Home Discrete probability model-based method for recognition of multicomponent combustible gas explosion hazard sources
Article Open Access

Discrete probability model-based method for recognition of multicomponent combustible gas explosion hazard sources

  • Wenhua Ye EMAIL logo , Yang Feng and Yangming Chen
Published/Copyright: March 22, 2022
Download Article (PDF)
Become an author with De Gruyter Brill
Submit Manuscript Author Information
High Temperature Materials and Processes
From the journal High Temperature Materials and Processes Volume 41 Issue 1

Abstract

Aiming at the problems of poor sensitivity and low accuracy of traditional combustible gas explosion hazard identification methods, the discrete probability model was introduced to design a multicomponent combustible gas explosion hazard identification method. Fem3 model is selected to analyze the diffusion mode of multicomponent combustible gas, and the calculation formula of diffusion velocity is obtained. The explosion limit of multicomponent combustible gas was observed by means of test, and the upper limit of explosion concentration of multicomponent combustible gas was obtained. According to the characteristics of explosion, a multicomponent combustible gas explosion hazard identification device is designed. The discrete probability model is used to identify the explosion hazard of multicomponent combustible gas. Experimental results show that the proposed method greatly improves the recognition sensitivity and accuracy, which fully shows that the proposed method has a better detection effect.

Keywords: discrete probability model; multicomponent; combustible gas; explosion; hazard source

1 Introduction

Combustible gas is a kind of substance that can ignite and present gas state at room temperature and pressure, for example, hydrogen (H2), acetylene (C2H2), ethylene (C2H4), ammonia, hydrogen sulfide (H2S), and so on. Combustible gases have general characteristics of gases [1]. One component of combustible gas is called single gas; the mixture of two or more combustible gases is called mixed combustible gas, also known as multicomponent combustible gas. Combustible gases can cause combustion or explosion by mixing in a certain proportion in the corresponding combustion-supporting medium and under the action of an ignition source [2]. The combustible gas is ejected from the nozzle at a certain velocity, and its combustion speed depends on the diffusion speed of combustible gas and air. There are many combustible gases, including H2, carbon monoxide (CO), methane (CH4), ethane (C2H6), propane (C3H8), butane (C4H10), C2H4, propylene (C3H6), butylene, C2H2, methyl acetylene (C3H4), butyne, H2S, hydrogen phosphide, etc. Multicomponent combustible gases and air must be uniformly mixed in a certain concentration range to form premixed gases, which will explode when ignited [3]. This concentration range is called explosion limit or explosion concentration limit. In a very short time, explosion will release a large amount of energy, produce high temperature, and release a large amount of gas, causing high-pressure chemical reactions or state changes in the surrounding media while being extremely destructive [4]. Multicomponent flammable gas explosion has many forms of damage. First, it has a direct destructive effect. After explosion of mechanical equipment, devices, containers, and so on, many debris will be produced, which will cause harm in a considerable range after flying out, general fragments are scattered within 100–500 m. Second is the destructive effect of a shock wave. When explosion occurs, the high-temperature and high-pressure gas expands at a very high speed, squeezing the surrounding air like a piston, transferring part of the energy released from explosion reaction to the compressed air layer. The air is disturbed by impact, which results in pressure, density, and so on [5]. A sudden change occurs when the disturbance propagates in the air, which is called a shock wave. Shock wave propagation speed is very fast; in the process of transmission, it can destroy the mechanical equipment and buildings in the surrounding environment and even cause casualties. Third, it can cause fire. After the explosion, the diffusion of explosive gas products occurs only in a very short time, which is not enough to cause fire for general combustibles. Combustion and explosive wind caused by shock wave also have a fire-extinguishing effect. However, the high temperature and pressure produced by the explosion and the large amount of heat or residual fire in buildings will ignite the vapor of flammable gas, or flammable liquid that continuously flows out from the damaged equipment, or may ignite other flammable substances to cause fire. Finally, it will cause poisoning and environmental pollution, in actual production, many substances are produced. It is not only flammable but also poisonous. When explosion accidents occur, a large number of harmful substances will leak out, causing human poisoning and environmental pollution. Therefore, the explosion risk of multicomponent combustible gas is very high. To avoid the occurrence of multicomponent combustible gas explosion, it is necessary to identify multicomponent combustible gas explosion hazard sources.

In recent years, the identification methods of multicomponent combustible gas explosion hazard sources have been studied at home and abroad, and some achievements have been made. Tianhao and Zhijie proposed an underground gas explosion positioning and testing system based on field-programmable gate array (FPGA) [6], which used theories related to seismic positioning to locate the underground explosions. An acceleration testing system based on FPGA was designed, and the ADXL345 triaxial acceleration sensor was used to test the triaxial acceleration data of each test point to locate the explosion center. Combined with the distributed test system theory, the data were analyzed reasonably, and the difference between the obtained positioning results and the actual results was less than 3%, which verified the accuracy and practicability of the test system to some extent. Qi Xinge et al. Proposed a quantitative division method of gas protection area based on equivalent gas cloud explosion risk assessment [7], considering gas leakage probability, site features elements such as joint probability distribution of wind speed and direction, and using the gauss diffusion model, get the gas leakage diffusion equivalent volume of gas cloud and gas leakage diffusion risk collection, screening and leakage scenario. Ignition probability analysis is carried out for scenarios with large diffusion risk, and the influence range of gas cloud explosion is calculated by using multienergy method. Risk assessment of gas cloud explosion accident is carried out to obtain the risk set of explosion accident consequence. Under the guidance of as low as reasonable practice standard and the scene coverage rate of fire and gas system detector, the quantitative classification standard of gas protection area is determined according to the risk value of different device area. Through the case analysis of an liquefied natural gas (LNG)-receiving station, the protection zone grade of different devices can be quantitatively obtained, and the gas protection zone grade can be quantitatively divided according to the specific leakage scene. The numerical calculation shows that the quantitative division of gas protection area can provide theoretical support for the detector layout of fire gas system. Zhong et al. Proposed an underground chemical explosion gas leakage analysis method based on dimensional analysis and Darcy’s law [8], dimensional analysis are used to get the main physical quantities affecting gas leak time, including gas dynamic viscosity, indoor overpressure, the square of porosity and thickness of surrounding rock and the ratio of permeability, etc., and gives the function relation between them. Then, based on Darcy’s law, the analytical formula for calculating the gas leakage time is derived. The formula for calculating the gas leakage time is in good agreement with the qualitative function relation obtained by the dimensional analysis. The two formulas are consistent with the same problem from different angles. It can provide research ideas and tools for the theoretical analysis and regular study of gas leakage in underground chemical explosion and provide reference for engineering estimation related to underground explosion. The above three methods have their own advantages, but they all have the defects of poor recognition sensitivity and low accuracy, which cannot meet the needs of today’s society. Therefore, the discrete probability model is introduced to design the multicomponent combustible gas explosion hazard identification method. In this study, the EM3 model is used to analyze the diffusion law of multicomponent combustible gas, and the diffusion speed calculation formula is determined. The upper limit of gas explosion concentration is obtained by analyzing the constraint conditions of explosion limit of multicomponent combustible gas through test. The explosion hazard source of multicomponent combustible gas is identified through the discrete probability model, and the identification of explosion hazard source of multicomponent combustible gas is realized. The proposed method greatly improves the recognition sensitivity and accuracy, which fully shows that the proposed method has a better detection effect.

