| OCR Text |
Show Use has been made of the axial symmetry condition, so that only half of the physical domain is solved in cases 1, 2, 3, 5 and 6. Near all solid surfaces, the velocity component parallel to the wall concerned, the turbulent kinetic energy and the turbulent dissipation rate are treated through a local wall log-law described by Launder [8]. The axisymmetric condition assumes that fluxes at the plane of symmetry are zero. Case 1 : Ramjet combustor The combusting flow investigated numerically and experimentally by Montazel is studied in this case. The experimental combustion chamber used by Montazel [14] is shown in Figure 1. It is a rectangular parallelepiped. The height and length of the combustion chamber are respectively 50 and 345 m m . Two premixed propane-air flows are injected through two narrow slots ( 5 m m width) in the upper and lower sides of the burner. The computational domain corresponds to the internal dimensions of the experimental combustion chamber (Figure 1) used by Montazel : 35 m m high and 295 m m long. The computational grid used in the present calculations has 80 x 25 nodes in the longitudinal and transversal directions, respectively. The domain is filled with a propane-air mixture with an equivalence ratio of 0.7. The inlet velocity Vo is equal to 24 m/s and the fresh gas temperature T 0 is 300 K. Figure 2 is a picture of the experimental combustion chamber showing the type of flame which is obtained with this arrangement : two reaction zones can be identified, one which is located up-stream with respect to the injection of the fresh gases (called the dome) and the other one downstream from the injection ports. The recent experimental measurements of Montazel [14] will be systematically compared to the predictions of the model described above. The experimental work of Montazel consisted in flame front visualisation by tomography imaging using C C D camera, heat release rate measurements derived from C H radical emission field and temperature measurements derived from gas composition analysis. Figure 3 gives a qualitative comparison between numerical prediction of flame surface density field (in W / m 3) from this work (Figure 3.b) and the experimental measurements of C H radical emission fields (Figure 3.a) from Montazel [14]. It can be seen that the numerical results predict well the qualitative general behaviour of the combustion in the chamber. They generate the main features observed in real situation : the two separate reaction zones in the dome and downstream from the injection ports, the minimum heat release rate obtained in the zone where the injection jets meet and the combustion taking place in the mixing zones on both sides of the fresh gas jets near the injectors. The location of the maximum heat release rate is also well reproduced. For more details, our previous work (see Ref [19]) is entirely devoted to this case. Case 2 : V-shaped gutter This configuration is simulating a simplified afterburner corresponding to the experimental set-up used by Maistret [9] and Maistret et al. [10] which is shown in Figure 4. A mixture of air and propane is injected through a long duct into a rectangular combustor. The height, depth and length of the combustion chamber are respectively 50, 80 and 380 m m . The inlet plane comprises a V-gutter flame holder placed at the duct center and producing a 50 % blockage. In this case, combustion is stabilized by hot gases recirculating behind the V-gutter. The computational grid used in the present calculations has 75 x 25 nodes in the longitudinal and transversal directions, respectively. The domain is filled with a propane-air mixture with an equivalence ratio of 0.75. The inlet velocity Vo is equal to 16 m/s and the fresh gas temperature To is 300 K. Figure 5 shows calculation prediction of the flame surface density field. The flame is stabilized behind the V-gutter and is composed of two distinct reaction zones which touch the boundaries of the combustion chamber exit. Maistret et al. [10] observed experimentally that for an equivalence ratio higher than 0.65, the light emission reaches its maximum in the vicinity of the sidewalls. This peculiar phenomenon is observed because the flame touches the boundaries which are in the present case nearly adiabatic. Following Maistret et al. [10], at the point where the flame reaches the walls, the flow is decelerated and its velocity is less then that existing in the main flow and the flame angle increases (this angle is measured with respect to the axial direction). The wall temperature takes large values and the wall region acts like a secondary stabilization zone for the incoming stream of fresh mixtures. Comparison of heat release rate fields measured experimentally (Figure 6.a) by Maister [10] and predicted numerically by our model (Figure 6.b) is displayed in Figure 6. One first observes that C H map and heat release rate field have the same shape. This is not a surprise because C H is a tracer of the local heat release in flame. Note 5 |
