Propagating Characteristic of Premixed Methane-Oxygen Deflagration in the Coal M

Huanjuan Zhao, Yiran Yan, Yinghua Zhang, and Yukun Gao

(1.Mine Emergency Technology Research Center(University of Science and Technology Beijing), State Administration of Production Safety Supervision and Administration, School of Civil and Resources Engineering, University of Science and Technology Beijing, Beijing 100083, China; 2.State Key Laboratory of High-Efficient Mining and Safety of Metal Mines (University of Science and Technology Beijing), Ministry of Education, Beijing 100083, China)

Explosion risk of gas and coal dust exists in more than 60% coal mines and 87.4% state-owned coal mines in China. Hence, the characteristics and prevention of gas explosion need further research[1- 2].

In developed country, mines for experiments of gas and coal dust explosion were established to investigate the flame propagation characteristic and pressure trend in gas and coal dust explosion. In 1982, a total length of 896 m (710 m for experiment) lane, which section was similar to the real coal mine in China, was built by Chongqing Branch of China Coal Mine Research General Institute. It is the first experimental lane and lays the foundation of preventing gas and coal dust explosion in China. Some experimental researches were conducted with that lane[3-5]. Recently, the small size tube was utilized in gas and coal dust experiment. New techniques such as high speed photograph and schlieren were adopted in the experiment of gas explosion and the results were demonstrated more directly and accurately. In the small size tube, the influence of obstacles in the explosion shock wave and flame propagation attracted researchers attention[6]and it is important to study the flow field distribution around the refuge chamber as an obstacle in lane[7]. In the explosion, the evolution of flame and shock wave links to the degree of the accident damage. Compared with experiment, simulation can be of lower cost, more efficient and more convenient in parameters analysis. The propagation of gas explosion shock wave[8-14]and the impact resistance of refuge chamber[15-18]were analyzed and discussed via experiment and simulation. However, the load in some previous studies may not agree with the fact and the characteristics of gas deflagration need to be investigated further.

Self-sustained deflagration can accelerate constantly and transfer to detonation with a proper boundary condition, which should be considered in the accident prevention. In detonation, leading shock wave ignites the mixture while the chemical energy released from ignition, and it can sustain the detonation wave propagating in supersonic[2]. Characteristics of CH4explosion were discussed and discovered by some Chinese researchers[19-22]. Gas can be seen as a hazard leading explosion in coal mine while more than 95% of gas is CH4so that it can be chosen instead of gas in study. Because CH4is instable, which is of low sensitive and high detonation limit pressure, it is difficult to obtain accurate experimental results. There are hardly discussion and conclusion about premixed CH4-O2mixture. In this paper, flow field load distribution for coal mine lane and pressure load for each part of the refuge chamber are analyzed and discussed via experiments in a small size tube and simulations in the real size.

1 Simulation for Flow Field Load of Refuge Chamber in the Coal Mine Lane

The finite element model for premixed CH4+2O2mixture explosion in a lane was established and AUTODYN was utilized to calculate the parameters of explosion wave propagating in the lane. The load against the surface of the refuge chamber was due to an instantaneous explosion source in the simulation. Pressure of CH4+2O2mixture explosion could 20 times higher than the initial pressure by experience, and the accurate numerical results are determined by the initial pressure and type of mixture. If the explosion pressure peak was 0.6 MPa, 0.6×25 MPa should be the initial value in the simulation. The corresponding explosion wave and initial pressure were calculated with the initial pressure 21.6 kPa.

1.1 Flow fieldload model

The arched lane with the height as 2.6 m and the width as 3.2 m was built as Fig.1. Considering the influence on the explosion waveform, the refuge chamber was at the geometric center in the horizontal direction.

