Suppression of nano-polymethyl methacrylate dust explosions by ABC powder

Accepted Manuscript Title: Suppression of nano-polymethyl methacrylate dust explosions by ABC powder Authors: Jianhua Zhou, Bei Li, Daqing Ma, Haipeng Jiang, Bo Gan, Mingshu Bi, Wei Gao PII: DOI: Reference:

S0957-5820(18)30748-1 https://doi.org/10.1016/j.psep.2018.11.023 PSEP 1584

To appear in:

Process Safety and Environment Protection

Received date: Revised date: Accepted date:

27 August 2018 22 November 2018 27 November 2018

Please cite this article as: { https://doi.org/ This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Suppression of nano-polymethyl methacrylate dust explosions by ABC powder

SC R

IP T

Jianhua Zhoua, Bei Lia, Daqing Maa,b, Haipeng Jianga, Bo Gana, Mingshu Bia, Wei Gaoa*

a School

of Safety Science and Technology, Beijing, 100012, China

A

N

b China Academy

U

of Chemical Machinery and Safety Engineering, Dalian University of Technology, Dalian, Liaoning 116024, China

*

PT

ED

M

Corresponding author: School of Chemical Machinery and Safety Engineering, Dalian University of Technology, Dalian 116024, China. Tel./Fax:+86 411 84986501. E-mail address: [email protected] (Wei Gao)

Highlights

The suppressed flame configurations, propagation velocities and temperatures in 100 nm PMMA dust explosions were investigated.



The critical suppression proportion of ABC powder on 100 nm PMMA dust explosions was determined.



Detailed suppression mechanism was revealed.

A

CC E



Abstract: The suppression effects of ABC powder on 100 nm polymethyl methacrylate (PMMA) dust explosions with different mass densities were experimentally studied. Results showed that the suppressed flames were dim and the flame propagation was slowed down with the addition of ABC powder. When

IP T

the proportion of suppressant increased, the average propagation velocity was decreased. After adding 30% ABC powder to 650 g/m3 PMMA dust cloud, the maximum suppression ratio for velocity was 93%. After adding 10% ABC powder, the temperatures of PMMA dust cloud with the mass densities of 250 g/m3, 450 g/m3 and 650 g/m3 were decreased to 19.8%, 27.4% and 33.1%, respectively. Meanwhile, the critical suppression proportion of PMMA dust explosion with three mass densities were 25%, 35% and 35%. Thermal analysis results showed that the addition of suppressant could increase the thermal stability of PMMA. The suppression mechanisms were comprised of physical suppression and chemical suppression which reflected on the consumption of H and OH radicals. The chemical kinetic models indicated that PO2 and HOPO played a catalytic role in the combination of H and OH.

SC R

Keywords: Nano-PMMA dust explosion; Explosion suppression; Flame propagations; Suppression mechanisms

1. Introduction

A

CC E

PT

ED

M

A

N

U

A range of new types of nano-materials are currently being produced and used in a number of plastic products, consumer products, solar cell products, microchip heat conductive substrates and their wiring products, high toughness ceramic products, catalyst products and body repair products (Eckhoff, 2011; Palanchoke et al., 2014). Compared with micro-materials, nano-materials exhibit larger specific surface area, higher surface energy and larger proportion of surface atoms, which lead to higher potential explosion severity (Eckhoff, 2012). As a consequence, “nano-safety” has attracted close attention. Most researchers focused on the measurement of characteristic parameters in nano-dust explosions. Bouillard (Bouillard et al., 2010) found that as the aluminium particle size decreased, minimum ignition temperature and minimum ignition energy decreased, indicating higher potential inflammation and explosion risks for the use of nano-dust. Wu (Wu et al., 2009) measured minimum ignition energy for Ti and Fe powders using a modified version of the 1.2-L Hartman apparatus. The experimental results indicated that nano-dust had the low minimum ignition energy. Jiang (Jiang et al., 2011), Mittal (Mittal, 2014) and Wu (Wu et al., 2010) investigated the maximum explosion pressure and the maximum rate of pressure rise of Al and Mg powders which indicated that the explosion severity notably increased as the particle size decreased to nano-size. Gao (Gao et al., 2016) pointed out that the maximum explosion pressure of nano-PMMA dust explosions were larger than those of micron dust explosions. Explosion mitigation and prevention are urgently developed for safe handling of nano-materials (Gao et al., 2015). Explosion suppression is one of the important methods to mitigate and prevent dust explosions. There are various categories of suppressants, including inert gases, water mist, halon replacements and non-combustible dust (Amyotte,

