CC BY
12ISHS 2019 Moscow, Russia
EXPERIMENTAL 2D MODEL OF HETEROGENEOUS COMBUSTION
S. G. Vadchenko*", E. V. Suvorova", and A. S. Rogachev"
aMerzhanov Institute of Structural Macrokinetics and Materials Science, Russian Academy of
Sciences, Chernogolovka, Moscow, 142432 Russia *e-mail: vadchenko@ism.ac.ru
DOI: 10.24411/9999-0014A-2019-10184
Theoretical models of heterogeneous combustion, based on relay-race or quasi-homogeneous heat transfer [1-5], need pictorial interpretation with experimental models. Existing simple models [6] cannot consider porosity or dilution of the mixture by inert addition. Here, we suggest two-dimensional (2D) experimental model of heterogeneous combustion. Macroscopic samples are built of small cylindrical pellets that imitate powder particles of real mixtures. The model allows close or loose packing, modelling inert additions and percolation combustion mechanism. This model also allows to change heat- and mass-transfer conditions on the "side" boundaries of the sample by means of placing various inert pellets along the perimeter. The pellets can be placed vertically or horizontally in one or several layers. Some examples of the packing of pellets are shown in Fig. 1.
az
12 3 4
Fig. 1. Schematics of close and loose packing of the pellets. 1 reactive pellet, 2 inert pellet, 3 pore, 4 igniting coil.
The reactive pellets with diameter d = 1 mm and height h = 1.3 mm were cold-pressed from Hf powder (HFN-1). Density of the pellets was 7.073 ±0.005 g/cm3, which corresponds to 0.536 theoretical density (bulk Hf). Burned pellets were used as inert ones. All experiments were made in air, and the combustion process was video-recorded at 50 fps. Photosensor with slit collimator was used for recording transient of combustion between pellets.
Selected video-frames of the process are presented in Fig. 2 for different packings. The combustion front was almost flat in closed samples, i.e., when side layers were made of inert pellets (Figs. 2a, 2b) and got curved in the open samples with reactive side layers (Figs. 2c, 2d). This can be caused by gas filtration difficulties and heat transfer into inert pellets. After some transition unsteady combustion taking place immediately after initiation, a steady-state regime of combustion appeared (Fig. 3). This effect was observed even for one row of the pellets. Average values of combustion velocity were measured for the steady regime along the central axis of the sample after formation of the cone-like front.
XV International Symposium on Self-Propagating High-Temperature Synthesis
Fig. 2. Initial samples, frames of video recordings (At = 2 s) and burned samples for different types of packing.
t, s
Fig. 3. Coordinate of the combustion front as function of time for one row of the pellets.
2,6.
2,4.
2,2.
f) 2,0. E . ^ 1,8. D • 1,6.
1,4.
1,2.
1,0.
o • 1 2
□ 3
■ 4
A 5
A 6
Fig. 4. Dependences of combustion velocity on the sample structure: 1 closed sample, close packed; 2 closed, square packed; 3 open sample, close packed; 4 open, square packed; 5 closed, close packed, with inert dilution; 6 closed, square packed, with pores.
Figure 4 demonstrates that dilution and porosity result in decreasing of combustion velocity (points 5 and 6), as compared to non-diluted and "pore-free" samples. Transition of the combustion front from one pellet to another generally may include up to 3 stages: flash in the end of the first pellet burning, depression, and second flash in the beginning of the next pellet burning. In some cases, not all stages were observed, depending on contacts between pellets that are influenced by packing type and external stress. Signals of the photosensor for one row of the pellets are shown in Fig. 5. Vertical lines indicate contact boundaries of pellets. In the
16-
4-
2-
10-
8-
6-
4-
2-
10
12
4
n
ÏSHS2019
Moscow, Russia
first case (Fig. 5a), two flashes of the transition process can be observed near the pellet contacts. Intermediate oscillations of the brightness appear due to self-oscillating combustion mode, which forms only if h/d > 1, and don't appear when h/d < 1. The oscillations can be due to cracking of the pellet induced by oxidation reaction. After burning, the samples height increased by 4 - 6%, resulting in stresses, bending of pellets and shifts of the rows. In the second case (Fig. 5b), only two stages, flash and depression, were observed.
0,300,280,260,240,220,200,180,160,140,12
9
t, s
\mt
9
t, s
(a) (b)
Fig. 5. Brightness signal from photosensor: (a) h = 1.2-1.4 mm; (b) h = 0.5-0.6 mm.
0,30-
0,25-
0,20-
0,15-
0,10-
0,05-
0,00
8
0
Conclusion
Experimental model allows pictorial imitation of heterogeneous combustion, including porosity, dilution, and packing density of the combustible mixture. Combustion propagating velocity can be controlled by controlling height of the pellets, which can be used in delay elements and initiators. Such models may be useful for imitation of industrial and forest fires.
1. A G. Merzhanov, Dokl. Academ. Nauk, 1997, no. 4, pp. 505-507.
2. I.A. Filimonov, Fizika gorenia i vzryva, 1998, vol. 34, no. 3, pp. 69-71.
3. A G. Merzhanov, A.S. Rogachev, Russ. J. Phys. Chem, 2000, vol. 74, pp. S20-S27.
4. P.M. Krishenik, A.G. Merzhanov, K.G. Shkadinskii, Fizika gorenia i vzryva, 2002, vol. 38, no. 3, pp. 70-79.
5. A.S. Mukasyan, A.S. Rogachev, Prog. Energy Combust. Sci., 2008, vol. 34, pp. 377-416.
6. S.G. Vadchenko, A.G. Merzhanov, Dokl. Russ. Academ. Nauk, 1997, vol. 352, no. 4, pp. 487-489.