2 Basic definitions

2.1 Analysis of multicomponent combustible gas diffusion model

The explosion of multicomponent combustible gas is mainly determined by the concentration of multicomponent combustible gas. That is, combustible substances (combustible gases, vapors, and dust) and air (or oxygen) must be uniformly mixed in a certain concentration range to form a premixed gas, which will only explode when the source of fire occurs. The explosion limit of combustible mixture is divided into explosion (ignition) lower limit and explosion (ignition) upper limit called as explosion lower limit and explosion upper limit, respectively. The upper limit refers to the high concentration at which a flammable mixture can explode. When above the upper limit of explosion, the air is insufficient, resulting in the flame cannot spread and cannot explode but can burn. The lower limit refers to the low concentration at which a flammable mixture can explode. Because the concentration of combustible is not enough, the cooling effect of excess air prevents the flame from spreading, so it does not explode or catch fire when it is below the lower limit of explosion. Therefore, an appropriate model is selected to analyze the diffusion mode of multicomponent combustible gas [9]. The specific analysis process of multicomponent combustible gas diffusion model is as follows.

Nowadays, there are mainly six kinds of multicomponent flammable gas diffusion model analysis models, which are Gauss model, Britter and McQuaid (BM) model, Sutton model, 3-D Finite Element Model (FEM3), Box model, and P-G model. To select the most suitable analysis model, the characteristics and defects of the existing models are analyzed. The specific analysis contents are shown in Table 1.

Table 1

Characteristic comparison table of multicomponent combustible gas diffusion model analysis model

Model name Gaussian model BM model Sutton model FEM3 model Box model P-G model
Applicable object Neutral gas Neutral and heavy gases Neutral gas Heavy gas Neutral and heavy gases Neutral gas
Scope of application Large scale, short term Large scale, long term Large scale, long term Unrestricted Unrestricted Unrestricted
Facility value Easier Easier Easier More difficult Easier Easier
Calculating amount Less Less Less Many Less Less
Accuracy of calculation Poor Commonly Poor Preferably Poor Poor
Correlation coefficient Gas properties, atmospheric stability, temperature, wind speed, and wind direction Average concentration and initial concentration on cross section of gas Diffusion parameters related to meteorological conditions Temperature, wind speed, and wind direction Average radius, average height, and average temperature of gases Wind speed, atmospheric stability, topography, leakage source height, and gas properties
Characteristic Simulable continuous leakage and instantaneous leakage The calculation diagram consists of experimental data Using turbulent diffusion theory Treatment of continuous source leakage and limited time discharge Predictable overall characteristics of gas clouds, regardless of details More detailed consideration of environmental factors
Defect It is only suitable for neutral gases, and its simulation accuracy is poor As an empirical model, the extensibility is poor Large error exists in simulating the leakage and diffusion of flammable gases The calculation is very heavy and difficult There are limitations and uncertainties There is a large error in determining the atmospheric stability

As shown in Table 1, the FEM3 model is found to be more suitable for the multicomponent combustible gas diffusion model analysis. This model belongs to the category of three-dimensional finite element method. It originated in 1979. It has carried out corresponding research on sudden leakage of LNG and other aspects. After the matching simulation operation, it is concluded that it has a relatively ideal simulation effect for LNG [10].

The core application of this model is the leakage of air with density exceeding, such as chlorine, which can be more efficiently applied to the continuous leakage and the leakage phenomenon in a short time. In the process of solving this kind of problem, the original model needs to be revised. The introduction of the finite element method belongs to the Galerkin method based on further optimization and adjustment; the key is to obtain the relevant unsteady equation, relying on the model to solve the diffusion problem, we need to introduce a supporting K theory (also known as local equilibrium) to deal with. In recent years, this model can be used to calculate relatively special terrain diffusion problems, such as those related to diffusion in buildings. However, the shortcomings of this model lie in its complicated operation and relatively high overall difficulty [11]. The core formula of the model is as follows:

(1) ∂ ( ρ u ) ∂ t + ρ u ∇ u = ∇ ( ρ K m ∇ u ) + ( ρ − ρ h ) g , ∇ ( ρ U ) = 0 , ∂ T ∂ t + u ∇ T = 1 ρ C p ∇ ( ρ C p K T ∇ T ) + C p A − C p N C p ( K T ∇ T ) ∇ T , ∂ w ∂ t + u ∇ w = 1 ρ ( ρ K w ⋅ ∇ w ) , ρ = P M R T = P R T w M N + M A .

Among them, ρ denotes the density of multicomponent combustible gases in kilogram per cubic meter, u denotes the friction velocity, t denotes the time parameter, K m denotes the temperature diffusion coefficient of multicomponent combustible gases, h denotes the height of gas cloud, g denotes the acceleration of gravity, U denotes the velocity of gas, and T denotes the temperature diffusion coefficient of multicomponent combustible gases. It is temperature, C p denotes the specific heat capacity of mixing index, K T denotes the diffusion coefficient of gas velocity, C p A denotes the specific heat capacity of controlling index, C p N denotes the specific heat capacity of diffusing index, w denotes the convection parameter, K w denotes the diffusion coefficient of concentration, P denotes the diffusion pressure, M denotes the specific heat capacity of controlling index molecular mass of mixed gases, R denotes the gas constant, M N denotes the molecular mass of gases, and M A denotes the molecular mass of air.

Assuming K m = K T , the vertical and horizontal constants are obtained as follows:

(2) K v = k [ ( u ⋅ z ) 2 + ( w ⋅ h ) 2 ] 1 / 2 Φ , K h = β k u Φ z Φ .

Among them, k denotes von Kaman constant, z denotes height, Φ denotes Monin–Obukhov function, and β denotes empirical constant.

The diffusion model of multicomponent combustible gas is analyzed by the above model.

(3) v = ∏ K v ∗ K h α ⋅ ∑ ρ ⋅ R ⋅ T .

Among them, α represents the parameters for calculating the diffusion velocity of multicomponent combustible gases.

Through the selected FEM3 model, the diffusion model of a multicomponent combustible gas is analyzed, and the calculation formula of diffusion velocity is obtained, which will prepare for the following observation of the explosion limit of a multicomponent combustible gas [12].

