt)_max, K_G and S_n of the fuel-lean CH₄-air mixture (7% CH₄-air mixture) increase nonlinearly. In contrast, addition of the mixed gas negatively affected the fuel-rich mixture (11% CH₄-air mixture), exhibiting a decreasing trend. Under stoichiometric conditions (9.5% CH₄-air mixture), the mixed gas slightly lowered P_max, (dp/dt)_max, K_G, and S_n. The P_max of CH₄-air mixtures at volume fractions of 7%, 9.5%, and 11% were 5.4, 6.9, and 6.8 bar, respectively. The S_n of CH₄-air mixtures at volume fractions of 7%, 9.5%, and 11% were 1.2 m/s, 2.0 m/s, and 1.8 m/s, respectively. The outcome of the study is comprehensive data that quantify the dependency of explosion severity parameters on the gas concentration. In the storage and transportation of flammable gases, the information is required to quantify the potential severity of an explosion, design vessels able to withstand an explosion and design explosion safety measures for installations handling this gas.     

应急状态下矿井瓦斯爆炸致灾因子传播快速分类器研究*

为了在矿井瓦斯爆炸灾变发生后，快速确定瓦斯爆炸冲击波的压力、温度、有毒有害气体等致灾因子在井巷网络中的传播情况。利用CFD数值模拟或爆炸实验获得瓦斯爆炸冲击波的压力、温度、有毒有害气体等致灾因子传播大数据，将影响瓦斯爆炸传播的因素以及观测点等参数作为人工神经网络的输入节点，压力、温度等致灾因子作为输出节点，建立瓦斯爆炸致灾因子传播快速预测机器学习模型，解决CFD数值模拟的建模、计算及数据分析处理等过程耗时大、不适应灾变应急的快速响应等问题。研究结果表明:在给定爆炸位置和爆炸当量的均直巷道，获得任一点的爆炸冲击波压力、温度以及有毒有害气体所需时间是瞬时的，人工神经网络平均训练误差为6.92 %，有训练样本的验证误差为5.24 %，无训练样本的验证误差为6.88 %。     

Inhibiting effects of gas–particle mixtures containing CO2, Mg(OH)2 particles,and NH4H2PO4 particles on methane explosion in a 20-L closed vessel

The main risk factors from methane explosion are the associated shock waves, flames, and harmful gases. Inert gases and inhibiting powders are commonly used to prevent and mitigate the damage caused by an explosion. In this study, three inhibitors (inert gas with 8.0 vol% CO₂, 0.25 g/L Mg(OH)₂ particles, and 0.25 g/L NH₄H₂PO₄ particles) were prepared. Their inhibiting effects on methane explosions with various concentrations of methane were tested in a nearly spherical 20-L explosion vessel. Both single-component inhibitors and gas–particle mixtures can substantially suppress methane explosions with varying degrees of success. However, various inhibitors exhibited distinct reaction mechanisms for methane gas, which indicated that their inhibiting effects for methane explosion varied. To alleviate amplitude, the ranking of single-component inhibitors for both explosion pressure (P_ex) and the rate of explosion pressure rise [(dP/dt)_ex] was as follows: CO₂, NH₄H₂PO₄ particles, and Mg(OH)₂ particles. In order of decreasing amplitude, the ranking of gas‒particle mixtures for both P_ex and (dP/dt)_ex was as follows: CO₂–NH₄H₂PO₄ mixture, CO₂‒Mg(OH)₂ mixture, and pure CO₂. Overall, the optimal suppression effect was observed in the system with the CO₂–NH₄H₂PO₄ mixture, which exhibited an eminent synergistic effect on methane explosions. The amplitudes of P_ex with methane concentrations of 7.0, 9.5, and 11.0 vol% decreased by 37.1%, 42.5%, and 98.6%, respectively, when using the CO₂–NH₄H₂PO₄ mixture. In addition, an antagonistic effect was observed with CO₂‒Mg(OH)₂ mixtures because MgO, which was generated by the thermal decomposition of Mg(OH)₂, can chemically react with water vapor and CO₂ to produce basic magnesium carbonate (xMgCO₃·yMg(OH)₂·zH₂O), thereby reducing the CO₂ concentration in a reaction system. This research revealed the inhibiting effects of gas‒particle mixtures (including CO₂, Mg(OH)₂ particles, and NH₄H₂PO₄ particles) on methane explosions and provided primary experimental data.     

