- Email: [email protected]
Self-ignition of hydrogen jet discharged under high pressure into a perforated channel
Accepted Manuscript Self-ignition of hydrogen jet discharged under high pressure into a perforated channel Sergey V. Golovastov, Olga Terekhova PII:
S0950-4230(16)30148-6
DOI:
10.1016/j.jlp.2016.05.027
Reference:
JLPP 3229
To appear in:
Journal of Loss Prevention in the Process Industries
Received Date: 22 March 2016 Revised Date:
20 May 2016
Accepted Date: 22 May 2016
Please cite this article as: Golovastov, S.V., Terekhova, O., Self-ignition of hydrogen jet discharged under high pressure into a perforated channel, Journal of Loss Prevention in the Process Industries (2016), doi: 10.1016/j.jlp.2016.05.027. 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.
ACCEPTED MANUSCRIPT
Self-ignition of hydrogen jet discharged under high pressure into a perforated channel
Sergey V. Golovastov*1 – corresponding author, Olga Terekhova2
125412, Moscow, Russia. Izhorskaya str., 13, build. 2.
2
Bauman Moscow State Technical University,
* [email protected]
M AN U
105005, Moscow, Russia. 2nd Baumanskaya str., 5.
RI PT
Joint Institute for High Temperatures of Russian Academy of Science,
SC
1
AC C
EP
TE D
Tel.: +7-495-4858463; mob.: +7-905-7065351; fax: +7-495-4842138
ACCEPTED MANUSCRIPT
Abstract
The self-ignition of a hydrogen jet during spontaneous release from a high-pressure chamber into a perforated channel was investigated experimentally. The experiments were devoted to the investigation of the possibility of preventing the self-ignition of hydrogen by using the perforated
RI PT
channel. Two lateral orifices with diameters of 2–6 mm were located in a side surface of a rectangular cross-section channel. The length of the channel was 180 mm and its sides were 2 and 10 mm. The initial hydrogen pressure varied from 3 to 9 MPa. Maps of the possibility of hydrogen
SC
self-ignition depending on the Mach number of the incident shock wave, the distance between the position of the orifices and the rupture diaphragm, and the orifice’s cross-sectional area are
M AN U
presented. Using the perforated channel as a pressure-relief device can be effective under certain conditions.
Keywords
1. Introduction
TE D
Hydrogen; perforated channel; self-ignition; high pressure
EP
One of the key problems when using and storing compressed hydrogen is its ability to self-ignite on sudden discharge from a high-pressure cylinder. In some cases, the self-ignition can occur in the
AC C
absence of external ignition sources. Pressure relief devices are used (Sunderland, 2008) to avoid a rupture of the pressure vessel. Specifications applicable to such devices must ensure the safe discharge of a pressurized hydrogen jet leaving out the possibility of self-ignition. The mechanism of such self-ignition of the pulsed hydrogen with the surrounding air was investigated in the papers of Wolanski and Wojcicki (1972), Golub et al. (2008), Xu et al. (2008), Yamada, Kitabayashi, Hayashi, and Tsuboi (2011), and Mironov, Penyazkov, and Ignatenko (2015). The ignition delays for the non-premixed mixtures were presented in the papers of Sakurai (1986), Xu, Wen, Dembele, Tam, and Hawksworth (2009), and Golovastov and Bocharnikov
ACCEPTED MANUSCRIPT
(2012). The minimum values of pressures and temperatures that cause the ignition of hydrogen are presented in the papers of Dryer et al. (2007), Mogi, Kim, Shiina, and Horiguchi (2008), Lee and Jeung (2009), and Grune, Sempert, Kuznetsov, and Jordan (2014). One of the ways to prevent the self-ignition of hydrogen during the discharge into a channel
RI PT
filled with air can be by constructing the lateral orifices in the channel side. These orifices can be placing to attenuate the intensity of the shock wave. However, the orifices on the sides of the channel means that an obstacle is present in the supersonic flow of hydrogen or compressed air.
SC
This can lead to an additional increase in pressure and temperature behind the incident shock wave. The influence of obstacles and variable cross-section on the self-ignition of hydrogen is presented in
M AN U
the papers of Mogi, Kim, Shiina, and Horiguchi (2008) and Lee and Jeung (2009). The actions of obstacles, channel cross-section, and channel length are presented by Baev, Buzukov, and Shumskii (2000) and Luikov, Mironov, Penyazkov, and Skilandz (2011). Numerical studies of spontaneous ignition were carried out in a tube with local contraction by Xu and Wen (2012), with an obstacle plate by Xu, Wen, and Tam (2011), in a T-channel by Bragin, Makarov, and Molkov (2013), and in
TE D
a long channel by Kitabayashi, Wada, Mogi, Saburi, and Hayashi (2013). Kaneko, Hayashi, and Ishii (2015) showed that the ignition centres are located on the inner surface of the channel. Thus,
EP
the presence of orifices can have a significant influence on the dynamics of hydrogen ignition. The aim of the present work was to experimentally investigate the influence of the orifice
AC C
parameters on the self-ignition of a hydrogen jet during the pulse discharge from the high-pressure chamber into the channel. The limit values of the Mach number of the generated shock wave were determined at which the spontaneous self-ignition of hydrogen occurs: 1. depending on the position of two lateral orifices; 2. depending on the area of two lateral orifices placed in a fixed location.
ACCEPTED MANUSCRIPT
2. Experimental set-up
The scheme of the experimental set-up is presented in Figure 1. Compressed hydrogen from the balloon (1) was supplied into the chamber (2) by a manual adjustment with the use of a regulating valve (3). Pressure in the chamber grew at a speed of 0.2 MPa/s and was measured by a precise
RI PT
manometer (4). When the required pressure was reached, a diaphragm (5) between the chamber and channel was broken, and hydrogen was discharged into the perforated channel (6). The air pressure
TE D
M AN U
SC
in the open channel was 0.1 MPa.
Figure 1.
