OCR Text |
Show Tampafatura m *urnaca maaaurad Dy rnarmocoupla Mlrtur a of gat • nd IK te "umict maintajnad at fl«ad ratio ••mp,< tf_ • of Controller a- Signal raiaBngto V«tr«n<« patwaan •dual and ••' point tamparaturaa H | B_rx_ H Compensator Moottted control • tgnai raking Into account non-inaanflaa m tha piocaae Fw ipa«d adtuatad by inodMad ttgral to auM currant operating condMoni Figure 9 : Bilinear controller applied to a furnace Supporting Technologies & Techniques Zone Mathematical Modelling:- A major step forward in the application of zone-type modelling is with us today by virtue of the dramatic increase in computing power over the last few years. Processor speeds of 500 M H z are now commonplace, while P C R A M memory size of 80 M b is also common. This means that the most accurate radiation prediction techniques, such as Hottel Zone and Monte-Carlo, can be used to assess furnace performance etc... in a matter of minutes, if not seconds: Figure 10. In addition, the models are based on algebraic heat balance equations, which do not require excessive iteration. Thus, for the first time, these techniques can be used on laptops, in-situ at plant, and even as realtime aids in sophisticated furnace control systems, as well as for checking other modelling techniques. The opportunities created by these computer hardware developments are tremendous. The basis of the zone models is to have one gas and three surface temperatures in each zone, with predefined combustion and flow. Because calculation of fluid flow, mixing and chemical kinetics governing the combustion of the fuel is not done, zone models must often be applied with the flow and combustion data derived from other sources such as confined jet correlations, physical models or Computational Fluid Dynamics (CFD) simulations. Alternatively, parameter studies can be done on the effect of varying, say, recirculation in a 1-dimensional model. Whichever approach is taken, the total amount of input data required is very limited. Heat transfer is via convection and real gas (wavelength-dependent) radiation and the gases are assumed to be at steady-state or to go through a series of steady states. The walls and load can be either steady or transient. In summary, zone models are simple to set up with limited input data required. In addition, they are robust and fast to run, with accurate handling of heat transfer. Thus they can be readilv transferred |