Investigations of flame front propagation between interconnected process vessels. Development of a new flame front propagation time prediction model

For the case where a dust or gas explosion can occur in a connected process vessel, it would be useful, for the purpose of designing protection measures and also for assessing the existing protection measures such as the correct placement, to have a tool to estimate the time for flame front propagation along the connecting pipe. Measurements of data from large-scale explosion tests in industrially relevant process vessels are reported. To determine the flame front propagation time, either a 1 m³ or a 4.25 m³ primary process vessel was connected via a pipe to a mechanically or pneumatically fed 9.4 m³ secondary silo. The explosion propagation started after ignition of a maize starch/air mixture in the primary vessel. No additional dust was present along the connecting pipe. Systematic investigations of the explosion data have shown a relationship between the flame front propagating time and the reduced explosion over-pressure of the primary explosion vessel for both vessel volumes. Furthermore, it was possible to validate this theory by using explosion data from previous investigations. Using the data, a flame front propagation time prediction model was developed which is applicable for:

• •gas and dust explosions up to a K value of 100 and 200 bar m s⁻¹, respectively, and a maximum reduced explosion over-pressure of up to 7 bar;
• •explosion vessel volumes of 0.5, 1, 4.25 and 9.4 m³, independent of whether they are closed or vented;
• •connecting pipes of pneumatic systems with diameters of 100–200 mm and an air velocity up to 30 m s⁻¹;
• •open ended pipes and pipes of interconnected vessels with a diameter equal to or greater than 100 mm;
• •lengths of connecting pipe of at least 2.5–7 m.

     

Experimental study on the minimum explosion concentration of anthracite dust: The roles of O2 mole fraction,inert gas and CH4 addition

The explosion characteristics of anthracite coal dust with/without small amount of CH₄ (1.14 vol %) were investigated by using a 20 L spherical explosion apparatus with an emphasis on the roles of oxygen mole fraction and inert gas. Two methods based on overpressure and combustion duration time were used to determine the minimum explosion concentration (MEC) or the lower explosion limit (LEL) of the pure anthracite coal dust and the hybrid coal-methane mixtures, respectively. The experiment results showed that increasing oxygen mole fraction increases the explosion risk of coal dust: with increasing oxygen mole fraction, the explosion pressure (P_ex) and the rate of explosion pressure rise ((dp/dt)_ex)) increase, while MEC decreases. The explosion risk of anthracite dust was found to be lower after replacing N₂ with CO₂, suggesting that CO₂ has a better inhibition effect on explosion mainly due to its higher specific heat. However, the addition of 1.14% CH₄ moderates the inhibition effect of CO₂ and the promotion effect of O₂ on anthracite dust explosion for some extent, increasing explosion severity and reducing the MEC of anthracite dust. For hybrid anthracite/CH₄ mixture explosions, Barknecht's curve was found to be more accurate and conservative than Chatelier's line, but neither are sufficient from the safety considerations. The experimental results provide a certain help for the explosion prevention and suppression in carbonaceous dust industries.     

基于FA-BAS-ELM的海洋油气管道外腐蚀速率预测

为提高海洋油气管道外腐蚀速率预测的精度和效率,建立基于因子分析(FA)和天牛须搜索算法(BAS)的极限学习机(ELM)腐蚀速率预测模型。利用FA对影响因素数据集进行降维处理,确定预测模型的输入变量;建立ELM预测模型,并采用BAS对ELM模型的参数进行优化,避免参数取值随机性对模型预测性能的影响;以实海挂片试验为例,通过建模仿真评价模型的预测性能,并与其他模型进行对比分析。结果表明:FA-BAS-ELM预测模型的平均绝对误差(MAPE)仅为1.92%,决定系数R²高达0.9949,相比于其他模型,该模型具有更优的预测性能。