EP
The length of the high-pressure chamber was 400 mm, which made it possible to avoid the influence of rarefaction waves on the characteristics of the hydrogen jet and the generated incident
AC C
shock wave in the 180-mm-long channel. For example, for minimum value of the Mach number 3.3 (see below, Figure 6) the velocity of the contact surface between air and hydrogen can be estimated by equation (1): u=
2 a air 1 m M − = 856 , γ +1 M s
(1)
where aair – sound speed in air for temperature 293 K, M – Mach number, γ = 1.4 – heat capacity ratio. The maximum period of moving of the contact surface through the channel is equaled to 180 mm = 210 µs . It suffices to note that this value does not exceed even the period of moving of 856 m s
ACCEPTED MANUSCRIPT
the rarefaction wave from the diaphragm to the closed end of the high pressure chamber 400 mm 400 mm = = 307 µs , where a H 2 – sound speed in hydrogen. The real delay of aH2 1305 m s interaction of the rarefaction wave with the contact surface will be significantly higher – see, for example, book of Gaydon and Hurle (1963).
RI PT
The internal diameter of the high-pressure chamber was 8–10 mm, and the diameter of the rupture diaphragm was 5 mm. The scheme of connection of the chamber with channel is shown in
be found in the paper of Golub et al. (2008).
SC
Figure 1. Detailed photographs of the rectangular cross-section channel and its main elements can
A channel with a rectangular cross-section with sides of 10 and 2 mm was connected to the
M AN U
high-pressure chamber. Thus, the cross section area of the rectangular channel was S0 = 20.0 mm2. The orifices (7) were located on the surface of the 10-mm side. Diameter of two symmetrically located orifices varied: 2 mm, 3 mm, 4 mm, and 6 mm. The full area of two orifices S* was respectively: 6.3 mm2, 14.1 mm2, 25.1 mm2, and 56.5 mm2. Thus, the channel of the rectangular
TE D
cross-section provides the opportunity to vary the area ratio σ = S*/S0 of two symmetrically located orifices and channels in a wide range σ = 0.32–2.83. Technical hydrogen, contained in a cylinder with a volume of 40 l at a pressure of 15 MPa,
EP
was used. Air was blown through the channel and the room was ventilated before each experiment.
AC C
The rupture of the copper diaphragm occurred at the moment when the pressure difference reached a certain value, which depended on the thickness of the diaphragm (5) and the depth of the cuts. The extent of the diaphragm rupture was controlled. Only experimental data obtained after the full opening of the diaphragms were used. For determination of the Mach number of the incident shock wave, two piezoelectric pressure transducers PCB 113A24 (8) were used, mounted along the channel. PD-256 photodiodes (9), mounted were used to register the flame. Figure 2 shows typical readings from the pressure transducers and photodiodes, when orifices are located at a distance of 40 mm.
RI PT
ACCEPTED MANUSCRIPT
Figure 2.
SC
In addition to the photodiodes, a digital video camera with a recording speed of 30 fps and resolution of 640*480 was used (Figure 1, pos. 10). Figure 3 shows a photograph of the flames
M AN U
during the self-ignition of the hydrogen jet discharged into the channel with two lateral orifices. The orifices were located at a distance of 40 mm (a) or 90 mm (b) from the diaphragm. The photograph shows the single flame flash along the axis of the channel for the distance of 40 mm and two lateral flame flashes for the distance of 90 mm. Location of the orifices at a distance of 140 mm gives the
AC C
EP
TE D
results like in Figure 3b (90 mm).
Figure 3.
ACCEPTED MANUSCRIPT
3. Experimental results. Discussion
Figure 4 shows a map of the self-ignition data in dependence on the Mach number of the incident shock wave and the distance between the lateral orifices and the rupture diaphragm. Two orifices, 4 mm in diameter, were placed at distances X of 40, 90, and 140 mm. Figure 4 also shows the critical
RI PT
values of the initial pressure of compressed hydrogen for self-ignition in dependence on the distance
M AN U
SC
X.
TE D
Figure 4.
When the distance from the diaphragm to the orifices was 40 mm, the self-ignition was
EP
observed at a shock-wave Mach number of 3.63. However, at the Mach number of 3.62, selfignition was not observed. With the increase in the distance between the orifices and the diaphragm,
AC C
the limiting Mach number increased monotonically. For a distance of 90 mm, the critical value was 3.62–3.69, and for the distance of 140 mm, the critical value was 3.72–3.80. With the increase of distance X the critical value of the initial hydrogen pressure rose from 4.6 to 5.3 MPa. Considering the position of the lateral orifices in which the self-ignition is registered as the location of ignition, it is possible to plot the dependence of the critical Mach number on the distance X (Figure 4, curve 3). The obtained dependence is different from the monotonically decreasing
dependence of the critical Mach number for a smooth channel (curve 5), which is presented in the paper of Golub et al. (2008). Using the values of the shock-wave pressure from the paper of Golub
ACCEPTED MANUSCRIPT
et al. (2008), it is possible to estimate the Mach number of the shock wave and to plot the dependence on the distance X for the smooth channel of rectangular cross-section. Location of the lateral orifices at the distance 40 mm makes the self-ignition of hydrogen possible at lower Mach numbers and hence at a lower initial hydrogen pressure. The Mach number
RI PT
is reduced from 3.81 (smooth) to 3.63 (perforated). A possible reason for the reduction of the critical Mach number could be a reflection of the incident shock wave from the orifice’s edges, as shown schematically on Figure 5. The presence of the reflected shock wave (3) leads to an increase
EP
Figure 5.
TE D
M AN U
ignition occurs mainly on the inner surface of the channel.