IP T

1996; Gatsonides et al., 2015; Li et al., 2017; Moore, 1996). As the main component of ABC extinguishing agent, ammonium phosphate powder is widely used in industrial fire extinguishment due to the excellent chemical and physical properties (Amerogowicz, 1991). As a medium for suppressing the flame propagations of coal dust deflagration, ABC powder could reduce the flame propagation velocities significantly, and prevent the transition from deflagration to detonation, thus eventually extinguish the flame propagated (Zhang, et al., 2012). Luo (Luo et al., 2014) found that the ABC powder exhibited an excellent suppression effect on the mine gas explosion. After injecting 0.25g/L ABC powder to the methane atmosphere, the explosion limit reduced by 13%.

SC R

From the above-mentioned studies, it was presented that ABC powder was the promising suppression. In this study, 100 nm PMMA dust explosions suppressed by ABC powder was experimentally studied. The suppressed flame configurations, flame propagation velocities and temperatures were observed. In addition, suppression mechanisms were revealed.

N

2.1 Experimental apparatus and procedures

U

2. Experimental

ED

M

A

The open-space dust explosion experimental system (Fig. 1) same as our previous study (Zhang et al., 2016) was used to study the suppression effects of ABC powder on 100nm PMMA dust explosions. It included a gas supplying system, a temperature measurement system, combustion tubes, an ignition system, a high-speed photography system, a data acquisition system and a time controller system.

A

CC E

PT

Before experiments, the experiment materials were dried in a drying oven for at least 24 h at 35 ℃ to reduce the agglomeration between particles. The experimental procedures were as follows: (1) weighted a certain amount of experimental materials; (2) mixed them fully in the mixing vessel for at least 20 minutes and put them in the basement uniformly; (3) opened the pressure reducing valve to make the air pressure in the buffer vessel was 0.5 MPa; (4) turned on the high –speed photography cameras and the data acquisition instrument; (5) started the time controller system. The subsequent processes were controlled by an OMRON CPM1A programmable logic controller. At 0 s, gas supplying system was triggered. Duration of dust dispersion was 0.5 s at the pressure of 0.5 MPa. After 0.2 s, the middle tube was moved down and the open space was formed. The tungsten electrode with 15kV high voltage was discharged for 0.01 s. Subsequently, the relatively stable and uniform dust cloud was ignited. Flame propagations were recorded by high-speed photography system. 2.2 Experimental materials The experimental materials were 100 nm (MP-300) PMMA particles and the ABC powder. As the main content of ABC powder, the content of ammonium

dihydrogen phosphate (NH4H2PO4) was more than 90%. The remaining ingredients were SiO2 and CaCO3. The theoretical mass densities 250 g/m3, 450 g/m3 and 650 g/m3 of 100 nm PMMA particles were used in our study. The relationships between actual dust concentration and nominal dust concentration had been presented at our previous study (Zhang, 2017). When the diameter of PMMA particle was 100 nm, the relationship could be stated as:

ycon, 100nm  21.13  0.59 x

(250 g/m3 < x < 650 g/m3)

(1)

SC R

IP T

The real concentrations of the 250 g/m3, 450 g/m3 and 650 g/m3 were 126 g/m3, 244 g/m3, 362 g/m3, respectively. As shown in Figure. 2, Malvern Mastersizer 2000 laser diffraction analyzer was used to measure the particles size distributions of the dusts. Table 1 lists the volume diameter D[4,3](μm), the Sauter diameter D[3,2] and the quantiles of the volumetric distribution Dv (10), Dv (50) and Dv (90). It was obviously found that the Sauter diameter of 100nm PMMA particles was 10.486 μm which was larger than the theoretical value due to the serious agglomeration.

ED

2.3 Kinetic model

M

A

N

U

The micrographs of dust particles shown in Fig.3 were observed by scanning electron microscopy (FEI Nova NanoSEM 450). It was observed in Fig.3 (a) that 100nm PMMA particles were in regularly spherical shape. By magnifying a single pure ABC particle with 40000 times, it was found that the surface of ABC particle was covered with some scraps like tiny ABC particles (Fig.3 (b)). When mixing PMMA particles with ABC powder, it could be seen that a lot of PMMA particles were attached to the ABC particle surface (Fig.3 (c)).

