OCR Text |
Show Tg = l173K, the value of K is in between 0.25 and 0.4. Frame 5 in Fig. 6 shows the effect of K, on the total heat flow rate Q to the fioor. The computed Q is consistently slightly lower than the experimental result for all three values of K. All values of computed Q at K = 0.25 to 0.4 are within the range of the experimental error. If the convective heat transfer is assumed to contribute 4% from the radiation, the prediction the of total heat fiux at K = 0.3 matches of the experimental value. The prediction at K = 0.4 is slightly higher than the experimental one. The value of K, = 0.3 is considered as an optimum value radiative heat transfer. The predictions of the the wall temperatures, Fig. 7, fairly well agree with the experiments. However, in some locations, the wall temperatures are under predicted and in other places, they are over predicted. This behavior is expected. The discrepancies between the predictions the measurement are less than 80 K. The predictions of the roof temperatures at y = l.Om are quite good. The front wall temperature profiles z = 3.0 are predicted very well. The only over predictions are the temperatures on the burner wall. The over prediction is likely because this wall is in the near field region. The effect of K on the gas temperatures is presented in Fig. 8. Lower values of K, give higher gas temperatures. The difference of predicted temperature at K, = 0.2 and K, = 0.4 is about 80 K. It is also shown that the temperature gradients in the far field region z > 1.5 are weak The computed hemispherical radiant heat fiux distribution qr in the gas chamber is more uniform, The lower value of the radiant heat flux shown by frame 3 is because the points see the cooled floor. The effect of the radiative absorption coefficient on the velocity components is evidently small. This is shown in Fig. 10. 6.2. Effect of the total burner momentum Because of the long quarl of the Bloom burner, the gases will lose momentum before leaving the burner exit plane. This loss can be measured, but in computations above the loss is assumed to be negligible. If the loss is taken account, the momentum of the burner exit will less than the inlet momentum on the gas entry to the burner cavity. For this reasoD, a sensitivity analysis for the effect of the total burner momentum Gb was also carried out in this work. Two levels of velocity were set up on the simplified burner exit in the computations to satisfy both mass and momentum flux to be the same as their values in the real gas inlet 11 |