SC
in temperature of the contact surface (4) of the hydrogen with air. Moreover, the centre of self-
AC C
The outflow of the compressed gas (air and hydrogen) through the lateral orifices of the channel can reduce the density, pressure, and temperature behind the incident shock wave. When the shock wave moves along the channel, the distance L between the incident shock wave and the ideal contact surface increases. The distance L can be evaluated by the gas-dynamic ratio: L = x SW − xCS = Ma air t −
2 a air 1 M − t . γ +1 M
(2)
Here, xSW – position of the shock wave, xCS – position of the contact surface, aair – sound speed in air, M – Mach number, γ = 1.4 – heat capacity ratio. Assuming that t = xSW Maair , it is possible to
ACCEPTED MANUSCRIPT
plot the length L in dependence on the position of the incident shock wave in the channel (Figure 5). For the first location of the orifices (40 mm), the length was 8–9 mm. This length corresponds to twice the diameter of the orifices (4 mm). This distance can also be close to the total path traversed by the incident (4 mm) and reflected (4 mm) shock wave; that is, the reflected shock wave interacts
RI PT
with the contact surface when it passes through the orifice. For large values of X, the action of the generated reflected shock wave on the contact surface can be neglected. Thus, for the self-ignition to occur at a long distance X, it is necessary to increase the intensity of the shock wave.
SC
Figure 6 shows the map of self-ignition data in dependence on the Mach number of the incident shock wave and the cross-sectional area of the lateral orifices. Two orifices were located at
M AN U
a fixed distance 40 mm from the rupture diaphragm. The total cross-sectional area σ varied in the range of 0.32–2.83. Figure 6 also shows the critical values of the initial pressure of compressed
Figure 6.
AC C
EP
TE D
hydrogen for the self-ignition in dependence on the area σ.
Figure 6 shows that for the orifices exceeding 1.27σ there is a significant increase in the critical Mach number, which leads to self-ignition. As expected, the increase in the cross-sectional area leads to an outflow of compressed gas from the lateral orifices and to weakening of the intensity of the shock wave. Contrariwise, orifices with areas of less than 0.72σ can create disturbances in the flow behind the incident shock wave. Perturbations of the flow near the orifices lead to a local intermixing of hydrogen with air and a local rise in temperature.
ACCEPTED MANUSCRIPT
The minimal value of the critical Mach number was equal to 3.6 for orifices of 0.72σ–1.27σ. These cross-sectional areas correspond to orifice diameters of 3 and 4 mm. The distance between the incident shock wave and the contact surface is close to twice the diameter of the orifice. Therefore, the presence of the minimum on Figure 6 for the range of 0.72σ–1.27σ can be caused by
RI PT
the interaction of the reflected shock wave with the contact surface when it passes through the orifice.
Thus, it is possible to assume that the discharge of the hydrogen jet into a geometrically non-
SC
uniform space is accompanied by an increased risk of self-ignition of hydrogen. The greatest effect of the obstacles/orifices on self-ignition occurs when they are located on the inner surface of the
M AN U
channel. The use of the perforated channel as a pressure relief device can be effective in comparison with the smooth channels only when the cross-sectional area of the orifices is above 1.2σ. Among the data, presented in the paper, the smallest Mach number leading to the self-ignition of hydrogen was observed to be equal to 3.63. This value corresponded to an initial pressure of 4.6 MPa, which is three times the value for the T-shaped channel. It was shown by Golovastov,
TE D
Baklanov, Golub, Lenkevich, and Volodin (2010) that self-ignition in the T-shaped channel occurs even at an initial hydrogen pressure of 1.35 MPa. Thus, the use of the perforated channel can be
EP
effective in comparison with T-shaped or other angular valves for initial pressures not exceeding ~5
AC C
MPa.
4. Conclusions
On the basis of the obtained results the following conclusions can be drawn. The dependence of the critical values of the Mach number, resulting in the self-ignition of hydrogen, on the distance between the orifices and the rupture diaphragm was presented. Increasing the distance from 40 mm to 90 mm increases the critical Mach number from 3.63 to 3.80 and initial pressure from 4.6 to 5.3 MPa. The dependence of the critical value of the Mach number on the cross-sectional area of orifices was presented too. The dependence has a non-monotonic character: for orifices larger than
ACCEPTED MANUSCRIPT
1.2σ or smaller than 0.4σ, the critical Mach number leading to self-ignition increases. Using the perforated channel as a pressure-relief device can be effective in comparison with smooth channels or T-shaped channels when the degree of perforation is 1.2σ or above and the initial hydrogen
RI PT
pressure does not exceed ~5 MPa.
5 Acknowledgements
This work was supported by the Russian Foundation for Basic Research under grant no. 15-38-
SC
70017, and grant of President of the Russian Federation no. SP-1501.2016.1. Authors are grateful to Prof. Victor Golub for useful discussions, to Ivan Tarasenko and Dmitry Lenkevich for help with
M AN U
the experiments.
REFERENCES
Baev, V. K., Buzukov, A. A., & Shumskii, V. V. (2000). Conditions of self-ignition at
Explosion, 36(3), 3–9.
TE D
impulse high pressure injection of combustible gases in confined volume. Combustion and
Bragin, M. V., Makarov, D. V., & Molkov, V. V. (2013). Pressure limit of hydrogen
8039-8052.
EP
spontaneous ignition in a T-shaped channel. International Journal of Hydrogen Energy, 38(19),
AC C
Dryer, F. L., Chaos, M., Zhao, Z., Stein, J. N., Alpert, J. Y., & Homer, C. J. (2007). Spontaneous ignition of pressurized releases of hydrogen and natural gas into air. Combustion science and technology, 179(4), 663-694.
Gaydon, A. G., & Hurle, I. R. (1963). The shock tube in high-temperature chemical physics. Chapman and Hall. Golovastov, S. V., Baklanov, D. I., Golub, V. V., Lenkevich, D.A., & Volodin, V. V. (2010). Influence of Lateral Orifices on Self-Ignition of Hydrogen at a Pulse Discharge Into Channel. 8th ISHPMIE September 5-10, 2010, Yokohama, Japan. No.ISH-112.
ACCEPTED MANUSCRIPT
Golovastov, S., & Bocharnikov, V. (2012). The influence of diaphragm rupture rate on spontaneous self-ignition of pressurized hydrogen: Experimental investigation. international journal of hydrogen energy, 37(14), 10956-10962.