A

CC E

PT

At nano-scale, PMMA particles would be decomposed to small gaseous molecules. The kinetic model for suppression of PMMA dust explosions by NH4H2PO4 was assembled using several kinetic sub-models. It was assumed that PMMA particles were completely decomposed to monomeric MMA (Lenmann et al., 1961). Some monomeric MMA continued to be decomposed to small molecules including CH4, CH2=CHCH3, CH2=C(CH3)2, CH3OH, HCHO et al. (Zeng et al., 2016a, b). Kinetic data for reactions with these small molecules were updated using the National Institute of Standards and Technology chemical kinetics database (NIST, 2001). In order to simplify the decomposition kinetic of NH4H2PO4, it was assumed that suppression particles had been completely decomposed in the flame zone. Decomposition of NH4H2PO4 was represented in the kinetic model by the overall chemical processes: NH4H2PO4 → H3PO4 + NH3 ↑. The chemical kinetic mechanism for phosphorus species was based on the works of TWAROWSKI and KOROBEINICHEV (O. P. Korobeinichev, 2000; Twarowski, 1995). Kinetic data for reactions with NH3 was adopted from Konnov (Konnov, 2009). Thermochemistry data was adopted from the JANAF table. For the numerical simulations, 0‒D Homogeneous reactor in the Chemkin software package (CHEMKIN‒PRO) was used, which numerically integrated

the mass and energy conservation equations for a homogeneous mixture at atmospheric pressure.

3. Results and discussions 3.1 Suppression effects of ABC powder on 100 nm PMMA dust flame configurations

N

U

SC R

IP T

Due to the similar flame propagation behaviors, the mass density of 450 g/m3 was selected as the typical case. As shown in Fig.4, ABC powder had significant effects on 100 nm PMMA dust configurations. For pure 100nm PMMA dust explosions (Fig. 4 (a)), the flame structures were continuous to the antic orange spherical shapes at 10 ms. It indicated that the smaller nano-PMMA particles adjacent to the ignition electrode were decomposed firstly and react as the premixed gas combustion. At 20 ms, with the influence of heat radiation, heat convection and turbulence, a large number of agglomerated PMMA particles were split into small particles and maintained the premixed combustion reactions. As a result, the flame propagation was accelerated and interior of flame exhibited lots of continuous yellow luminous spot flames. With the rapid propagation of flame, the front of 100 nm PMMA dust flame almost reached the upper boundary of the open combustion space at 60 ms.

A

CC E

PT

ED

M

A

After suppressing by the 10% ABC powder, the flame was shown in Fig. 4 (b). The flame as a whole showed a continuous dim orange spherical shapes at any moment. A portion of the PMMA particles near the ignition electrode were decomposed and ignited, exhibiting a small and dim flame at 10 ms. After 20 ms, the yellow luminous spot flames were observed. The scale of flame at 60ms was same as the flame size of pure PMMA combustion flame at 30ms due to the slow flame propagation. These phenomenon indicated that ABC powder produced excellent suppression effects. It could be inferred that a part of heat which released by the combustion of PMMA particles would be absorbed by the ABC decomposition. Consequently, the rate of PMMA decomposition was slowed down, which in turn slowed down the flame propagation velocity. The suppression effects of 20% ABC powder on flame propagation behaviors shown in Fig. 4 (c) were especially significant. Combined with additives of 10% ABC powder, it was found that the flames of upper were dim but the lower parts were luminous yellow flames. Under the suppression effects of ABC powder, the flame propagation velocity was decreased. In this case, gravity caused the unburned agglomerated particles to continue falling. As a result, the premixed gas combustion of small particles occurred at the upper parts. At the lower parts, agglomerated particles were split into a large number of small particles to form a high concentration of flammable volatiles and further reacted with O2 to exhibit yellow luminous flames. From the microstructure of the flame shown in Fig. 5(a), it was observed that the flame front of pure PMMA dust cloud was clear and smooth. The rear flame zone presented lots of continuous yellow luminous spot

flames. On the contrary, for the flame microstructure of 450 g/m3 PMMA suppressed by 10% ABC shown in Fig. 5(b), the flame front was fuzzy and presented discrete structure. Only a part of the rear flame area was yellow continuous luminous flame, and the rest area was the dim orange flame. 3.2 Suppression effects of ABC powder on 100 nm PMMA dust flame propagation velocities