2.2 Observation of explosion limits for multicomponent flammable gases

To improve the accuracy of a multicomponent combustible gas explosion limit observation, it is necessary to calculate the minimum oxygen concentration of combustible gas. The minimum oxygen concentration (also known as the maximum allowable oxygen content by some researchers) refers to the minimum oxygen concentration required for combustible gas to maintain combustion or explosion at a certain concentration. It exists in the whole explosion range. Near the lower limit of combustible gas, the oxygen concentration is relatively more, and the combustible gas concentration is relatively less. At this time, if it is required to reach the minimum oxygen concentration, the amount of inert gas added should be relatively more. With the increase of the limit value of combustible gas, the oxygen concentration in the mixed gas decreases. At this time, if it is required to reach the minimum oxygen concentration, less inert gas will be added, and less inert gas will be added at the upper explosion limit.

In the actual industrial production, to improve the production efficiency and the conversion rate of raw materials, it is necessary to have a fixed concentration of combustible gas as the raw material to participate in the reaction. If the concentration of combustible gas is within the explosion range, the explosion accident can be avoided only by reducing the oxygen concentration to just below the minimum oxygen concentration of the reaction. It can also maximize production efficiency.

The minimum oxygen concentration of combustible gas can be calculated theoretically. In case of complete combustion of combustible gas and oxygen, the concentration of combustible gas components in complete reaction can be calculated from the following chemical reaction equation:

(4) C n H m O λ F f + n + m − f − 2 λ 4 O 2 → n CO 2 + m − f 2 H 2 O + f HF ,

where n , m , λ , f represent the atomic number of carbon, H2, oxygen, and halogen elements, respectively.

Theoretically, the concentration of combustible gas completely burned with 1 M of air is as follows:

(5) C st = 100 1 + 4.733 n + m − f − 2 λ 4 ( V % ) ,

where 4.773 is the reciprocal of 0.2095 M of oxygen in the air. The concentration C st of combustible gas completely combusted with 1 M of oxygen can be simplified as follows:

(6) C st = 100 1 + n + m − f − 2 λ 4 ( V % ) .

The theoretical calculation formula of the minimum oxygen concentration of combustible gas is as follows:

(7) φ ( O 2 ) = L n + m − f − 2 λ 4 ,

where L is the concentration of combustible gas.

Of course, the theoretical minimum oxygen concentration is calculated when the combustible gas completely reacts with the oxygen molecules in the air. It is only an estimated value, which has a certain error with the actual measured minimum oxygen concentration, but it does not lose a reference value when there is no specific experimental data, and it is relatively safe.

Oxygen is one of the three elements of combustion and explosion. If the oxygen content in combustible gas mixture can be controlled beyond the critical oxygen concentration and oxygen volume percentage, explosion can be prevented. The oxygen content of combustible gas is regarded as an important safety technical index in many industrial production occasions. With the frequent occurrence of coal mine gas explosion accidents, there are extensive and in-depth studies on coal mine gas explosion conditions.

When inert gases, such as carbon dioxide and nitrogen, are added to the combustible gas mixture, the oxygen concentration in the combustible gas mixture will be reduced accordingly. At the same time, it will effectively reduce the explosion limit range of the combustible gas mixture, increase the lower explosion limit and decrease the upper explosion limit. The explosion range finally converges to one point. Beyond this point, the mixed gas will exit the explosion range. This point is the explosion limit critical point of combustible gas mixture, and the oxygen concentration corresponding to this point is the explosion limit critical oxygen concentration. According to this law, the method of filling inert gas into the system is often used in safety engineering to dilute the oxygen concentration below the critical oxygen concentration, so as to reduce the possibility of combustion and explosion accidents and improve safety. This measure is called inerting protection. However, for the sake of reliability, the oxygen concentration in the system should be controlled at a level about 10% lower than the critical oxygen concentration in practical application. The critical oxygen content of some combustible gases at room temperature and 1 atmospheric pressure is shown in Table 2.

Table 2

Critical oxygen content of some combustible gases

Combustible gas Critical oxygen content (volume)/% Combustible gas Critical oxygen content (volume)/%
Dilute with CO2 Dilute with N2 Dilute with CO2 Dilute with N2
CH4 14.6 12.1 Propylene 14.1 11.7
C2H6 13.4 11.0 Cyclopropene 13.9 5.0
C3H8 14.3 11.4 H2 5.9 5.6
C4H10 14.5 12.1 CO 5.9 10.4
n-Pentane 14.4 11.4 Butadiene 13.9 11.2
Hexane 14.5 11.6 Benzene 13.9 8.5
Petrol 14.5 11.0 Dimethyl ketone 10.5 13.5
C2H4 11.7 11.5 Propanone 15.0 —

In this study, pure CH4 and the four-component mixed gas of CH4 with CO, C2H6, and H2 are selected, and nitrogen and carbon dioxide are added, respectively. The critical oxygen concentration is measured experimentally and compared with the literature value, as shown in Table 3.

Table 3

Comparison of experimental and theoretical critical oxygen concentration

Theoretical value Experimental value Error
CH4 Dilute with CO2 14.6 16.29 −10.3
Dilute with N2 12.1 14.81 −18.2
CH4 gas mixture Dilute with CO2 16.05 16.05 0
Dilute with N2 14.49 14.49 0

It is recorded in the literature that when diluted with carbon dioxide, the critical oxygen concentration of CH4 is 14.6%. In the experiment, when diluted with carbon dioxide, the critical oxygen concentration of CH4 is 16.29%, which is slightly larger than the literature value, and the error is 10.3%. When diluted with nitrogen, the literature value is 12.1%, the experimental value is 14.81%, and the error is 18.2%. No matter which inert medium is used, the experimental value of critical oxygen concentration is slightly larger than the literature value. The literature value is the data obtained through theoretical calculation in the ideal state, which still has a certain gap with the field practice. However, the experiment realistically simulates the real situation and is closer to the reality, so the error between the theoretical data and the experimental data is within the normal range.

The influence of inhomogeneity of concentration distribution of multicomponent combustible gas on explosion process is the core of the research. At room temperature and pressure, the explosion limit values of multicomponent combustible gases are different. As a highly reactive substance, the density of multicomponent combustible gas is relatively small, which is quite different from the air density. In this section, the explosion limit of multicomponent combustible gas in 5 L cylinder tank is observed [13].

The calculation model of a 5 L cylindrical tank is shown in Figure 1.

Figure 1 A 5 L cylindrical tank model.
Figure 1

A 5 L cylindrical tank model.

As shown in Figure 1, the ignition point is located at 1/5 and 4/5 of the Z-axis height of the tank. The ignition radius is 2 mm, the initial temperature is 2,500 K, the wall thickness is 10 mm, the nonadiabatic wall is 871 J·(kg·K−1), and the wall thermal conductivity is 202.4 W·(m·K−1).

A tetrahedral unstructured grid is used for gas explosion. Because the radius of ignition source is small (2 mm), to refine the ignition source area, the local refined grid of ignition point is adopted for the gas explosion model [14]. The minimum mesh size is 1.1 mm × 1.1 mm × 1.1 mm, and the maximum mesh size is 10 mm × 10 mm × 10 mm, with a growth rate of 1.07. To verify the grid independence, different grid sizes were used to simulate the gas explosion process. The mesh sizes are Grid1 and 1.1 mm × 1.1 mm × 1.1 mm (Grid2), respectively. The mesh numbers are 316,502 and 68,864,968, respectively. The simulation results of different mesh sizes are shown in Figure 2.