Golub, V. V., Baklanov, D. I., Golovastov, S. V., Ivanov, M. F., Laskin, I. N., Saveliev, A. S.,
RI PT
.Semin, N. V., & Volodin, V. V. (2008). Mechanisms of high-pressure hydrogen gas self-ignition in tubes. Journal of Loss Prevention in the Process Industries, 21(2), 185-198.
Grune, J., Sempert, K., Kuznetsov, M., & Jordan, T. (2014). Experimental study of ignited
SC
unsteady hydrogen releases from a high pressure reservoir. international journal of hydrogen energy, 39(11), 6176-6183.
M AN U
Kaneko, W., Hayashi, K., & Ishii, K. (2015). Self-Ignition of High-Pressure Hydrogen Released by Reproducible Rupture of Diaphragm. 25th ICDERS August 2 – 7, 2015 Leeds, UK. Paper 0155.
Kitabayashi, N., Wada, Y., Mogi, T., Saburi, T., & Hayashi, A. K. (2013). Experimental study on high pressure hydrogen jets coming out of tubes of 0.1–4.2 m in length. International Journal of
TE D
Hydrogen Energy, 38(19), 8100-8107.
Lee, B. J., & Jeung, I. S. (2009). Numerical study of spontaneous ignition of pressurized
EP
hydrogen released by the failure of a rupture disk into a tube. international journal of hydrogen energy, 34(20), 8763-8769.
AC C
Luikov, D. G., Mironov, V, N., Penyazkov, O. G., & Skilandz A.V. (2011). Self-ignition of a high pressure hydrogen jet outflowing into the encumbered space. In: Proc 2nd Minsk Int Colloq Phys Shock Waves Comb Deton, Minsk; Paper 60.
Mironov, V. N., Penyazkov, O. G., & Ignatenko, D. G. (2015). Self-ignition and explosion of a 13-MPa pressurized unsteady hydrogen jet under atmospheric conditions. International Journal of Hydrogen Energy, 40(16), 5749-5762.
ACCEPTED MANUSCRIPT
Mogi, T., Kim, D., Shiina, H., & Horiguchi, S. (2008). Self-ignition and explosion during discharge of high-pressure hydrogen. Journal of Loss Prevention in the Process Industries, 21(2), 199-204. Sakurai, A. (1986). Auto-ignition of hydrogen by a shock-compressed oxidizer. In Shock
RI PT
Waves and Shock Tubes: Proceedings of the Fifteenth International Symposium on Shock Waves and Shock Tubes, Berkeley, California, July 28-August 2, 1985 (p. 77). Stanford University Press.
Sunderland, P. B. (2008). Pressure relief devices for hydrogen vehicles. Third European
SC
Summer School on Hydrogen Safety.
Wolanski, P., & Wojcicki, S. (1972). Investigation into the mechanism of the diffusion
M AN U
ignition of a combustible gas flowing into an oxidizing atmosphere. In Proc. Combust. Instit (Vol. 14, pp. 1217-1223).
Xu, B. P., Wen, J. X., Dembele, S., Tam, V. H. Y., & Hawksworth, S. J. (2009). The effect of pressure boundary rupture rate on spontaneous ignition of pressurized hydrogen release. Journal of Loss Prevention in the Process Industries, 22(3), 279-287.
TE D
Xu, B. P., El Hima, L., Wen, J. X., Dembele, S., Tam, V. H. Y., & Donchev, T. (2008). Numerical study on the spontaneous ignition of pressurized hydrogen release through a tube into
EP
air. Journal of Loss Prevention in the Process Industries, 21(2), 205-213. Xu, B. P., & Wen, J. X. (2012). Numerical study of spontaneous ignition in pressurized
AC C
hydrogen release through a length of tube with local contraction. International Journal of Hydrogen Energy, 37(22), 17571-17579.
Xu, B. P., Wen, J. X., & Tam, V. H. Y. (2011). The effect of an obstacle plate on the spontaneous ignition in pressurized hydrogen release: A numerical study. International journal of hydrogen energy, 36(3), 2637-2644.
Yamada, E., Kitabayashi, N., Hayashi, A. K., & Tsuboi, N. (2011). Mechanism of highpressure hydrogen auto-ignition when spouting into air. International journal of hydrogen energy, 36(3), 2560-2566.
ACCEPTED MANUSCRIPT
FIGURE CAPTIONS
Figure 1. Scheme of the experimental set-up (a), drawing of the diaphragm mounting (b) and photo of the connection (c). 1: balloon; 2: high-pressure chamber; 3: regulating valve; 4: manometer; 5: rupture diaphragm; 6: perforated channel; 7: orifices; 8: piezoelectric pressure transducers; 9:
RI PT
photodiodes; 10, videocamera; 11, locking nut; 12, coupling.
Figure 2. Readings from the pressure transducers (P, solid lines) and photodiodes (I, dashed lines,
SC
relative units). Positions 40 mm and 90 mm.
M AN U
Figure 3. Photography of the flame jets of hydrogen during the self-ignition. The distance between orifices and rupture diaphragm is 40 mm (a) and 90 mm (b). Time interval between frames 1/30 s. 1: elements of high-pressure chamber; 2: perforated channel; 3: orifices; 4: piezoelectric pressure transducers; 5: photodiodes.
TE D
Figure 4. Map of self-ignition of hydrogen in dependence on the Mach number M of the incident shock wave and distance X between lateral orifices and rupture diaphragm. 1: self-ignition occurs;
EP
2: self-ignition does not occur; 3: critical Mach number; 4: critical initial pressure; 5: critical Mach
AC C
number for a smooth channel, Golub et al. (2008).
Figure 5. Evaluated distance L between incident shock wave (M = 3.6÷4.0) and contact surface in dependence on the position of the incident shock wave. 1: orifice; 2: incident shock wave; 3: reflected shock wave; 4: contact surface.
Figure 6. Map of self-ignition of hydrogen in dependence on the Mach number M of the incident shock wave and cross-sectional area σ of two lateral orifices. 1: self-ignition occurs; 2: self-ignition does not occur; 3: critical Mach number; 4: critical initial pressure.