Fig.1 Simplified physical arched lane model

In the model,the length of explosion segment was 28 m, while the length of shock wave propagation segment was 100 m and the one of refuge chamber was 22 m. The explosion source and air area were built of solid element, the outlet was determined as outflow and other boundaries were rigid walls. Mesh size was 100 mm. As Fig.2, flow field detection points were set at front and rearface (detection point 2 and 27), the lateral middle position of each section (detection point 3, 4, 5, 6, 7, 20, 21, 22, 23, 24, 25, 26), the top surface of each section (detection point 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19) and it is 1 m in front of the refuge chamber (detection point 1).

Fig.2 Meshing and flow field pressure detection points distribution of the chamber

1.2 Results and discussion

Fig.3 Pressure-time history of the chamber in the lane

The pressure-time history of the chamber is shown in Fig.3. The flow field pressure is shown in Fig.3a. The detection points at the front and rear face are shown in Fig.3b and Fig.3c. There are 12 sections in the refuge chamber. In Fig.3d, curves c(1)-c(12) demenstrate the pressure on the lateral of each section, while curves d(13)-d(24) are the pressures on the top surface. In Fig.4 and Fig.5, the pressure of the shock wave rises suddenly at the front surface, then propagates along the chamber to the rear surface in a short time (10-1s magnitude). The process of impact on the refuge chamber comes to end.

Fig.4 Corresponding pressure-time history of detection points on lateral surface of the chamber

Fig.5 Corresponding pressure-time history of detection points on upside of the chamber

The explosion shock wave form can change with the propagating distance and time. The time of shock wave propagating at each segment is different. These can affect the structural response of refuge chamber under the explosion. In fact, the explosion shock wave propagates constantly so that it is difficult to record waveform in time and calculate the structural response. In the simulation, pressure results are higher and attenuation coefficient is smaller than one of experiments, for which the gravity of gas, heat loss and roughness at tube wall were not considered. The peak pressure change of each segment is in Fig.6.

Fig.6 Peak pressure change of each segment

2 CH4+2O2 Mixture Detonation in Small Size Tube

Fig.7 Design of the rectangular tube

A small size rectangular experimental tube was built (Fig.7) to study the explosion shock wave in the lane. The rectangular tube consists of two plates and an aluminium alloy plate which was slotted on each side to fit the rubber seal ring. In Fig.8, two pressure transducers were fitted on the aluminium plate and the distance is ΔL=20.4 cm, while Δtwas the peak distance of two waveforms on the oscilloscope. CH4+2O2mixture was pre-mixture for more than 24 h to ensure an intensive mixing.

Fig.8 Installation of the rectangular tube

The velocity of CH4+2O2mixture detonation under different initial pressures is shown in Fig.9. The results can prove that the detonation is stable in the experiment.

Fig.9 Velocity of CH4+2O2 mixture detonation under different initial pressures

Δt1=77.6 μs, Δt2=75.2 μs, Δt3=78.8 μs,Δt4=76.8 μs, and Δt5=76.4 μs were measured on the oscilloscope and the pressure-time curves are in Fig.10 and Fig.11. Under the initial pressures 5 kPa and 10 kPa, the peak pressures are 1 MPa and 2 MPa, respectively. The shock wave form in the small tube experiment illustrates that the pressure-time curve will change with the time. The differences about the peak pressure and propagation time cannot be ignored.

Fig.10 Pressure-time curve under initial pressure of 5 kPa

Fig.11 Pressure-time curve under initial pressure of 10 kPa

3 Impact Simulation on the Refuge Chamber

A fluid-structure interaction method was utilized to calculate and analyze the structural dynamic response of the refuge chamber under the pressure load which was calculated in the coal dust explosion simulation. ANSYS/LS-DYNA was used to analyze the response of the refuge chamber structure under the shock wave changing with the time.