ED

M

A

N

U

SC R

IP T

Robert operator of MATLAB was used to recognize the flame edges. Figure. 6 shows the flame propagation velocities of 100 nm PMMA dust clouds with different mass densities suppressed by different proportions of ABC powders. The propagated velocities were pulsated in all conditions, which were caused by the feedback of radiation heat transfer between particles (Hironao Hanai, 2000). When the flame velocity was rapid, it took insufficient time for particles to be heated by radiation from burned regions. As a consequence, the flame velocity was slowed down. When the flame velocity was slow, the unburned particles, sufficiently heated by the burned regions, could be decomposed and reacted with O2 so that the flame was accelerated. After that, the flame met the cold particles and the velocity was again decreased. However, as the proportion of ABC powder increased in the dust cloud, the magnitude of flame velocity oscillations decreased which contributed to the endothermic decomposition of ABC powder. As the proportion of suppressant increased, the average propagation velocity decreased, indicating that the suppression effect gradually increased. As the proportion of ABC increased, more heat was absorbed by the pyrolysis of ABC powder and the rates of PMMA decomposition were slowed down, resulting in the decrease of average velocities. To evaluate the suppression effects of ABC powder on flame propagation velocities, the suppression ratio  was defined as:

PT

v0  vi  100% (i=1, 2, 3) v0

CC E

i =

(2)

Where v0 was the average pulsating flame propagation velocities of pure

A

PMMA particles; vi was the average pulsating flame propagation velocities of PMMA dust with different proportions of ABC powder. The values of  with PMMA dust cloud mass densities of 250 g/m3, 450 g/m3 and 650 g/m3 are shown in Table 2. The greater the value of  , the better suppression effect. The maximum suppression ratio for velocity was 93%. The values of  with PMMA mass density of 650 g/m3 were all larger than those of 450 g/m3. O2 content kept constant in a confined area. When the mass density of PMMA dust under the conditions of excess, subjected to limited O2 content, a part of PMMA particles could not participate in the combustion reaction process. But these particles

consumed a large number of energy through endothermic reactions and collisions so that the lower average pulsating flame propagation velocities. 3.3 Suppression effects of ABC powder on 100 nm PMMA dust flame temperatures The temperature value measured by the thermocouple can be compensated as follows (Zhang et al., 2017): dTm dt

(3)

IP T

T  Tm  8.75  103

SC R

A series of typical instantaneous temperature curves of the PMMA dust cloud mass densities of 250 g/m3, 450 g/m3 and 650 g/m3 with different proportions of ABC powder corrected by Eq. (2) are shown in Fig. 7. The time when the electrode discharging ended was regarded as the initial moment of flame propagation and the position at which the temperature began to increase indicated the front of the preheated zone.

A

CC E

PT

ED

M

A

N

U

The maximum temperatures of the 250 g/m3, 450 g/m3 and 650 g/m3 of 100 nm PMMA dust explosions were dramatically decreased with the addition of ABC powder. The maximum flame temperatures of 250 g/m3, 450 g/m3 and 650 g/m3 PMMA dust explosion without ABC powder were 1305 ℃, 1691 ℃ and 1539 ℃ respectively. After adding the 10 % ABC powder, the maximum flame temperatures of 250 g/m3, 450 g/m3 and 650 g/m3 PMMA dust explosion were 1047 ℃, 1228℃ and 1029℃, respectively. The temperatures dropped separately 19.8%, 27.4% and 33.1%. It was found that the more mass proportion of ABC powder was added, the lower temperatures were. The critical suppression proportion of ABC powder with 100 nm PMMA mass densities of 250 g/m3, 450 g/m3 and 650 g/m3 were 25%, 35% and 35%. Continuing to increase the proportion of suppressant, flame were no longer propagated and the temperatures could not be measured due to the suppression effect of ABC powder. Meanwhile, as the proportion of suppressant increased, the time at which the temperature began to rise was also delayed. After adding ABC powder to the dust clouds, PMMA particle surfaces could be carbonized by the ammonium phosphate which was the product of ABC decomposition. Under the protection of the carbide layer, the flame propagation was slowed down and the flame temperature was reduced (Zhou et al., 2004). 3.4 Suppression mechanisms SDT-DSC Thermal analyzer was applied to evaluate the thermal characteristics of ABC powder, 100 nm PMMA particles and the mixtures. The temperature increased from the room temperature to 600 ℃ with rate of 10 K/min at atmospheric pressure. The TG, DTG and DSC curves are shown in Fig. 8. As shown in Fig. 8 (a), the ABC powder was continuously decomposed