Figure 2 Grid validation diagram.
Figure 2

Grid validation diagram.

Grid1 and Grid2 at different times calculate the peak overpressure and its time as shown in Table 4. The relative deviation of overpressure peak value is less than 1.1% and the relative deviation of overpressure peak time is less than 5.5% for different grid sizes.

Table 4

Computation of peak explosion overpressure with Grid1 and Grid2 grids and their time comparison

Grid1 Grid2 Difference value Relative deviation (%)
Peak overpressure 0.0461 MPa 0.0456 MPa 0.0005 MPa 1.09
Overpressure peak time 0.0198 s 0.0188 s 0.001 s 5.32

The grid density has some influence on the gas explosion process, but the relative deviation between the calculation results of peak overpressure and its time is less than 6%. Considering that the grid size is too small and the computing resources are too large, the Grid1 is selected to calculate without affecting the calculation accuracy. Based on the grid size, the influence of multicomponent flammable gas–air concentration distribution on flammability in a 5 L tank was studied [15].

Gas mixing model was used to fill a 5 L tank with 4% macro concentration and 75% multicomponent combustible gas, respectively. After standing for 5 min, the concentration and relative deviation of multicomponent combustible gas vary with the height of Z-axis of the tank as shown in Figures 3 and 4.

Figure 3 Variation of multicomponent combustible gas concentration with Z-axis height of tank.
Figure 3

Variation of multicomponent combustible gas concentration with Z-axis height of tank.

Figure 4 Variation of relative deviation of multicomponent combustible gas with Z-axis height of tank.
Figure 4

Variation of relative deviation of multicomponent combustible gas with Z-axis height of tank.

Figures 3 and 4 show that when the macro concentration of multicomponent combustible gas is 4% after 5 min, the top concentration is 6.07%, the bottom concentration is 1.9%, and the concentration difference is 4.17%. When the macro concentration of multicomponent combustible gas is 75%, the top concentration is 76.8%, the bottom concentration is 73.3%, and the concentration difference is 3.5%. When the macro concentration of multicomponent combustible gas is 4%, the relative deviations of tank top and bottom concentration are 51.75 and 52.5%, respectively. When the macro concentration of multicomponent combustible gas is 75%, the relative deviations of tank top and bottom concentration are 2.4 and 2.27%, respectively. This is because when the macro concentration is 75%, the base of relative deviation is large, so the relative deviation of concentration is small.

Multicomponent combustible gas and air are mixed in a 5 L tank. After standing for 5 min, the concentration distribution varies with the height of the tank. The concentration distribution of multicomponent combustible gas is different in different positions. When ignited at different positions in the tank, the concentration of multicomponent combustible gas at ignition point affects the development trend of high temperature area after ignition. Two ignition points were set in the upper part (4/5 height) and the lower part (1/5 height) of the 5 L tank, respectively. The influence of concentration distribution at different ignition positions on gas flammability at explosion limit concentration was analyzed.

When the macro concentration of multicomponent combustible gas is 4%, the overpressure–time curve when the ignition points are set at the upper and lower parts of the tank is shown in Figure 5.

Figure 5 Overpressure–time curves (4%) when the ignition point is located at the top and bottom of the tank, respectively.
Figure 5

Overpressure–time curves (4%) when the ignition point is located at the top and bottom of the tank, respectively.

As shown in Figure 5, when the ignition point is 4/5 high, the peak overpressure after ignition is 0.046 MPa, and the combustible mixture reacts chemically, because at 4/5 high, the concentration of multicomponent combustible gas is higher than 4%, which is within the explosion limit. The flame propagates downward. Because the concentration of multicomponent combustible gas in the lower part is less than 4%, the quantity of combustible gas is less, and the wall heat dissipates, the flame temperature decreases gradually, and its overpressure decreases gradually after reaching the peak value. When the ignition point is 1/5 high, the concentration of multicomponent combustible gas is less than 4%; beyond the explosion limit, the high-temperature spark area gradually reduces to normal temperature, the combustible mixture does not undergo chemical reaction, and the overpressure curve does not change.

When the macro concentration of multicomponent combustible gas is 75%, the overpressure–time curve when the ignition points are set at the upper and lower parts of the tank is shown in Figure 6.

Figure 6 Overpressure–time curves (75%) when the ignition point is located at the top and bottom of the tank, respectively.
Figure 6

Overpressure–time curves (75%) when the ignition point is located at the top and bottom of the tank, respectively.

When the macros concentration of multicomponent combustible gas is 75%, when the ignition point is 4/5 high, the concentration of multicomponent combustible gas is higher than 75%, which exceeds the explosion limit. The high-temperature spark area is gradually reduced to normal temperature. The multicomponent combustible gas–air mixture does not undergo chemical reaction, and its overpressure curve remains unchanged. When the ignition point is 1/5 high, the peak overpressure after ignition is 0.049 MPa, and the combustible gas reacts chemically, because at 1/5 high, the concentration of multicomponent combustible gas is less than 75%, which is within the explosion limit. The flame propagates upward. Because the concentration of multicomponent combustible gas in the upper part is higher than 75%, the amount of oxygen is less, and the wall heat dissipates, the flame temperature decreases gradually, and its overpressure decreases gradually after reaching the peak value.

Through the above analysis, it is concluded that the explosion limit of multicomponent flammable gas is between 3.6 and 6.5% of the lower explosion limit concentration and 74.7 and 77% of the upper explosion limit concentration.

2.3 Design of multicomponent flammable gas explosion hazard identification device

Based on the explosion limit of multicomponent flammable gas obtained above, a multicomponent flammable gas explosion hazard identification device is designed. The specific design process is shown below.

The automatic detection technology of explosion occurrence can be understood as extracting appropriate and effective features from the explosion process as parameters of detection and identification of explosion. On the basis of previous studies, this section tries to extract the ultraviolet, infrared, and ion signals of the flame in the process of explosion and analyze the explosion process based on these three signals to explore effective methods for identifying explosion hazards. In this section, a multicomponent flammable gas explosion hazard identification device is designed. Three infrared sensors, three ultraviolet sensors, and six ionization probes are simultaneously arranged on the device to detect the infrared, ultraviolet, and ion current characteristic signals of multicomponent flammable gas explosion flame.

Among them, the photosensitive surface of infrared sensor is very small, and its pin is thin and short, easy to fall off, but its photosensitive surface is perpendicular to the pin, so it is easier to fix the direction. The sensitive surface of infrared sensor also needs to be protected from the damage of shock wave. In the design of infrared sensor packaging, the following aspects should be considered: first, considering the connection mode between sensor and protective packaging, not only to ensure good fixed connection but also to ensure the safety of small and short pins of infrared sensor; second, to protect the sensitive surface of infrared sensor, but the sensitive surface is small, so it cannot be used. Overprotection to prevent the effect of detection, and finally, design the sealing surface of the encapsulated parts to ensure the sealing of the whole device. The schematic of the infrared sensor is shown in Figure 9.