S0950-4230(16)30148-6
DOI:
10.1016/j.jlp.2016.05.027
Reference:
JLPP 3229
To appear in:
Journal of Loss Prevention in the Process Industries
Received Date: 22 March 2016 Revised Date:
20 May 2016
Accepted Date: 22 May 2016
Please cite this article as: Golovastov, S.V., Terekhova, O., Self-ignition of hydrogen jet discharged under high pressure into a perforated channel, Journal of Loss Prevention in the Process Industries (2016), doi: 10.1016/j.jlp.2016.05.027. 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.
ACCEPTED MANUSCRIPT
Self-ignition of hydrogen jet discharged under high pressure into a perforated channel
Sergey V. Golovastov*1 – corresponding author, Olga Terekhova2
125412, Moscow, Russia. Izhorskaya str., 13, build. 2.
2
Bauman Moscow State Technical University,
* [email protected]
M AN U
105005, Moscow, Russia. 2nd Baumanskaya str., 5.
RI PT
Joint Institute for High Temperatures of Russian Academy of Science,
SC
1
AC C
EP
TE D
Tel.: +7-495-4858463; mob.: +7-905-7065351; fax: +7-495-4842138
ACCEPTED MANUSCRIPT
Abstract
The self-ignition of a hydrogen jet during spontaneous release from a high-pressure chamber into a perforated channel was investigated experimentally. The experiments were devoted to the investigation of the possibility of preventing the self-ignition of hydrogen by using the perforated
RI PT
channel. Two lateral orifices with diameters of 2–6 mm were located in a side surface of a rectangular cross-section channel. The length of the channel was 180 mm and its sides were 2 and 10 mm. The initial hydrogen pressure varied from 3 to 9 MPa. Maps of the possibility of hydrogen
SC
self-ignition depending on the Mach number of the incident shock wave, the distance between the position of the orifices and the rupture diaphragm, and the orifice’s cross-sectional area are
M AN U
presented. Using the perforated channel as a pressure-relief device can be effective under certain conditions.
Keywords
1. Introduction
TE D
Hydrogen; perforated channel; self-ignition; high pressure
EP
One of the key problems when using and storing compressed hydrogen is its ability to self-ignite on sudden discharge from a high-pressure cylinder. In some cases, the self-ignition can occur in the
AC C
absence of external ignition sources. Pressure relief devices are used (Sunderland, 2008) to avoid a rupture of the pressure vessel. Specifications applicable to such devices must ensure the safe discharge of a pressurized hydrogen jet leaving out the possibility of self-ignition. The mechanism of such self-ignition of the pulsed hydrogen with the surrounding air was investigated in the papers of Wolanski and Wojcicki (1972), Golub et al. (2008), Xu et al. (2008), Yamada, Kitabayashi, Hayashi, and Tsuboi (2011), and Mironov, Penyazkov, and Ignatenko (2015). The ignition delays for the non-premixed mixtures were presented in the papers of Sakurai (1986), Xu, Wen, Dembele, Tam, and Hawksworth (2009), and Golovastov and Bocharnikov
ACCEPTED MANUSCRIPT
(2012). The minimum values of pressures and temperatures that cause the ignition of hydrogen are presented in the papers of Dryer et al. (2007), Mogi, Kim, Shiina, and Horiguchi (2008), Lee and Jeung (2009), and Grune, Sempert, Kuznetsov, and Jordan (2014). One of the ways to prevent the self-ignition of hydrogen during the discharge into a channel
RI PT
filled with air can be by constructing the lateral orifices in the channel side. These orifices can be placing to attenuate the intensity of the shock wave. However, the orifices on the sides of the channel means that an obstacle is present in the supersonic flow of hydrogen or compressed air.
SC
This can lead to an additional increase in pressure and temperature behind the incident shock wave. The influence of obstacles and variable cross-section on the self-ignition of hydrogen is presented in
M AN U
the papers of Mogi, Kim, Shiina, and Horiguchi (2008) and Lee and Jeung (2009). The actions of obstacles, channel cross-section, and channel length are presented by Baev, Buzukov, and Shumskii (2000) and Luikov, Mironov, Penyazkov, and Skilandz (2011). Numerical studies of spontaneous ignition were carried out in a tube with local contraction by Xu and Wen (2012), with an obstacle plate by Xu, Wen, and Tam (2011), in a T-channel by Bragin, Makarov, and Molkov (2013), and in
TE D
a long channel by Kitabayashi, Wada, Mogi, Saburi, and Hayashi (2013). Kaneko, Hayashi, and Ishii (2015) showed that the ignition centres are located on the inner surface of the channel. Thus,
EP
the presence of orifices can have a significant influence on the dynamics of hydrogen ignition. The aim of the present work was to experimentally investigate the influence of the orifice
AC C
parameters on the self-ignition of a hydrogen jet during the pulse discharge from the high-pressure chamber into the channel. The limit values of the Mach number of the generated shock wave were determined at which the spontaneous self-ignition of hydrogen occurs: 1. depending on the position of two lateral orifices; 2. depending on the area of two lateral orifices placed in a fixed location.
ACCEPTED MANUSCRIPT
2. Experimental set-up
The scheme of the experimental set-up is presented in Figure 1. Compressed hydrogen from the balloon (1) was supplied into the chamber (2) by a manual adjustment with the use of a regulating valve (3). Pressure in the chamber grew at a speed of 0.2 MPa/s and was measured by a precise
RI PT
manometer (4). When the required pressure was reached, a diaphragm (5) between the chamber and channel was broken, and hydrogen was discharged into the perforated channel (6). The air pressure
TE D
M AN U
SC
in the open channel was 0.1 MPa.
Figure 1.