SC R

IP T

from the room temperature to 600 ℃.The rate of decomposition was accelerated significantly from 128 ℃ to 600 ℃. According to the DTG curve, it was found that the temperatures corresponding to DTG peaks were the temperature values with the maximal weight loss rate. The pyrolysis process of ABC powder could be divided into three stages, 128-254 ℃,254-445 ℃ and 445-600 ℃. Meanwhile, the weight losses of ABC powder in the three stages were 16%, 15% and 28%, respectively. NH4H2PO4 was decomposed to ammonia (NH3) and phosphoric acid (H3PO4) during the initial thermal decomposition stage from 128 ℃ to 254 ℃ . In the next stage, the phosphoric acid was quickly decomposed to pyrophosphoric acid (H4P2O7) and H2O from 254 ℃ to 445 ℃ .In the last stage, pyrophosphoric acid was decomposed to metaphosphoric acid (HPO3) and H2O from 445 ℃ to 600 ℃ . The decomposition processes mentioned above are as follows.

NH 4 H 2 PO4  NH3   H3PO4 (128-254 ℃) 2H3PO4  H 4 PO 2 7  H 2O 

(225-444 ℃)

U

(5)

(445-600 ℃)

N

H 4 PO 2 7  2 HPO3  H 2O 

(4)

(6)

A

CC E

PT

ED

M

A

Fig. 8 (b) gives the thermal analysis of 100 nm PMMA dust particles. The rate of decomposition was accelerated significantly from 218 ℃ to 504 ℃. According to the DTG curve, the decomposition process of PMMA particles could be divided into two stages, 218-406 ℃ and 406-504 ℃. The first stage was regarded as the rapid weight loss stage when the weight loss rate increased first and then decreased. The second stage was considered as the slow weight loss stage when the DSC curve presented an endothermic peak. It might be contribute to the heat absorption of the agglomerate particles. Figure. 9 gives the TG curves, it was defined the temperature corresponding to the mass loss of 10 % (T0.1) was a criterion for judging the thermal stability (Dai et al., 2013; Yang et al., 2017). The T0.1 for PMMA, PMMA/10 % ABC and PMMA/20 % ABC dust clouds were 286 ℃, 313 ℃ and 318 ℃, respectively. The amount of residue increased when the proportion of ABC increased. The results indicated that the thermal stability of PMMA could be augmented by the introduction of ABC powder or that was to say the greater mass proportion of ABC powder, the lower decomposition rates. The suppression mechanisms of ABC powder were revealed from the combined efforts of physical and chemical suppression effects as shown in Fig. 10. For the physical effects, PMMA particle surfaces could be carbonized by the ammonium phosphate. Under the protection of the carbide layer, the flame temperatures were reduced and the flame propagations were slowed down. Meanwhile, ABC powder was decomposed instantly in the beginning and absorbed heat from the reaction zone, resulting in a reduction in the temperature

of the reaction zone. On the other hand, HPO3, decomposed products of NH4H2PO4 were attached to the surfaces of the PMMA particles with high temperature. It would be melted to form the glassy covering layers which could infiltrate to the small holes of PMMA particles (Zhou et al., 2004). In addition, the layers secluded the PMMA particle from the air atmosphere.

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Nano-PMMA particles absorbed heat from the combustion reactions and decomposed to generate the monomer methacrylate (MMA) and small molecules. The combustion reactions of these small molecules (CH4, CH2=CHCH3, CH2=C(CH3)2, CH3OH, HCHO et al.) consisted of many elementary reactions in which the H and OH radicals were the main active molecules. The kinetic model for suppression of PMMA dust explosions by NH4H2PO4 was established to reveal the chemical suppressive effects. Figure. 11 presents the mole fractions of H and OH with and without 10% ABC powder for PMMA dust cloud at the mass density of 450 g/m3. NH4H2PO4 reduced the concentration of the radicals fairly in the two conditions. Figure. 12 is the maximum rate of production (MROP) for H and OH. Without ABC powder, the reaction with the highest rate of H atom production was R2: OH+H2=H+H2O, and the reaction with the highest rate of OH atom production was R9: H+O2=O+OH. After adding 10% ABC powder to the PMMA dust cloud, the MROP of H and OH drastically decreased. The negative MROP represented consumption rates with H and OH. Reaction 6-8 consumed a large amount of H radicals. Similarly, reaction 13-16 consumed lots of OH radicals. R8 was the reaction which consumed the highest rate of H, indicating that R8 (H+PO3=HO+PO2) played a key role in suppressing PMMA combustion reaction. It could also be seen from the reaction 15-16 that NH3 binded a portion of OH and reduced the concentration of OH. Furthermore, P-containing species promoted catalytic recombination of radicals (Twarowski, 1995). Figure. 13 is the rate of production with PO2. The primary reaction for PO2 production was OH+HOPO=H2O+PO2. The same key reaction for PO2 consumption was H+PO2+M=HOPO+M. As a result, it formed a catalytic cycle with HOPO  PO2 where the net effect was that H and OH recombined to form H2O. Figure. 14 shows the suppression cycles. The two main suppression cycles are as follows:

