Surface heat loss and chemical kinetic response in deflagration-to-detonation transition in microchannels

The effects of cold, hot, and adiabatic walls on flame propagation and deflagration-to-detonation transition (DDT) in a microscale channel are investigated by high-resolution numerical simulation. Results show that the conducting, cold, and hot walls lower the flame acceleration rate, while DDT can occur and originate from local explosion near the flame tip for both the hot and adiabatic walls. Furthermore, for the adiabatic wall, autoignition near the wall produces fast flames in the boundary layer, inducing two shocks propagating and colliding at the center, inducing a local explosion near the flame tip. However, for the hot wall, fast flames do not appear in the boundary layer due to heat loss at the wall; DDT occurs due to coupling of the compression waves with the stretched flame, and needs strong local explosion due to the absence of autoignition in the boundary layer. Nevertheless, compared with the adiabatic wall, occurrence of DDT is delayed while the run-up distance is reduced because of heat input from the hot wall to the fresh gas. For the cold wall, a flame propagates oscillatorily and fails to develop to detonation. It is identified that the flame retreat is caused by thermal contraction due to heat loss at the cold wall. It is further demonstrated that realistic chemistry is needed for an accurate description of the occurrence of autoignition within the boundary layer and DDT.

Figure 1

Ignition delay time as a function of temperature for ${\mathrm{H}}_2\text{−}{\mathrm{O}}_2$ system with the developed San Diego mechanism [40] and Burke's mechanism [44, 45, 46].

Figure 2

Flame-tip position and velocity as a function of time.

Figure 3

Entire process of flame acceleration and DDT: (a) adiabatic wall; (b) hot wall.

Figure 4

Temperature and pressure profiles during DDT: (a) adiabatic wall, $t=23.18$, 37.49, 51.91, 61.50, 69.85, 75.68, 80.84, 86.76, and 89.68 μs; (b) hot wall, $t=15.98$, 30.91, 45.19, 53.29, 61.14, 68.78, 75.88, 83.3, 91.0, 96.52, 101.07, 104.57, 107.27, and 108.37 μs.

Figure 5

Temperature and mass fraction profiles of species (a) at center and (b) near wall at $t=16.01\times {10}^{-6}$ s for the adiabatic wall.

Figure 6

Temperature and mass fraction profiles of species (a) at center and (b) near wall at $t=86.7\times {10}^{-6}\mathrm{s}$ for the adiabatic wall.

Figure 7

Reaction rate of critical species at center for the adiabatic wall: (a) 85.72 μs, (b) 86.61 μs, (c) 88.34 μs, and (d) 89.12 μs.

Figure 8

Temperature and species mass fraction profiles at (a) center and (b) near wall at $t=107.5\mu \mathrm{s}$ for the hot wall.

Figure 9

DDT process (density gradient): (a) adiabatic wall; (b) hot wall.

Figure 10

Histories of maximum pressure in DDT and detonation propagation: (a) adiabatic wall; (b) hot wall.

Figure 11

Histories of maximum pressure at axis of channel in Fig. 10.

Figure 12

Evolution of the flame front: $t=1.5-129.1\mu \mathrm{s}$.

Figure 13

Pressure, $x$-flow velocity, and species mass fraction profiles along the center of Fig. 12: 1–8, $t=94.76$, 97.89, 101.06, 116.94, 120.0, 123.02, 126.03, and 129.05 μs.

Figure 14

Positions of the flame surface at the center and wall, as well as flame stretch versus time for cases with (a) adiabatic wall, (b) hot wall, and (c) cold wall: black (center), red (wall), blue (stretch).

Figure 15

Verification of grid resolution: (a) $x\text{−}t$ plots for three cases; (b) pressure profiles at axis for hot-wall boundary.