3.1 Model parameter

The length×width×height of KJYF-96/12 refuge chamber is 10 662 mm×2 072 mm×1 800 mm, skin and flange is 34 mm, respectively. In the simulation, the model was established as the real size. Solid beams and columns were set as solid elements, hollow beams and columns as shell elements, shell structure as shell elements. Structural mesh was used mostly to discretize the refuge chamber. On the each side of solid beams and columns section were more than 2 rows. The mesh size at shell elements was 5-20 times larger than shell structures. Key area and small components were in the real size. The domain was determined by the boundary if the boundary existed around the refuge chamber was 0.5 times less than the structural maximal size. The maximum size in the shell element was 50 mm, in the solid element as 25 mm, and in the small component as 10-15 mm. Shell element, solid element, and rigid element were 548 975 totally (Fig.12). Component and materials of the refuge chamber were shown in Tab.1 and parameters of materials in Tab.2. Fixed or simply-supported connection was chosen due to the real connection method and the connection points, the parts and method should be identical with the refuge chamber in reality.

3.2 Results and discussion

The shock wave time delay in different parts cannot be ignored in the simulation according to the waveform and characteristics of explosion shock wave mentioned before. In the lane, the calculated flow field pressure load was determined as the pressure load on the refuge chamber. Single domain size was less than one cabin surface size and the maximum load in the domain was considered as the pressure load.

Fig.12 Mesh generation of KJYF-96/12 refuge chamber

StructuresMaterialThickness/mmNoteMaincabindoorQ345-B18FlangeatmaincabindoorQ345-B30EmergencyexitQ345-B30FlangeatemergencyexitQ345-B22FlangeatjointofcabinQ345-B35SkinQ345-B12StiffenerincabinQ345-BWeldedTsteel

Tab.2 Material characteristic parameters of KJYF-96/12 refuge chamber

The detection points diagram is shown in Fig.13a, the field pressure load was calculated and the pressure-time history on all parts are shown in Fig.13b-Fig.13e.

Fig.13 Detection points and pressure-time history on all parts of KJYF-96/12 refuge chamber

Fig.14 Stress distribution for KJYF-96/12 refuge chamber

In Fig.14a-Fig.14f, a stress distribution is demonstrated from low to high. With the explosion wave propagating, stress change can be observed at the front surface of the cabin. The maximum stress is 339.3 MPa, at the connection of the door frame and the connection of door handle, which is lower than the yield strength.

In Fig.15a-Fig.15f, the maximum displacement is at the middle cabin and middle and upper part of the front face, which is 15.12 mm maximumly. Refuge chamber maximum deformation is lower than 20 mm, and no local fracture or crack exist.

Fig.15 Displacement distribution for KJYF-96/12 refuge chamber

4 Conclusions

① The explosion initial pressure was calculated with “Chemical Equilibrium Analysis”. The lane with a refuge chamber model was built in real size and the flow field pressure load and other pressure loads on the cabin were calculated.

② The pressure-time curve in the small size tube agrees with the waveform change in the simulation, which validates the model of CH4+2O2mixture and the explosion load.

③ The inpressure-time curve and the explosion wave form changes with the distance and time. The peak pressure and propagation time are the main different points which can have influences on the simulation. Flow field load history was analyzed and more accurate pressure load on the refuge chamber were calculated, which is important on the method and practical applications for preventing coal mine gas explosion.

[1] Wang Cheng, Hui Yan, Hu Binbin, et al. Effect of obstacle shape on gas explosion flame propagation process[J]. Transactions of Beijing Institute of Technology, 2015, 35(7): 661-676. (in Chinese)

[2] Jiang Zonglin, Teng Honghui．Research on some fundamental problems of the universal frame work for regular gaseous detonation initiation and propagation [J]．Sci Sin-Phys Mech Astron, 2012, 42(4): 421-435. (in Chinese)

[3] Jing Guoxun, Duan Zhenwei, Cheng Lei, et al．Research progress in explosion characteristics and spread law of gas and coal dust[J].China Safety Science Journal, 2009, 19(4): 67-72. (in Chinese)

[4] Si Rongjun. Study on mine gas and coal dust explosion propagation [D]. Qingdao: Shandong University of Science and Technology, 2007. (in Chinese)