PO2  H  HOPO and HOPO  OH  PO2  H 2O

(7)

PO2 +OH  HOPO2 and HOPO2  H  PO2  H 2O

(8)

As can be seen in both cycles, the phosphorus compounds were acting catalytically to recombine H and OH to form H2O. Some H+H recombination between PO2 and HOPO was also observed. Therefore, chemical suppression was mainly carried out by the consumption of H and OH radicals, in which the catalytic cycling of phosphorus-containing substances was an important factor.

4 Conclusions In this study, the suppression effects of ABC powder on 100 nm PMMA dust explosions were experimentally investigated. The systematic conclusions could be described as follows:

IP T

(1) Suppressant had significant effects on 100 nm PMMA dust explosion flame configurations. The flames exhibited lighting luminous continuous shapes and the flame propagations were rapid in the pure 100 nm PMMA dust explosions. After adding suppressant, the flame as a whole showed a continuous dim orange spherical shapes and the flame propagations were slowed down.

N

U

SC R

(2) As the proportion of suppressant increased, the average propagation velocities and the maximum flame temperatures decreased, indicating that the suppression effect gradually increased. After adding 10% ABC powder, the flame temperatures dropped 19.8%, 27.4% and 33.1% in PMMA dust explosions with the mass densities of 250 g/m3, 450 g/m3 and 650 g/m3. The critical suppression proportions of PMMA dust explosions with three mass densities were 25%, 35% and 35%, respectively.

PT

ED

M

A

(3) The suppression mechanisms of ABC powder were revealed from the combined efforts of physical and chemical suppressive effects. For the physical effects, PMMA particle surfaces were carbonized by the ammonium phosphate and ABC powder was decomposed instantly in the beginning and absorbed heat from the reaction zone. The decomposed products of ABC powder could infiltrate to the small holes of PMMA particles. The chemical effects were mainly carried out by the consumption of H and OH radicals in which the catalytic cycling of phosphorus-containing substances was an important factor.

A

CC E

According to our experimental investigation, the critical suppression proportions of ABC powder for nano-PMMA dust explosions and the key reactions of suppression at chemical reaction kinetics were determined. Hence, the obtained results in this study can be utilized in practice, for instance in the installations used for the explosion suppression of nano-PMMA dust-air mixtures to determine the minimum mass of the filled suppressant. Meanwhile, it will promote the research of surface modification and new suppressants to improve the capacity of explosion suppression in the future.

Acknowledgments The authors gratefully acknowledge the financial support from the National Key R&D Program of China (No. 2017YFC0804705), National Natural Science Foundation of China (No. 51674059), the Fundamental Research Funds for the Central Universities (DUT16RC(4)04).

A ED

PT

CC E

IP T

SC R

U

N

A

M

References Amerogowicz, J., Amerogowicz, W., 1991. Effectiveness of Dust Explosion Suppression by Carbonates and Phosphates. Combust. Flame 85, 520-522.

IP T

Amyotte, P.R., 1996. Introduction to the special issue on dust explosions. Theme Hazard evaluation, prevention and mitigation of dust explosions. J. Loss Prev. Process Ind. 9, 1-2.

SC R

Bouillard, J., Vignes, A., Dufaud, O., Perrin, L., Thomas, D., 2010. Ignition and explosion risks of nanopowders. J. Hazard. Mater. 181, 873-880.

Chemical Kinetics Database, 2001. NIST Standard Reference Database 17. http://kinetics.nist.gov/

N

U

Dai, K., Song, L., Jiang, S., Yu, B., Yang, W., Yuen, R.K.K., Hu, Y., 2013. Unsaturated polyester resins modified with phosphorus-containing groups: Effects on thermal properties and flammability. Polym. Degrad. Stab. 98, 2033-2040.