EP
The length of the high-pressure chamber was 400 mm, which made it possible to avoid the influence of rarefaction waves on the characteristics of the hydrogen jet and the generated incident
AC C
shock wave in the 180-mm-long channel. For example, for minimum value of the Mach number 3.3 (see below, Figure 6) the velocity of the contact surface between air and hydrogen can be estimated by equation (1): u=
2 a air 1 m M − = 856 , γ +1 M s
(1)
where aair – sound speed in air for temperature 293 K, M – Mach number, γ = 1.4 – heat capacity ratio. The maximum period of moving of the contact surface through the channel is equaled to 180 mm = 210 µs . It suffices to note that this value does not exceed even the period of moving of 856 m s
ACCEPTED MANUSCRIPT
the rarefaction wave from the diaphragm to the closed end of the high pressure chamber 400 mm 400 mm = = 307 µs , where a H 2 – sound speed in hydrogen. The real delay of aH2 1305 m s interaction of the rarefaction wave with the contact surface will be significantly higher – see, for example, book of Gaydon and Hurle (1963).
RI PT
The internal diameter of the high-pressure chamber was 8–10 mm, and the diameter of the rupture diaphragm was 5 mm. The scheme of connection of the chamber with channel is shown in
be found in the paper of Golub et al. (2008).
SC
Figure 1. Detailed photographs of the rectangular cross-section channel and its main elements can
A channel with a rectangular cross-section with sides of 10 and 2 mm was connected to the
M AN U
high-pressure chamber. Thus, the cross section area of the rectangular channel was S0 = 20.0 mm2. The orifices (7) were located on the surface of the 10-mm side. Diameter of two symmetrically located orifices varied: 2 mm, 3 mm, 4 mm, and 6 mm. The full area of two orifices S* was respectively: 6.3 mm2, 14.1 mm2, 25.1 mm2, and 56.5 mm2. Thus, the channel of the rectangular
TE D
cross-section provides the opportunity to vary the area ratio σ = S*/S0 of two symmetrically located orifices and channels in a wide range σ = 0.32–2.83. Technical hydrogen, contained in a cylinder with a volume of 40 l at a pressure of 15 MPa,
EP
was used. Air was blown through the channel and the room was ventilated before each experiment.
AC C
The rupture of the copper diaphragm occurred at the moment when the pressure difference reached a certain value, which depended on the thickness of the diaphragm (5) and the depth of the cuts. The extent of the diaphragm rupture was controlled. Only experimental data obtained after the full opening of the diaphragms were used. For determination of the Mach number of the incident shock wave, two piezoelectric pressure transducers PCB 113A24 (8) were used, mounted along the channel. PD-256 photodiodes (9), mounted were used to register the flame. Figure 2 shows typical readings from the pressure transducers and photodiodes, when orifices are located at a distance of 40 mm.
RI PT
ACCEPTED MANUSCRIPT
Figure 2.
SC
In addition to the photodiodes, a digital video camera with a recording speed of 30 fps and resolution of 640*480 was used (Figure 1, pos. 10). Figure 3 shows a photograph of the flames
M AN U
during the self-ignition of the hydrogen jet discharged into the channel with two lateral orifices. The orifices were located at a distance of 40 mm (a) or 90 mm (b) from the diaphragm. The photograph shows the single flame flash along the axis of the channel for the distance of 40 mm and two lateral flame flashes for the distance of 90 mm. Location of the orifices at a distance of 140 mm gives the
AC C
EP
TE D
results like in Figure 3b (90 mm).
Figure 3.
ACCEPTED MANUSCRIPT
3. Experimental results. Discussion
Figure 4 shows a map of the self-ignition data in dependence on the Mach number of the incident shock wave and the distance between the lateral orifices and the rupture diaphragm. Two orifices, 4 mm in diameter, were placed at distances X of 40, 90, and 140 mm. Figure 4 also shows the critical
RI PT
values of the initial pressure of compressed hydrogen for self-ignition in dependence on the distance
M AN U
SC
X.
TE D
Figure 4.
When the distance from the diaphragm to the orifices was 40 mm, the self-ignition was
EP
observed at a shock-wave Mach number of 3.63. However, at the Mach number of 3.62, selfignition was not observed. With the increase in the distance between the orifices and the diaphragm,
AC C
the limiting Mach number increased monotonically. For a distance of 90 mm, the critical value was 3.62–3.69, and for the distance of 140 mm, the critical value was 3.72–3.80. With the increase of distance X the critical value of the initial hydrogen pressure rose from 4.6 to 5.3 MPa. Considering the position of the lateral orifices in which the self-ignition is registered as the location of ignition, it is possible to plot the dependence of the critical Mach number on the distance X (Figure 4, curve 3). The obtained dependence is different from the monotonically decreasing
dependence of the critical Mach number for a smooth channel (curve 5), which is presented in the paper of Golub et al. (2008). Using the values of the shock-wave pressure from the paper of Golub
ACCEPTED MANUSCRIPT
et al. (2008), it is possible to estimate the Mach number of the shock wave and to plot the dependence on the distance X for the smooth channel of rectangular cross-section. Location of the lateral orifices at the distance 40 mm makes the self-ignition of hydrogen possible at lower Mach numbers and hence at a lower initial hydrogen pressure. The Mach number
RI PT
is reduced from 3.81 (smooth) to 3.63 (perforated). A possible reason for the reduction of the critical Mach number could be a reflection of the incident shock wave from the orifice’s edges, as shown schematically on Figure 5. The presence of the reflected shock wave (3) leads to an increase
EP
Figure 5.
TE D
M AN U
ignition occurs mainly on the inner surface of the channel.
