1. Centre for Fire and Explosion Studies Numerical Study of Spontaneous Ignition of Pressurized Hydrogen Release through a length of tube with local contraction B.P. Xu and J. X. Wen Centre for Fire and Explosion Studies Faculty of Science, Engineering and Computing Kingston University, London Centre for Fire and Explosion Studies

2. Shock-induced diffusion ignition Leading shock driven into air; Air behind the leading shock is heated; Expansion waves developed into hydrogen; Hydrogen behind the contact surface is cooled; Occurrence of mixing at contact surface – mass and energy exchange via molecular diffusion; Spontaneous ignition takes place after an ignition delay time. Adopted from Dryer et al. (2007)

3. Centre for Fire and Explosion Studies Background Objective: To study the effect of internal geometry on the propensity to spontaneous ignition using a release tube with a local contraction. local contraction Experimental studies: 1.Wolanski & Wojicicki (1973): direct release into confine chamber filled with oxygen. 2.Dryer et al.(2007): release through tubes with different internal geometries into ambient air. 3.Golub et al. (2008): release through tubes with different shapes of cross section into air. 4.Mogi et al. (2008,2009): release through a straight tube into ambient air. 5.Lee et al. (2011): release through a straight tube into ambient air. Numerical Studies: 1.Liu et al. (2006): direct release into ambient air. 2.Golub et al. (2007, 2008): direct release and release through a straight tube into ambient air. 3.Lee et al. (2009): release inside a straight tube filled with ambient air. 4.Yamada et al. (2009, 2011): release through a straight tube into ambient air. 5.Xu & Wen (2008- 2011): direct release and release through a straight tube into ambient air.

4. Centre for Fire and Explosion Studies Numerical Methods  ALE numerical scheme: convective term solved separately from diffusion terms;  In Lagrangian stage, 2 nd -order Crank-Nicolson scheme + 2 nd -order central differencing for diffusion and pressure related terms;  In Eulerian phase, 3 rd -order Runge-Kutta method + 5 th -order upwind WENO scheme for convection terms;  Detailed chemical-kinetic scheme - 8 reactive species and 21 elementary steps –third body and “fall off” behavior considered (Williams 2006);  Multi-component diffusion approach for mixing - thermal diffusion;  Iris model mimics the rupture process;

5. Centre for Fire and Explosion Studies Problem Description ParametersValues Rupture time (µs)5 Release pressure (bar)50, 25, 18 Initial Temperature (K)293 Diameter of tube (mm)3 Length of tube (mm)60 Contraction ratio0.6 Thickness of film(mm)0.1 Minimum grid spacing (µm)15 Figure. 1. Schematic of the release tube with local contraction. Computational details

6. Centre for Fire and Explosion Studies Predicted contours of pressure, Mach number and axial velocity at a time interval of 1 µs starting from 13 µs for the case of 50 bar. Release pressure – 50 bar (a) Logarithm of pressure (bar) (b) Mach number (c) Axial velocity (m/s)

7. Centre for Fire and Explosion Studies Predicted contours of Hydrogen mass fraction, temperature and OH mass fraction at a time interval of 1 µs starting from 13 µs for the case of 50 bar. Release pressure – 50 bar (a) Hydrogen mass fraction (b) Temperature (c) OH mass fraction

8. Centre for Fire and Explosion Studies (c) Temperature (b) Hydrogen mass fraction (a) Axial velocity Release pressure – 50 bar

9. Release pressure – 25 bar (a) Mach number (b) Temperature

10. Centre for Fire and Explosion Studies Scattered plots of temperature versus mixture fraction Scattered plots of OH mass fraction versus mixture fraction

11. Centre for Fire and Explosion Studies Scattered plots of temperature versus mixture fraction Scattered plots of OH mass fraction versus mixture fraction

12. Centre for Fire and Explosion Studies Maximum temperature versus release time

13. Centre for Fire and Explosion Studies Summaries 1.The internal geometry of a local contraction can significantly increase the propsensity to spontaneous ignition by producing elevated flammable mixture and enhancing turbulent mixing from shock formation, reflection and interaction. 2.The first ignition kernel is observed upstream the contraction. It then quickly propagates along the contact interface and transits to a partially premixed flame due to the enhanced turbulent mixing. 3.The partially premixed flames are highly distorted and overlapped with each other. Flame thickening is observed due to the merge of thin flames. 4.The predictions suggested that sustained flames could develop for release pressure as low as 25 bar. For the release pressure of 18 bar, spontaneous ignition was predicted but the flame was soon quenched inside the tube. To some extent this finding is consistent with Dryer’s experimental observation in that the minimum release pressure for the release through a tube with internal geometries is only 20.4bar.