M

A

Eckhoff, R.K., 2011. Are enhanced dust explosion hazards to be foreseen in production, processing and handling of powders consisting of nano-size particles? J. Phys. Conf. Ser. 304.

PT

ED

Eckhoff, R.K., 2012. Does the dust explosion risk increase when moving from μm-particle powders to powders of nm-particles? J. Loss Prev. Process Ind. 25, 448-459.

CC E

Gao, W., Yu, J., Zhang, X., Li, J., Wang, B., 2015. Characteristics of vented nano-polymethyl methacrylate dust explosions. Powder Technol. 283, 406-414. Gatsonides, J.G., Andrews, G.E., Phylaktou, H.N., Chattaway, A., 2015. Fluorinated halon replacement agents in explosion inerting. J. Loss Prev. Process Ind. 36, 544-552.

A

Gao, W., Li, J., Li, Y., Yan, X., Yu, J., Zhang, X., 2016. Explosion characteristics of nano-PMMA particles. Combust. Explo. Shock. 52, 117-123. Hironao Hanai, H.K.a.T.N., 2000. A numerical study of pulsating flame propagation in mixtures of gas and particles. P. Combust Inst. 28, 815-822. Jiang, B., Lin, B., Shi, S., Zhu, C., Li, W., 2011. Explosive characteristics of nanometer and micrometer aluminum-powder. Mining Sci. Tech. (China) 21, 661-666.

Konnov, A.A., 2009. Implementation of the NCN pathway of prompt-NO formation in the detailed reaction mechanism. Combust. Flame 156, 2093-2105. Lenmann, F.A., Brauer, G.M., 1961. Analysis of pyrolyzates of polystyrene and poly(methylmathacrylate) by gas chromatography. Anal. Chem. 33, 673-676. Li, J., Zhou, F., Li, S., 2017. Experimental study on the dust filtration performance with participation of water mist. Process Saf. Environ. Prot. 109, 357-364.

SC R

IP T

Luo, Z.M., Wang, T., Tian, Z., Cheng, F., Deng, J., Zhang, Y., 2014. Experimental study on the suppression of gas explosion using the gas–solid suppressant of CO2/ABC powder. J. Loss Prev. Process Ind. 30, 17-23. Moore, P.E., 1996. Suppressants explosions for the control of industrial. J. Loss Prev. Process Ind. 9, 119-123.

U

Mittal, M., 2014. Explosion characteristics of micron- and nano-size magnesium powders. J. Loss Prev. Process Ind. 27, 55-64.

M

A

N

O. P. Korobeinichev, S.B.I., T. A. Bolshova, V. M. Shvartsberg and A. A. Chernov 2000. The chemistry of the destruction of organphorus compounds in flames—Ⅲ：The destruction of DMMP and TMP in a flame of hydrogen and oxygen. Combust. Flame 121, 593-609.

PT

ED

Palanchoke, U., Kurz, H., Noriega, R., Arabi, S., Jovanov, V., Magnus, P., Aftab, H., Salleo, A., Stiebig, H., Knipp, D., 2014. Tuning the plasmonic absorption of metal reflectors by zinc oxide nano particles: Application in thin film solar cells. Nano Energy 6, 167-172.

CC E

Twarowski, A., 1995. Reduction of a phosphorus oxide and acid reaction set. Combust. Flame 102, 41-54. Wu, H.C., Chang, R.-C., Hsiao, H.-C., 2009. Research of minimum ignition energy for nano Titanium powder and nano Iron powder. J. Loss Prev. Process Ind. 22, 21-24.

A

Wu, H.C., Ou, H.-J., Peng, D., Jr., Hsiao, H.-C., Gau, C.-Y., Shih, T.-S., 2010. Dust Explosion Characteristics of Agglomerated 35 nm and 100 nm Aluminum Particles. Inter J. Chem. Eng. 2010, 1-6. Yang, M., Chen, X., Wang, Y., Yuan, B., Niu, Y., Zhang, Y., Liao, R., Zhang, Z., 2017. Comparative evaluation of thermal decomposition behavior and thermal stability of powdered ammonium nitrate under different atmosphere conditions. J. Hazard Mater. 337, 10-19.