SC
in temperature of the contact surface (4) of the hydrogen with air. Moreover, the centre of self-
AC C
The outflow of the compressed gas (air and hydrogen) through the lateral orifices of the channel can reduce the density, pressure, and temperature behind the incident shock wave. When the shock wave moves along the channel, the distance L between the incident shock wave and the ideal contact surface increases. The distance L can be evaluated by the gas-dynamic ratio: L = x SW − xCS = Ma air t −
2 a air 1 M − t . γ +1 M
(2)
Here, xSW – position of the shock wave, xCS – position of the contact surface, aair – sound speed in air, M – Mach number, γ = 1.4 – heat capacity ratio. Assuming that t = xSW Maair , it is possible to
ACCEPTED MANUSCRIPT
plot the length L in dependence on the position of the incident shock wave in the channel (Figure 5). For the first location of the orifices (40 mm), the length was 8–9 mm. This length corresponds to twice the diameter of the orifices (4 mm). This distance can also be close to the total path traversed by the incident (4 mm) and reflected (4 mm) shock wave; that is, the reflected shock wave interacts
RI PT
with the contact surface when it passes through the orifice. For large values of X, the action of the generated reflected shock wave on the contact surface can be neglected. Thus, for the self-ignition to occur at a long distance X, it is necessary to increase the intensity of the shock wave.
SC
Figure 6 shows the map of self-ignition data in dependence on the Mach number of the incident shock wave and the cross-sectional area of the lateral orifices. Two orifices were located at
M AN U
a fixed distance 40 mm from the rupture diaphragm. The total cross-sectional area σ varied in the range of 0.32–2.83. Figure 6 also shows the critical values of the initial pressure of compressed
Figure 6.
AC C
EP
TE D
hydrogen for the self-ignition in dependence on the area σ.
Figure 6 shows that for the orifices exceeding 1.27σ there is a significant increase in the critical Mach number, which leads to self-ignition. As expected, the increase in the cross-sectional area leads to an outflow of compressed gas from the lateral orifices and to weakening of the intensity of the shock wave. Contrariwise, orifices with areas of less than 0.72σ can create disturbances in the flow behind the incident shock wave. Perturbations of the flow near the orifices lead to a local intermixing of hydrogen with air and a local rise in temperature.
ACCEPTED MANUSCRIPT
The minimal value of the critical Mach number was equal to 3.6 for orifices of 0.72σ–1.27σ. These cross-sectional areas correspond to orifice diameters of 3 and 4 mm. The distance between the incident shock wave and the contact surface is close to twice the diameter of the orifice. Therefore, the presence of the minimum on Figure 6 for the range of 0.72σ–1.27σ can be caused by
RI PT
the interaction of the reflected shock wave with the contact surface when it passes through the orifice.
Thus, it is possible to assume that the discharge of the hydrogen jet into a geometrically non-
SC
uniform space is accompanied by an increased risk of self-ignition of hydrogen. The greatest effect of the obstacles/orifices on self-ignition occurs when they are located on the inner surface of the
M AN U
channel. The use of the perforated channel as a pressure relief device can be effective in comparison with the smooth channels only when the cross-sectional area of the orifices is above 1.2σ. Among the data, presented in the paper, the smallest Mach number leading to the self-ignition of hydrogen was observed to be equal to 3.63. This value corresponded to an initial pressure of 4.6 MPa, which is three times the value for the T-shaped channel. It was shown by Golovastov,
TE D
Baklanov, Golub, Lenkevich, and Volodin (2010) that self-ignition in the T-shaped channel occurs even at an initial hydrogen pressure of 1.35 MPa. Thus, the use of the perforated channel can be
EP
effective in comparison with T-shaped or other angular valves for initial pressures not exceeding ~5
AC C
MPa.
4. Conclusions
On the basis of the obtained results the following conclusions can be drawn. The dependence of the critical values of the Mach number, resulting in the self-ignition of hydrogen, on the distance between the orifices and the rupture diaphragm was presented. Increasing the distance from 40 mm to 90 mm increases the critical Mach number from 3.63 to 3.80 and initial pressure from 4.6 to 5.3 MPa. The dependence of the critical value of the Mach number on the cross-sectional area of orifices was presented too. The dependence has a non-monotonic character: for orifices larger than
ACCEPTED MANUSCRIPT
1.2σ or smaller than 0.4σ, the critical Mach number leading to self-ignition increases. Using the perforated channel as a pressure-relief device can be effective in comparison with smooth channels or T-shaped channels when the degree of perforation is 1.2σ or above and the initial hydrogen
RI PT
pressure does not exceed ~5 MPa.
5 Acknowledgements
This work was supported by the Russian Foundation for Basic Research under grant no. 15-38-
SC
70017, and grant of President of the Russian Federation no. SP-1501.2016.1. Authors are grateful to Prof. Victor Golub for useful discussions, to Ivan Tarasenko and Dmitry Lenkevich for help with
M AN U
the experiments.
REFERENCES
Baev, V. K., Buzukov, A. A., & Shumskii, V. V. (2000). Conditions of self-ignition at
Explosion, 36(3), 3–9.
TE D
impulse high pressure injection of combustible gases in confined volume. Combustion and
Bragin, M. V., Makarov, D. V., & Molkov, V. V. (2013). Pressure limit of hydrogen
8039-8052.
EP
spontaneous ignition in a T-shaped channel. International Journal of Hydrogen Energy, 38(19),
AC C
Dryer, F. L., Chaos, M., Zhao, Z., Stein, J. N., Alpert, J. Y., & Homer, C. J. (2007). Spontaneous ignition of pressurized releases of hydrogen and natural gas into air. Combustion science and technology, 179(4), 663-694.
Gaydon, A. G., & Hurle, I. R. (1963). The shock tube in high-temperature chemical physics. Chapman and Hall. Golovastov, S. V., Baklanov, D. I., Golub, V. V., Lenkevich, D.A., & Volodin, V. V. (2010). Influence of Lateral Orifices on Self-Ignition of Hydrogen at a Pulse Discharge Into Channel. 8th ISHPMIE September 5-10, 2010, Yokohama, Japan. No.ISH-112.
ACCEPTED MANUSCRIPT
Golovastov, S., & Bocharnikov, V. (2012). The influence of diaphragm rupture rate on spontaneous self-ignition of pressurized hydrogen: Experimental investigation. international journal of hydrogen energy, 37(14), 10956-10962.
