| OCR Text |
Show in h~t environment (e.g. in furnaces). Further, none of the papers cite any efforts utilizing a mult~-flame DFI system. In his comprehensive review R. Viskanta [7] commented that no publ~shed data on ~ame jet array could be found. Most researchers have investigated rel~t1v~ly low vel~cI~ premixed flame jets, because of the potential for flame separation, which Increases WIth mcreasing the flame velocity. To provide maximum convective heat transfer rates, however, the velocity levels should be maximized. The benefits of using high velocity multi-flame DFI have been assessed through experimental as well as theoretical studies. The experiments were carried out at conditions representing real industrial heating processes. Initial experiments showed that, within appropriate firing rate and temperature ranges, multi-flame DFI systems can provide stable combustion at nozzle exit velocities up to Mach 1, without any special components (tunnels, flame holders etc.). Measurements of dynamic pressures, composition of combustion products, heat transfer rates at the load surface, and temperature fields around the load were carried out. In parallel, a simple 2-D model of fluid dynamics and heat transfer was developed, using a coarse zone scheme for radiative heat transfer. The main challenges faced in developing the model were to set the boundary conditions on the 'free' surfaces and to couple the heat exchange between the fme and the coarse grids. To take into account the influence of turbulent temperature fluctuations on combustion rate, the Arrenius expression for reaction rate was modified. The experimental data are compared here with the model predictions for convective and radiant heat transfer. EXPERIMENTAL APPARATUS The experimental furnace used was designed to be comparable in size to actual electric induction furnaces. The design and operating parameters of the experimental furnace were specified based on the results of preliminary laboratory experiments. The minimum distance between the brickwork and the load was specified to be approximately the same as in industrial induction furnaces designed for heating 100 mm ( 4 in) tubes. The nozzle size was specified to ensure high flame jet core temperatures before impingement. The maximum firing rate was specified to provide heat fluxes approaching those achieved with induction heating. A standard 50 kPa (200 in.w.c.) industrial air blower was used to provide nozzle exit velocities of up to 300 rnls (980 ftls). The furnace chamber was lined with refractory bricks; and had internal dimensions of 250mm (10 in.) by 230 mm (9 in.) by 930 mm (37 in.) long (see Figure 1). A water-cooled cylindrical calorimeter with a diameter of 108 mm (4.3 in.) was mounted on the furnace axis to simulate an elongated billet. A total of 104 nozzles made from stainless steel were installed into the brickwork and directed towards the surface of the calorimeter. The nozzles were arranged in four rows around the furnace axis with a spacing of 30 nun (1.3 in.) between the nozzles. Preheated or ambient combustion air-natural gas mixture was distributed unifonnly to the individual nozzles. The system was designed to initiate combustion close to, but downstream of the nozzle tips and continue it over and around the calorimeter's surface. The strong recirculation zones created by the g~ streams. ~xiting the nozzles provide inte~se and stable combustion over a wide range of Jet velOCItIes. No problems occurred WIth 4 |
