(353f) Modeling a Direct- Fired Industrial Tunnel Furnace Using an Implicit Multiple 1D Approach

Today’s expectations regarding CFD simulations for industrial furnaces are ambitious since they represent a viable option to reduce the need of conducting usually expensive and time-consuming experiments. Furthermore, CFD simulations provide the possibility to estimate certain process parameters, as for example local concentrations or temperature profiles, where conventional experiments and measurements fail. Although offering a broad spectrum of advantages, most simulations require immense computational power because of the high number of cells and algorithms, which must be iteratively solved. Due to these disadvantages the time required to deliver results is high. Therefore, both experiments and CFD simulations, are incapable of providing essential data needed for process related decisions, raising the need of reduced furnace models that can generate results faster.

This paper proposes a method that can decrease computational cost significantly, by separating the simulation domain into several coupled one-dimensional regions. Thus, hours of process time and associated furnace parameters can be predicted within minutes of calculation. The advantage by introducing reduced furnace models is emphasized by several authors who are using similar mathematical concepts and discretization strategies for computation. The fundamental difference between reduced furnace models and CFD simulations is the significant difference in cell resolution and connectivity. In consequence of the interconnected one- dimensional discretization the total number of cells is reduced drastically, hence reducing the cell resolution and improving the simulation speed. Additionally, the connectivity of the respective cell regions is only towards one dimension which further improves the calculation performance because of the reduced number of equations to solve. Based on this approach the generated results represent global process parameters like the average gas temperature or concentration within a specific furnace compartment. Those results can be utilized afterwards to determine the energy efficiency of a specific furnace setup or the effect on the heat-treated products in regard of quality control.

Although, if using reduced furnace models as a data generating method in addition to CFD simulations and experiments seem to be promising, caution must be exercised when using them. More specifically, a detailed knowledge of the heat-treated products, as well as the required dew temperatures are of uppermost importance in order to implement optimal solutions. This work concentrates on a direct- fired industrial tunnel furnace that treats products at temperatures above 1400°C, making radiation to the major mode of heat transfer. The modeling of the direct- firing process bases on an equilibrium approach utilizing calculated gas concentrations to determine the released energy supply in the burning zone. Furthermore, the additional modes of heat transfer, conduction and convection, are added to give a complete physical description. The product feed is done on a continuous basis depending on the pulling time of the respective products, which is deposited in a production plan provided to the simulation. Subsequently to the model implementation, an existing furnace of industry partners is modeled and afterwards validated by measurement data. Ultimately, the generated temperature profiles show a high potential to predict the influence of the applied heat on the processed products properties along the furnace axis.