Golub, V. V., Baklanov, D. I., Golovastov, S. V., Ivanov, M. F., Laskin, I. N., Saveliev, A. S.,
RI PT
.Semin, N. V., & Volodin, V. V. (2008). Mechanisms of high-pressure hydrogen gas self-ignition in tubes. Journal of Loss Prevention in the Process Industries, 21(2), 185-198.
Grune, J., Sempert, K., Kuznetsov, M., & Jordan, T. (2014). Experimental study of ignited
SC
unsteady hydrogen releases from a high pressure reservoir. international journal of hydrogen energy, 39(11), 6176-6183.
M AN U
Kaneko, W., Hayashi, K., & Ishii, K. (2015). Self-Ignition of High-Pressure Hydrogen Released by Reproducible Rupture of Diaphragm. 25th ICDERS August 2 – 7, 2015 Leeds, UK. Paper 0155.
Kitabayashi, N., Wada, Y., Mogi, T., Saburi, T., & Hayashi, A. K. (2013). Experimental study on high pressure hydrogen jets coming out of tubes of 0.1–4.2 m in length. International Journal of
TE D
Hydrogen Energy, 38(19), 8100-8107.
Lee, B. J., & Jeung, I. S. (2009). Numerical study of spontaneous ignition of pressurized
EP
hydrogen released by the failure of a rupture disk into a tube. international journal of hydrogen energy, 34(20), 8763-8769.
AC C
Luikov, D. G., Mironov, V, N., Penyazkov, O. G., & Skilandz A.V. (2011). Self-ignition of a high pressure hydrogen jet outflowing into the encumbered space. In: Proc 2nd Minsk Int Colloq Phys Shock Waves Comb Deton, Minsk; Paper 60.
Mironov, V. N., Penyazkov, O. G., & Ignatenko, D. G. (2015). Self-ignition and explosion of a 13-MPa pressurized unsteady hydrogen jet under atmospheric conditions. International Journal of Hydrogen Energy, 40(16), 5749-5762.
ACCEPTED MANUSCRIPT
Mogi, T., Kim, D., Shiina, H., & Horiguchi, S. (2008). Self-ignition and explosion during discharge of high-pressure hydrogen. Journal of Loss Prevention in the Process Industries, 21(2), 199-204. Sakurai, A. (1986). Auto-ignition of hydrogen by a shock-compressed oxidizer. In Shock
RI PT
Waves and Shock Tubes: Proceedings of the Fifteenth International Symposium on Shock Waves and Shock Tubes, Berkeley, California, July 28-August 2, 1985 (p. 77). Stanford University Press.
Sunderland, P. B. (2008). Pressure relief devices for hydrogen vehicles. Third European
SC
Summer School on Hydrogen Safety.
Wolanski, P., & Wojcicki, S. (1972). Investigation into the mechanism of the diffusion
M AN U
ignition of a combustible gas flowing into an oxidizing atmosphere. In Proc. Combust. Instit (Vol. 14, pp. 1217-1223).
Xu, B. P., Wen, J. X., Dembele, S., Tam, V. H. Y., & Hawksworth, S. J. (2009). The effect of pressure boundary rupture rate on spontaneous ignition of pressurized hydrogen release. Journal of Loss Prevention in the Process Industries, 22(3), 279-287.
TE D
Xu, B. P., El Hima, L., Wen, J. X., Dembele, S., Tam, V. H. Y., & Donchev, T. (2008). Numerical study on the spontaneous ignition of pressurized hydrogen release through a tube into
EP
air. Journal of Loss Prevention in the Process Industries, 21(2), 205-213. Xu, B. P., & Wen, J. X. (2012). Numerical study of spontaneous ignition in pressurized
AC C
hydrogen release through a length of tube with local contraction. International Journal of Hydrogen Energy, 37(22), 17571-17579.
Xu, B. P., Wen, J. X., & Tam, V. H. Y. (2011). The effect of an obstacle plate on the spontaneous ignition in pressurized hydrogen release: A numerical study. International journal of hydrogen energy, 36(3), 2637-2644.
Yamada, E., Kitabayashi, N., Hayashi, A. K., & Tsuboi, N. (2011). Mechanism of highpressure hydrogen auto-ignition when spouting into air. International journal of hydrogen energy, 36(3), 2560-2566.
ACCEPTED MANUSCRIPT
FIGURE CAPTIONS
Figure 1. Scheme of the experimental set-up (a), drawing of the diaphragm mounting (b) and photo of the connection (c). 1: balloon; 2: high-pressure chamber; 3: regulating valve; 4: manometer; 5: rupture diaphragm; 6: perforated channel; 7: orifices; 8: piezoelectric pressure transducers; 9:
RI PT
photodiodes; 10, videocamera; 11, locking nut; 12, coupling.
Figure 2. Readings from the pressure transducers (P, solid lines) and photodiodes (I, dashed lines,
SC
relative units). Positions 40 mm and 90 mm.
M AN U
Figure 3. Photography of the flame jets of hydrogen during the self-ignition. The distance between orifices and rupture diaphragm is 40 mm (a) and 90 mm (b). Time interval between frames 1/30 s. 1: elements of high-pressure chamber; 2: perforated channel; 3: orifices; 4: piezoelectric pressure transducers; 5: photodiodes.
TE D
Figure 4. Map of self-ignition of hydrogen in dependence on the Mach number M of the incident shock wave and distance X between lateral orifices and rupture diaphragm. 1: self-ignition occurs;
EP
2: self-ignition does not occur; 3: critical Mach number; 4: critical initial pressure; 5: critical Mach
AC C
number for a smooth channel, Golub et al. (2008).
Figure 5. Evaluated distance L between incident shock wave (M = 3.6÷4.0) and contact surface in dependence on the position of the incident shock wave. 1: orifice; 2: incident shock wave; 3: reflected shock wave; 4: contact surface.
Figure 6. Map of self-ignition of hydrogen in dependence on the Mach number M of the incident shock wave and cross-sectional area σ of two lateral orifices. 1: self-ignition occurs; 2: self-ignition does not occur; 3: critical Mach number; 4: critical initial pressure.