| OCR Text |
Show behavior of the actual furnace, one has to consider that the actual fuel distribution between the port burners was not uniform. Furthermore, the actual furnace volume is larger than that . the virtual volume of twice the end-section considered in the current model set-up. Some gradients of predicted refractory temperatures ace due to the fact that the refractory material and outer insulation differed from section to section. This effect is especially obvious for the uninsulated tuckstones located just over the glass surface. Relatively low surface temperatures are predicted for these stones. Fig. 8a shows the symmetrized net heat flux distribution predicted at the glass surface for air combustion and assuming uniform effective temperatures of the glass surface of 26000 F (1700 K). The shape and the inhomogeneity of these heat flux profiles indicate a considerable influence of direct flame radiation on overall heat transfer to the glass. The heat fluxes to the glass vary from maximal 98 kW/m2 below the flame to less than 40 kW/m2 adjacent to the bridge wall. Q2-Burner Load Optimization and 02 Base Case In the first step carried out to optimize the 02-firing configuration, the thermal load of the AS and A6 02 burners of the end-section (see Fig. 4) was considered to be equal (Case 2) . The resulting net heat flux distribution to the glass surface is shown in Fig. 8b. Like in the air case, the 02-flames generate a radiation image on the glass surface. Maximum heat fluxes of 114 kW/m2 predicted for flame AS was slightly higher than those predicted for air combustion. However, by comparing Fig. 8b with Fig. 8a, it is obvious, that the glass surface section near the bridgewall receives too much heat in the 02 case. If two glass surface sections are defined so that section 1 is the glass area between the centerlines of Ports 4 and 5 and section 2 is the remaining area up to the bridge wall, then the heat flux ratio between sections 1 and 2 is 1.54 for air Case 1 compared to only 1.03 for 02 Case 2. Furthermore in the 02 case, maximum refractory temperatures of the bridge wall exceeded the ones predicted for air Case 1 by 67°F (37 K). In order to lower the bridge wall temperatures and in order to achieve a heat flux ratio between Sections 1 and 2 which is more similar to that predicted for air combustion, the load ratio AS/A6 was increased from 1 in Case 2 to 3.71 in Case 3. This resulted in an i~crease of the heat flux ratio between Sections 1 and sections 2 from 1.03 to 1.31. It also led to a local heat 12 |