As shown in Figure 7, the infrared sensor includes seven parts: extension wire of pin, fixing plug, sealing cavity, protective cover, controller, display, and pin. The pin of sensor is welded together with circular circuit board. The sensor is fixed in the package room through the circuit board. The protective cover with optical glass is assembled with the encapsulation chamber through threads. To make the whole package of the infrared sensor fit into the reserved sensor base on the pipeline, the outer diameter of the protective cover is smaller than the inner diameter of the sensor base on the pipeline. The packaging method of the infrared sensor has a good sealing property, can effectively protect the infrared sensor, and is convenient to use.

Figure 7 Infrared sensor diagrams.
Figure 7

Infrared sensor diagrams.

Compared with the infrared photodiode, the ultraviolet sensor has larger volume, longer and thicker pins, so it is more convenient to weld and assemble the pins in the packaging process. But the ultraviolet sensor has a large light window, which is easy to damage. Once it is damaged, the inert gas in the sensor will leak and cause the sensor to fail. According to the shape, detection principle and working environment of the ultraviolet sensor, three points should be considered when designing the packaging of the ultraviolet sensor: first, the connection mode between the sensor and the protective packaging to ensure that the pins of the sensor are fixed. Second, the window of the ultraviolet sensor is effectively protected so that the ultraviolet rays in the flame can reach. Reaching the anode and cathode, the light window of the sensor is not directly impacted by explosion shock wave. Third, the sealing surface of the encapsulated parts is designed to ensure the sealing of the whole device. The schematic of the ultraviolet sensor is shown in Figure 8.

Figure 8 Schematic diagram of ultraviolet sensor.
Figure 8

Schematic diagram of ultraviolet sensor.

Ionization probes are mainly installed in polytetrafluoroethylene (PTFE) with two probes 3 mm in diameter. The distance between the two probes is 5 mm, and the probe head is exposed 25 mm. The outer diameter of PTFE is the same as the inner diameter of the sensor base used in the pipeline. Then, PTFE is fixed in the pipeline base through a bolt cap with internal threads. The packaged ion probe separation method can not only fix the ion probe well but also ensure that the whole device has a good sealing performance.

The designed infrared sensor, ultraviolet sensor, and ionization probe are arranged around the multicomponent combustible gas according to certain rules, and the related data of the multicomponent combustible gas are collected to provide data support for the identification of the following multicomponent combustible gas explosion hazards.

2.4 Identification of multicomponent flammable gas explosion hazards

Based on the related data of multicomponent combustible gas, the multicomponent combustible gas explosion hazard sources are identified based on discrete probability model. The specific identification process is shown below.

The flow chart of multicomponent flammable gas explosion hazard identification based on discrete probability model is shown in Figure 9.

Figure 9 Flow chart for identifying multicomponent flammable gas explosion hazards based on discrete probability model.
Figure 9

Flow chart for identifying multicomponent flammable gas explosion hazards based on discrete probability model.

As shown in Figure 9, the results of multicomponent flammable gas explosion hazard identification mainly have three parameters: input weight, hidden layer bias, and output weight. The calculation of the output weights only involves matrix inversion and matrix multiplication. The training speed is very fast, and the generalization performance is good. It shows great superiority in solving such problems with complex features and large amount of data. To optimize the structure of the training network model, the differential evolution method is used to optimize the input weights and hidden layer biases. Multiple initial values are randomly generated. The root mean square error between the actual output and the ideal output is used as the fitness index to train the training network model. The training network evolves continuously until the maximum number of population iterations is reached. The optimal parameters are obtained, and the training set is constructed with the optimized parameters, which makes the identification accuracy and sensitivity of multicomponent flammable gas explosion hazard sources higher.

Then, the steps of multicomponent flammable gas explosion hazard identification based on a discrete probability model are as follows:

Input: The collected multicomponent combustible gas-related data sets are recorded as U = { U i ∣ i = 1 , 2 , … , n , U i = { U 1 , U 2 , … , U n } } , the hazard components table T , the number of hidden layer nodes L , the activation function G ( a , b , x ) = exp ( − ∥ x − a ∥ 2 / b ) , the initial training data q , and the size p of the new data set arrived in each sequence.

Output: Multicomponent flammable gas explosion hazard identification results.

Step 1: Input multicomponent combustible gas-related data sets.

Step 2: Initial learning.

  1. Random selection of q data from U as the initial sample set N D .

  2. For each sample in N D , find out whether the hazard factor table T has its information, and generate input weight a i randomly from (0.5, 1], otherwise, generate input weight BB randomly from the number field (0, 0.5].

  3. Initial output matrix H 0 of hidden layer is calculated according to activation function.

  4. The generalized inverse matrix H 0 + of H 0 is calculated and the initial output weight β 0 is calculated.

Step 3: Online learning. Repeat the second part of step 2; calculate the output matrix at this time according to the activation function, and mark it as H 1 ; calculate the output weight β 1 ; and repeat step 3 until there is no new data.

Step 4: Output multicomponent flammable gas explosion hazard identification results.

Through the above process, the multicomponent flammable gas explosion hazard source identification based on discrete probability model is realized, which provides more effective guarantee for people’s safety.

3 Method performance simulation experiment

The above process realizes the design of multicomponent flammable gas explosion hazard identification method based on discrete probability model, but it is uncertain whether it can solve the problems existing in the existing methods. Therefore, the design of simulation comparative experiments is carried out. To make the experiment proceed smoothly, the methods of identifying multicomponent flammable gas explosion hazards based on causal analysis, brainstorming and hazard and operability analysis are used in the experiment. To carry out the experiment smoothly, the experimental process is designed first, as shown in Figure 10.

Figure 10 Experimental flow chart.
Figure 10

Experimental flow chart.

According to the above process, the identification effect of the method is reflected by the sensitivity and accuracy of the explosion hazard identification. The detailed experimental results are analyzed as follows.

3.1 Comparative analysis of recognition sensitivity

The sensitivity index of the identification directly determines the effect of the explosion hazard identification. In general, it is considered that the higher the sensitivity parameter value of recognition is, the better the recognition effect of the method is. The comparison of the value of the identified sensitivity parameters is shown in Figure 11.

Figure 11 Comparison of recognition sensitivity parameter values.
Figure 11

Comparison of recognition sensitivity parameter values.

According to the analysis of Figure 11, the sensitivity parameters of different methods are different. For the sample 1 of combustible gas, the sensitivity parameter value of the method in literature [6] is 3.6, the sensitivity parameter value of the method in literature [7] is 1.3, and the sensitivity parameter value of this method is 7.2. For the sample of combustible gas no. 5, the sensitivity parameter value of the method in literature [6] is 3.2, the sensitivity parameter value of the method in literature [7] is 3.8, and the sensitivity parameter value of this method is 8.1. The sensitivity parameter value of this method is significantly higher than that of other methods, which shows that the method has a better recognition effect.