[5] Jiang Congguang, Lin Baiquan, Zhai Cheng. Structure variation law of explosion wave in gas explosion[J].Journal of China University of Mining and Technology, 2003, 32(4): 363-366. (in Chinese)

[6] Zheng Youshan, Wang Cheng. Numerical simulation for the influence of variable cross-section tube on explosion characteristics of methane[J].Transactions of Beijing Institute of Technology, 2009, 29(11): 947-949. (in Chinese)

[7] Ciccarelli G, Fowler C J, Bardon M. Effect of obstacle size and spacing on the initial stage of flame acceleration in a rough tube[J]. Shock Waves, 2005 14(3): 161-166.

[8] Xu Shengli, Zhang Hongyue, Yue Pengtao, et al. Study on properties of pressure waves generated by steady flames in a duct[J]. Journal of China University of Science and Technology, 2000, 30(8): 387-392. (in Chinese)

[9] Zhu Chuanjie, Lin Baiquan, Jiang Bingyou, et al. Numerical simulation on the oscillation and shock of gas explosion in a closed end pipe[J]. Journal of Vibration and Shock, 2012, 31(16): 8-12.(in Chinese)

[10] Wang Zhirong. Gas explosion propagation and dynamic process research in confined space [D]. Nanjing: Nanjing University of Technology, 2005. (in Chinese)

[11] Qu Zhiming, Zhou Xinquan, Zhang Yikai, et al. Dynamic mathematical model and numerical analysis of gas explosion in excavation roadway[J]. Chinese Journal of Engineering, 2006, 28(10): 907-911. (in Chinese)

[12] Kobiera A, Kindracki J, Zydak P. A new phenomenological model of gas explosion based oncharacteristics of flame surface [J]. Journal of Loss Prevention in the Process Industries, 2007, 20(3): 271-280.

[13] Chen Dongliang, Sun Jinhua, Liu Yi, et al. Propagation characteristics of premixed methane-air flames[J].Explosion and Shock Waves, 2008, 28(5): 385-390. (in Chinese)

[14] Hong Yidu, Lin Baiquan, Zhu Chuanjie. Simulation on dynamic pressure of premixed methane/air explosion in open-end pipes[J]. Explosion and Shock Waves, 2016, 36(2): 198-209. (in Chinese)

[15] Zhao Huanjuan, Huang Ping, Qian Xinming. Structure safety analysis and optimization of refuge chamber shell under explosion load[J]. Journal of China Coal Society, 2012,38(6): 1095-1100. (in Chinese)

[16] Zhao Huanjuan, Qian Xinming. Simulation analysis on structure safety of coal mine mobile refuge chamber under explosion load[J].Safety Science, 2012,50(4): 674-678. (in Chinese)

[17] Zhao Huanjuan, Qian Xinming, Huang Ping, et al. A one-piece coal mine mobile refuge chamber with safety structure and less sealing risk based on FEA[J].Journal of Beijing Institute of Technology, 2014, 23(2):152-157.

[18] Rong Jili, Liu Qian, Xiang Dalin. Numerical simulation on the dynamic response of mine rescue chamber subjected to explosion impact in the tunnel[J].Journal of Vibration and Shock, 2016, 35(11): 28-33. (in Chinese)

[19] Wu Yuwen, Lee John H S. Stability of spinning detonation waves [J]. Combustion and Flame, 2015, 162(6):2660-2669. (in Chinese)

[20] Lee John H S. The detonation phenomenon [M]. New York: Cambridge University Press, 2008: 29-38, 173-175, 194-200.

[21] Zhang Bo, Shen Xiaobo, Pang Lei, et al. Methane-oxygen detonation characteristics near their propagation limits in ducts [J].Fuel, 2016, 177(1): 1-7. (in Chinese)

[22] Anatoliy V Trotsyuk, Pavel A Fomin, Anatoly A Vasil’ev. Numerical study of cellular detonation structures of methane mixtures[J].Journal of Loss Prevention in the Process Industries, 2015, 36: 394-403.