Zhou, W.Y., 2004. Ammonium phosphate dry powder extinguishant, Fire Techn. Products Infor. 7, 70-76. Zeng, W.R., Li, S.F., Chow, W.K., 2016a. Preliminary Studies on Burning Behavior of Polymethylmethacrylate (PMMA). J. Fire Sci. 20, 297-317. Zeng, W.R., Li, S.F., Chow, W.K., 2016b. Review on Chemical Reactions of Burning Poly(methyl methacrylate) PMMA. J. Fire Sci. 20, 401-433.

IP T

Zhang, X.Y., Yu, J.L., Yan, X., Xie, Q., Gao, W., 2016. Flame propagation behaviors of nano- and micro-scale PMMA dust explosions. J. Loss Prev. Process Ind. 40, 101-111.

SC R

Zhang, X.Y., Yu, J.L., Gao, W., Zhang, D., Sun, J., Guo, S., Dobashi, R., 2017. Effects of particle size distributions on PMMA dust flame propagation behaviors. Powder Technol. 317, 197-208.

A

CC E

PT

ED

M

A

N

U

Zhang, X.Y., 2017. Flame Propagation Behaviors of Nano PMMA Dust Explosions. Ph.D. thesis, Dalian University of Technology, China.

IP T SC R U

A

CC E

PT

ED

M

A

N

Fig. 1. Experimental apparatus. 1-PC, 2-Transformer, 3-High speed camera, 4-Data acquisition, 5-Time controller, 6-Stopper, 7-Metal mesh, 8-Electeodes, 9-Cylinder, 10-Thermocouple, 11-Moveable tube, 12-Solenoid valve, 13-Buffer tank, 14-Air tank

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 2. Particle size distributions of 100nm PMMA particles and ABC powder.

IP T SC R U N

A

CC E

PT

ED

M

A

(a) 100nm PMMA (b) ABC powder (c) PMMA and ABC powder Fig. 3. SEM images of PMMA particles, ABC powder and the mixtures of them.

IP T SC R U N A M ED PT CC E A Fig. 4. Flame configurations of 100 nm PMMA dust explosions suppressed by different proportions of ABC powder.

IP T SC R

U

(b) 450 g/m3 PMMA dust cloud with 10%

N

(a) 450 g/m3 PMMA dust cloud ABC powder

A

CC E

PT

ED

M

A

Fig. 5. Flame microstructures.

IP T

M

A

N

U

SC R

(a)

A

CC E

PT

ED

(b)

(c)

Fig. 6. Flame propagation velocities of 100 nm PMMA dust clouds with mass densities of 250 g/m3 (a), 450 g/m3 (b), 650 g/m3 (c).

IP T

M

A

N

U

SC R

(a)

A

CC E

PT

ED

(b)

(c)

Fig. 7. Flame temperatures of 100 nm PMMA dust clouds with mass densities of 250 g/m³ (a), 450 g/m³ (b) and 650 g/m³ (c).

IP T SC R U N A M ED PT CC E

(a)

(b)

A

Fig 8. Thermal analysis results of ABC powder (a) and 100 nm PMMA particles (b).

IP T SC R

Fig. 9. Weight losses of PMMA particles vs. temperatures with various mass

A

CC E

PT

ED

M

A

N

U

proportions of ABC powder.

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 10. Suppression mechanisms of nano-PMMA dust explosions by ABC powder.

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 11. The mole fraction of H and OH.

IP T SC R U N A M

(a)

(b)

A

CC E

PT

ED

Fig. 12. Maximum rate of production for H (a) and OH (b) with different proportions of ABC powder.

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 13. Rate of production of PO2 due to various reactions.

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 14. Reaction pathway diagram for the key recombination pathways via phosphorus.

Table 1 Characteristic diameters of 100nm PMMA particles and ABC powder. ABC(≤14μm) 13.963 4.395 1.972 6.873 33.700

MP-300(100nm) 24.877 10.486 6.151 22.543 46.764

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Diameters D[4,3](μm) D[3,2](μm) Dv(10) (μm) Dv(50) (μm) Dv(90) (μm)

Table 2

δ1 (%) 43 45 50

δ2 (%) 80 79 82

δ3 (%) 90 93

A

CC E

PT

ED

M

A

N

U

SC R

Density 250 (g/m3) 450 (g/m3) 650 (g/m3)

IP T

Suppression ratio for density of 250 g/m3, 450 g/m3 and 650 g/m3.