3.2 Error analysis of multicomponent combustible gas explosion hazard identification

To verify the recognition effect of this method on the multicomponent combustible gas explosion hazard source, the methods of literature [6, 7] and this method are used to verify the recognition error, and the results are shown in Figure 12.

Figure 12 Identification error of explosion hazard.
Figure 12

Identification error of explosion hazard.

It can be seen from the analysis of Figure 12 that the identification error of multicomponent combustible gas explosion hazard source is different under different methods. For the gas sample whose sample number is 1, the error of multicomponent combustible gas explosion hazard identification in a study by Tianhao and Zhijie [6] is 0.4, the error of multicomponent combustible gas explosion hazard identification in an earlier study [7] is 0.45, the error of multicomponent combustible gas explosion hazard identification in this article is 0.42, and the actual error of multicomponent combustible gas explosion hazard identification is 0.42. For the gas sample with sample number 4, the error of multicomponent combustible gas explosion hazard identification in an earlier study [6] is 0.32, the error of multicomponent combustible gas explosion hazard identification in an earlier study [7] is 0.34, the error of multicomponent combustible gas explosion hazard identification in this article is 0.37, and the actual error of multicomponent combustible gas explosion hazard identification is 0.38. The prediction results of this method are close to the actual error values. This shows that the method of this article has a good effect on the identification of explosion hazards.

4 Conclusion

In this study, a method of multicomponent combustible gas explosion hazard identification based on discrete probability model is proposed. Fem3 model is selected to analyze the diffusion mode of multicomponent combustible gas, and the explosion limit of multicomponent combustible gas is observed by test to obtain the upper limit of explosion concentration of multicomponent combustible gas. According to the characteristics of explosion, a multicomponent combustible gas explosion hazard identification device is designed. The discrete probability model is used to identify the explosion hazard of multicomponent combustible gas. The results are as follows:

  1. For combustible gas sample no. 1, the recognition sensitivity parameter value of this method is 7.2. The recognition effect of this method is good.

  2. For the gas sample whose sample number is 1, the error of multicomponent combustible gas explosion hazard identification is 0.42, and the actual error of multicomponent combustible gas explosion hazard identification is 0.42. This shows that the method of this article has a good effect on the identification of explosion hazards.

However, the recognition sensitivity and accuracy of the proposed method still have a large room for improvement, which needs further research and optimization.

  1. Funding information: A fund project of Jiangsu Key Research and Development Program: Research and Application of Key Technologies of Explosion-proof Waste Metal Crusher Equipment (BE2016758) and Research on Key Technologies of Robot Sorting System for Solid Waste (BE2018722).

  2. Author contribution: Ye Wenhua uses fem3 model to analyze multi-component combustible gas diffusion mode, determines the calculation formula of multi-component combustible gas diffusion rate, observes the explosion limit of multi-component combustible gas through test, and obtains the upper limit of explosion concentration of multi-component combustible gas; Yang Fengyong and Chen Yangming designed a multi-component combustible gas explosion hazard identification device; Ye Wenhua, Chen Yangming and Yang Feng conducted comparative experiments based on causal analysis, brainstorming, HAZOP Analysis and the proposed method to verify the recognition sensitivity and accuracy of the proposed method. The experimental results were recorded and analyzed. Ye Wenhua and Chen Yangming wrote the article.

  3. Conflict of interest: The authors declare that they have no competing interests.

  4. Data availability statement: The data used to support the findings of this study are available from the corresponding author upon request.

References

[1] Pulak, N., T. Srijeeta, and C. Pinaki. An adaptive mutation simulated annealing based investigation on Coulombic explosion and identification of dissociation patterns in (CO2) n2 + clusters. Physical Chemistry Chemical Physics, Vol. 19, No. 14, 2017, pp. 327–346.10.1039/C7CP00655ASearch in Google Scholar PubMed

[2] Khatabakhsh, A., Z. Maleki, T. H. Hejazi, and M. Pouyakian. Analysis of hazard identification methods in process industries using analytic network process technique. Iran Occupational Health, Vol. 16, No. 2, 2019, pp. 48–60.Search in Google Scholar

[3] Li, R. Y., K. W. Chau, W. Lu, and D. C. Ho. A research agenda for neuroactivities in construction safety knowledge sharing, hazard identification and decision making. International Conference on Applied Human Factors and Ergonomics, Springer, Cham, 2019, 383–389.10.1007/978-3-030-20473-0_37Search in Google Scholar

[4] Anderson, D. N., D. K. Fagan, M. A. Tinker, G. D. Kraft, and K. D. Hutchenson. A mathematical statistics formulation of the teleseismic explosion identification problem with multiple discriminants. Bulletin of the Seismological Society of America, Vol. 97, No. 5, 2007, pp. 1730–1741.10.1785/0120060052Search in Google Scholar

[5] Yuan, C. F., H. Cui, B. Tao, and W. Wang. Risk source identification method based on the essential safety in emergency process of fire accident for petroleum storage and transportation. 2017 14th International Conference on Service Systems and Service Management (ICSSSM), IEEE, 2017. 10.1109/ICSSSM.2017.7996294 Search in Google Scholar

[6] Tianhao, W. and Z. Zhijie. Design of underground gas explosion location test system based on FPGA. Electronic Measurement Technology, Vol. 42, No. 21, 2019, pp. 6–9.Search in Google Scholar

[7] Qi, X., H. Wang, Y. Tian, and G. Chen. Quantitative division of gas protection area based on risk assessment of equivalent gas cloud explosion. Chemical Industry and Engineering Progress, Vol. 38, No. 3, 2019, pp. 447–454.Search in Google Scholar

[8] Zhong, W. T. Tian, and Z. Wang. Regularity of gas leakage from underground chemical explosions based on dimensional analysis and darcy’s law. Chinese Journal of Computational Physics, Vol. 36, No. 4, 2019, pp. 449–456.Search in Google Scholar

[9] Fu, G., J. H. Wang, and M. W. Yan. Anatomy of Tianjin port fire and explosion: process and causes. Process Safety Progress, Vol. 35, No. 3, 2016, pp. 3547–3554. 10.1002/prs.11837.Search in Google Scholar

[10] Lavinius, I. G., C. Houcine, P. Dumitru, and C. Lupu. Fault detection and identification for fire and explosion detection. 2017 21st International Conference on Control Systems and Computer Science (CSCS), IEEE, 2017. 10.1109/CSCS.2017.13.Search in Google Scholar

[11] Aniket, R., S. C. Rajat, S. Udaya, and R. Naskar. Camera source identification using discrete cosine transform residue features and ensemble classifier. 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), IEEE, 2017. 10.1109/CVPRW.2017.231.Search in Google Scholar

[12] Qian, D. U., D. Shuai, Y. Q. Xiang, X. Pan, and C. Qian. Design and implementation of display platform for transmission line mountain fire disaster information base on support vector machine. Electric Power Information and Communication Technology, Vol. 97, No. 3, 2018, pp. 918–922.Search in Google Scholar

[13] Alananbeh, K. M., N. Boquellah, N. A. Kaff, and A. A. Majid. Evaluation of aerial microbial pollutants in Al-Haram Al-Nabawi during pilgrimage of 2013. Saudi J Biol Sci, Vol. 24, No. 1, 2017, pp. 217–225. 10.1016/j.sjbs.2015.08.003.Search in Google Scholar PubMed PubMed Central

[14] Brunetkin, O., M. V. Maksymov, A. Maksymenko, and M. M. Maksymov. Development of the unified model for identification of composition of products from incineration, gasification, and slow pyrolysis. Eastern-European Journal of Enterprise Technologies, Vol. 4, No. 6, 2019, pp. 25–31.10.15587/1729-4061.2019.176422Search in Google Scholar

[15] Wang, J., X. M. Wu, X. M. Hu, and X. Xu. Event source identification and strength estimation in wireless sensor networks. IEEE International Conference on Networking, IEEE, 2016. 10.1109/ICNSC.2016.7479019.Search in Google Scholar
Received: 2021-03-08
Revised: 2021-09-29
Accepted: 2021-11-23
Published Online: 2022-03-22

© 2022 Wenhua Ye et al., published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Research Articles
  2. Numerical and experimental research on solidification of T2 copper alloy during the twin-roll casting
  3. Discrete probability model-based method for recognition of multicomponent combustible gas explosion hazard sources
  4. Dephosphorization kinetics of high-P-containing reduced iron produced from oolitic hematite ore
  5. In-phase thermomechanical fatigue studies on P92 steel with different hold time
  6. Effect of the weld parameter strategy on mechanical properties of double-sided laser-welded 2195 Al–Li alloy joints with filler wire
  7. The precipitation behavior of second phase in high titanium microalloyed steels and its effect on microstructure and properties of steel
  8. Development of a huge hybrid 3D-printer based on fused deposition modeling (FDM) incorporated with computer numerical control (CNC) machining for industrial applications
  9. Effect of different welding procedures on microstructure and mechanical property of TA15 titanium alloy joint
  10. Single-source-precursor synthesis and characterization of SiAlC(O) ceramics from a hyperbranched polyaluminocarbosilane
  11. Carbothermal reduction of red mud for iron extraction and sodium removal
  12. Reduction swelling mechanism of hematite fluxed briquettes
  13. Effect of in situ observation of cooling rates on acicular ferrite nucleation
  14. Corrosion behavior of WC–Co coating by plasma transferred arc on EH40 steel in low-temperature
  15. Study on the thermodynamic stability and evolution of inclusions in Al–Ti deoxidized steel
  16. Application on oxidation behavior of metallic copper in fire investigation
  17. Microstructural study of concrete performance after exposure to elevated temperatures via considering C–S–H nanostructure changes
  18. Prediction model of interfacial heat transfer coefficient changing with time and ingot diameter
  19. Design, fabrication, and testing of CVI-SiC/SiC turbine blisk under different load spectrums at elevated temperature
  20. Promoting of metallurgical bonding by ultrasonic insert process in steel–aluminum bimetallic castings
  21. Pre-reduction of carbon-containing pellets of high chromium vanadium–titanium magnetite at different temperatures
  22. Optimization of alkali metals discharge performance of blast furnace slag and its extreme value model
  23. Smelting high purity 55SiCr automobile suspension spring steel with different refractories
  24. Investigation into the thermal stability of a novel hot-work die steel 5CrNiMoVNb
  25. Residual stress relaxation considering microstructure evolution in heat treatment of metallic thin-walled part
  26. Experiments of Ti6Al4V manufactured by low-speed wire cut electrical discharge machining and electrical parameters optimization
  27. Effect of chloride ion concentration on stress corrosion cracking and electrochemical corrosion of high manganese steel
  28. Prediction of oxygen-blowing volume in BOF steelmaking process based on BP neural network and incremental learning
  29. Effect of annealing temperature on the structure and properties of FeCoCrNiMo high-entropy alloy
  30. Study on physical properties of Al2O3-based slags used for the self-propagating high-temperature synthesis (SHS) – metallurgy method
  31. Low-temperature corrosion behavior of laser cladding metal-based alloy coatings on EH40 high-strength steel for icebreaker
  32. Study on thermodynamics and dynamics of top slag modification in O5 automobile sheets
  33. Structure optimization of continuous casting tundish with channel-type induction heating using mathematical modeling
  34. Microstructure and mechanical properties of NbC–Ni cermets prepared by microwave sintering
  35. Spider-based FOPID controller design for temperature control in aluminium extrusion process
  36. Prediction model of BOF end-point P and O contents based on PCA–GA–BP neural network
  37. Study on hydrogen-induced stress corrosion of 7N01-T4 aluminum alloy for railway vehicles
  38. Study on the effect of micro-shrinkage porosity on the ultra-low temperature toughness of ferritic ductile iron
  39. Characterization of surface decarburization and oxidation behavior of Cr–Mo cold heading steel
  40. Effect of post-weld heat treatment on the microstructure and mechanical properties of laser-welded joints of SLM-316 L/rolled-316 L
  41. An investigation on as-cast microstructure and homogenization of nickel base superalloy René 65
  42. Effect of multiple laser re-melting on microstructure and properties of Fe-based coating
  43. Experimental study on the preparation of ferrophosphorus alloy using dephosphorization furnace slag by carbothermic reduction
  44. Research on aging behavior and safe storage life prediction of modified double base propellant
  45. Evaluation of the calorific value of exothermic sleeve material by the adiabatic calorimeter
  46. Thermodynamic calculation of phase equilibria in the Al–Fe–Zn–O system
  47. Effect of rare earth Y on microstructure and texture of oriented silicon steel during hot rolling and cold rolling processes
  48. Effect of ambient temperature on the jet characteristics of a swirl oxygen lance with mixed injection of CO2 + O2
  49. Research on the optimisation of the temperature field distribution of a multi microwave source agent system based on group consistency
  50. The dynamic softening identification and constitutive equation establishment of Ti–6.5Al–2Sn–4Zr–4Mo–1W–0.2Si alloy with initial lamellar microstructure
  51. Experimental investigation on microstructural characterization and mechanical properties of plasma arc welded Inconel 617 plates
  52. Numerical simulation and experimental research on cracking mechanism of twin-roll strip casting
  53. A novel method to control stress distribution and machining-induced deformation for thin-walled metallic parts
  54. Review Article
  55. A study on deep reinforcement learning-based crane scheduling model for uncertainty tasks
  56. Topical Issue on Science and Technology of Solar Energy
  57. Synthesis of alkaline-earth Zintl phosphides MZn2P2 (M = Ca, Sr, Ba) from Sn solutions
  58. Dynamics at crystal/melt interface during solidification of multicrystalline silicon
  59. Boron removal from silicon melt by gas blowing technique
  60. Removal of SiC and Si3N4 inclusions in solar cell Si scraps through slag refining
  61. Electrochemical production of silicon
  62. Electrical properties of zinc nitride and zinc tin nitride semiconductor thin films toward photovoltaic applications
  63. Special Issue on The 4th International Conference on Graphene and Novel Nanomaterials (GNN 2022)
  64. Effect of microstructure on tribocorrosion of FH36 low-temperature steels
https://doi.org/10.1515/htmp-2021-0052
Keywords for this article
discrete probability model; multicomponent; combustible gas; explosion; hazard source
Creative Commons
BY 4.0

Articles in the same Issue

  1. Research Articles
  2. Numerical and experimental research on solidification of T2 copper alloy during the twin-roll casting
  3. Discrete probability model-based method for recognition of multicomponent combustible gas explosion hazard sources
  4. Dephosphorization kinetics of high-P-containing reduced iron produced from oolitic hematite ore
  5. In-phase thermomechanical fatigue studies on P92 steel with different hold time
  6. Effect of the weld parameter strategy on mechanical properties of double-sided laser-welded 2195 Al–Li alloy joints with filler wire
  7. The precipitation behavior of second phase in high titanium microalloyed steels and its effect on microstructure and properties of steel
  8. Development of a huge hybrid 3D-printer based on fused deposition modeling (FDM) incorporated with computer numerical control (CNC) machining for industrial applications
  9. Effect of different welding procedures on microstructure and mechanical property of TA15 titanium alloy joint
  10. Single-source-precursor synthesis and characterization of SiAlC(O) ceramics from a hyperbranched polyaluminocarbosilane
  11. Carbothermal reduction of red mud for iron extraction and sodium removal
  12. Reduction swelling mechanism of hematite fluxed briquettes
  13. Effect of in situ observation of cooling rates on acicular ferrite nucleation
  14. Corrosion behavior of WC–Co coating by plasma transferred arc on EH40 steel in low-temperature
  15. Study on the thermodynamic stability and evolution of inclusions in Al–Ti deoxidized steel
  16. Application on oxidation behavior of metallic copper in fire investigation
  17. Microstructural study of concrete performance after exposure to elevated temperatures via considering C–S–H nanostructure changes
  18. Prediction model of interfacial heat transfer coefficient changing with time and ingot diameter
  19. Design, fabrication, and testing of CVI-SiC/SiC turbine blisk under different load spectrums at elevated temperature
  20. Promoting of metallurgical bonding by ultrasonic insert process in steel–aluminum bimetallic castings
  21. Pre-reduction of carbon-containing pellets of high chromium vanadium–titanium magnetite at different temperatures
  22. Optimization of alkali metals discharge performance of blast furnace slag and its extreme value model
  23. Smelting high purity 55SiCr automobile suspension spring steel with different refractories
  24. Investigation into the thermal stability of a novel hot-work die steel 5CrNiMoVNb
  25. Residual stress relaxation considering microstructure evolution in heat treatment of metallic thin-walled part
  26. Experiments of Ti6Al4V manufactured by low-speed wire cut electrical discharge machining and electrical parameters optimization
  27. Effect of chloride ion concentration on stress corrosion cracking and electrochemical corrosion of high manganese steel
  28. Prediction of oxygen-blowing volume in BOF steelmaking process based on BP neural network and incremental learning
  29. Effect of annealing temperature on the structure and properties of FeCoCrNiMo high-entropy alloy
  30. Study on physical properties of Al2O3-based slags used for the self-propagating high-temperature synthesis (SHS) – metallurgy method
  31. Low-temperature corrosion behavior of laser cladding metal-based alloy coatings on EH40 high-strength steel for icebreaker
  32. Study on thermodynamics and dynamics of top slag modification in O5 automobile sheets
  33. Structure optimization of continuous casting tundish with channel-type induction heating using mathematical modeling
  34. Microstructure and mechanical properties of NbC–Ni cermets prepared by microwave sintering
  35. Spider-based FOPID controller design for temperature control in aluminium extrusion process
  36. Prediction model of BOF end-point P and O contents based on PCA–GA–BP neural network
  37. Study on hydrogen-induced stress corrosion of 7N01-T4 aluminum alloy for railway vehicles
  38. Study on the effect of micro-shrinkage porosity on the ultra-low temperature toughness of ferritic ductile iron
  39. Characterization of surface decarburization and oxidation behavior of Cr–Mo cold heading steel
  40. Effect of post-weld heat treatment on the microstructure and mechanical properties of laser-welded joints of SLM-316 L/rolled-316 L
  41. An investigation on as-cast microstructure and homogenization of nickel base superalloy René 65
  42. Effect of multiple laser re-melting on microstructure and properties of Fe-based coating
  43. Experimental study on the preparation of ferrophosphorus alloy using dephosphorization furnace slag by carbothermic reduction
  44. Research on aging behavior and safe storage life prediction of modified double base propellant
  45. Evaluation of the calorific value of exothermic sleeve material by the adiabatic calorimeter
  46. Thermodynamic calculation of phase equilibria in the Al–Fe–Zn–O system
  47. Effect of rare earth Y on microstructure and texture of oriented silicon steel during hot rolling and cold rolling processes
  48. Effect of ambient temperature on the jet characteristics of a swirl oxygen lance with mixed injection of CO2 + O2
  49. Research on the optimisation of the temperature field distribution of a multi microwave source agent system based on group consistency
  50. The dynamic softening identification and constitutive equation establishment of Ti–6.5Al–2Sn–4Zr–4Mo–1W–0.2Si alloy with initial lamellar microstructure
  51. Experimental investigation on microstructural characterization and mechanical properties of plasma arc welded Inconel 617 plates
  52. Numerical simulation and experimental research on cracking mechanism of twin-roll strip casting
  53. A novel method to control stress distribution and machining-induced deformation for thin-walled metallic parts
  54. Review Article
  55. A study on deep reinforcement learning-based crane scheduling model for uncertainty tasks
  56. Topical Issue on Science and Technology of Solar Energy
  57. Synthesis of alkaline-earth Zintl phosphides MZn2P2 (M = Ca, Sr, Ba) from Sn solutions
  58. Dynamics at crystal/melt interface during solidification of multicrystalline silicon
  59. Boron removal from silicon melt by gas blowing technique
  60. Removal of SiC and Si3N4 inclusions in solar cell Si scraps through slag refining
  61. Electrochemical production of silicon
  62. Electrical properties of zinc nitride and zinc tin nitride semiconductor thin films toward photovoltaic applications
  63. Special Issue on The 4th International Conference on Graphene and Novel Nanomaterials (GNN 2022)
  64. Effect of microstructure on tribocorrosion of FH36 low-temperature steels
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 7.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/htmp-2021-0052/html
Scroll to top button