B. Golchert

A Simulation Approach for Bubble Flow in a Glass Melter

Combustion heat is used in glass furnaces to melt sand and cullet (scrap glass) into liquid glass to make products. The glass flow in a melter consists of solid particles of sand/cullet, liquid glass, and bubbles. Bubbles formed in the melting processes due to the glass reactions have strong impacts on glass quality and furnace efficiency. Smaller bubbles entrained in the liquid flow degrade the glass quality. Larger bubbles rise to the top of the melter and form a foam layer that impedes the radiation heat transfer from the combustion space and lowers the furnace efficiency. An Eulerian approach was developed to simulate the bubble flow in a glass melter. The approach divides bubbles into various groups and treats each group of bubbles as a continuum. The mass, momentum, and energy conservation equations of the bubble flow are derived to solve for local bubble properties. The approach was incorporated into a multiphase reacting flow computational fluid dynamics code that simulates overall furnace flows to evaluate the impacts of bubbles on glass quality and furnace efficiency.Copyright

Effect of Nitrogen and Oxygen Concentration on NOx Emissions in an Aluminum Furnace

For many aluminum melting furnaces, natural gas is mixed with air. The ensuing heat from combustion is then used to melt the solid aluminum and heat the liquid metal. Of increasing concern to the industry are the more stringent regulations in regard to NOx emissions from these plants. The formation of NOx mainly depends on the concentration of nitrogen and the temperature of the gas. One problem that affects this formation that has not been adequately addressed is the variability of the local natural gas supply. Natural gas has molecular nitrogen as a portion of its composition. This percentage ranges from approximately one to seven percent of the total mass fraction. In addition, the aluminum industry is investigating methods to reduce NOx emissions. One method is to replace some of the combustion air with pure oxygen. This reduces the amount of nitrogen coming into the furnace, but also raises the combustion temperature which could promote NOx production. This paper details a systematic computational fluid dynamics study on how the variability of the nitrogen concentration coupled with the partial replacement of air with pure oxygen affects heat transfer and pollutant formation in an aluminum furnace. Trends will be discussed as will the ideal oxygen concentration for a given nitrogen mass fraction.Copyright

Heat Loss Analysis of Dross Removal in an Aluminum Furnace

During the melting/production of aluminum, dross forms on the surface of the molten aluminum. Dross is a contaminant and an insulator so the aluminum industry regularly removes this dross from the aluminum surface. This is done by opening a door on the side of the furnace while the burners are still in operation and manually raking off the dross. As a result, a lot of heat is lost while the door is open. This lost heat depends upon the length of time the door is open and on the firing rate of the burners. This paper will present computational results on how much heat is lost while the door is open, the effect of this heat loss on the combustion space flow field, and the length of time needed to return to equilibrium conditions once the door is closed again.© 2006 ASME

The Effect of Glass Foam on Heat Transfer in a Glass Furnace

In high temperature environments such as a glass furnace, radiation is the dominant mode of heat transfer from the combustion process to the glass melt. It has long been thought that the presence of glass foam on the surface of the molten glass will impede the transmission of the radiant energy from the flames/superstructure to the melt. Several correlations relating foam thickness and material properties to the transmission of radiant energy have been incorporated into the Argonne National Laboratory Glass Furnace Model. Previously, a multiphase computational model to calculate foam formation had been developed and incorporated into the melt model. Prior to this work, no computational fluid dynamics code had the capability to calculate foam distribution or the effect of foam on heat transfer. This paper will present the formalism behind this modeling effort and will indicate the effect of this computed foam layer on the overall heat transfer in a glass furnace.

Modeling of Regenerative Furnace Ports

In order to increase overall efficiency, many industrial glass furnaces are regenerative; that is, the heat from the exhaust gases is used to preheat in the in-coming combustion air. The ports on these furnaces inject stream(s) of fuel into the preheated air stream and then combustion occurs inside the combustion chamber. Modeling of the exact detail of these furnace ports in addition to modeling the combustion space proper becomes computationally burdensome since many of these furnaces are extremely large. This paper presents an engineering approach using computational fluid dynamics to model both the major effects of the furnace ports in addition to calculating the detailed flow field in the combustion space. This approximation has been incorporated into a complete (combustion space/glass melt) furnace simulation. This engineering approach significantly reduces run time while still maintaining results that represent the conditions seen in the furnace. This paper will present this approach as well as some preliminary comparisons with actual furnace data/observations.Copyright

The effect of gases emitted from batch/glass reactions on the combustion space flow field.

The concept of ‘coupling’ a combustion space CFD code to a code that models the molten glass flow is not a new idea. However, this concept has been limited to an energy coupling; the heat flux calculated in the combustion space model is used to drive the glass melt while the calculated surface temperature is used in the radiative heat transport calculation in the combustion space computation. In reality, there is significant mass (mostly gas) transport from the batch/molten glass into the combustion space. This is an important phenomenon to be modeled since these gases, particulates and volatiles will be removed from the combustion chamber and hence raise environmental considerations. In addition, these released gases have a distinct influence on the flow field in the combustion space. The ANL Glass Furnace Model has been augmented to calculate the chemical reactions that release gases in the batch and in the melt. This work presents preliminary results indicating the effect of these gases on the combustion space.Copyright

Modeling and Preliminary Validation of a Regenerative Furnace Using the ANL Glass Furnace Model

The ANL Glass Furnace Model (GFM) was developed for steady state simulation of industrial glass furnaces. Unfortunately, a large fraction of the operating glass furnaces do not operate in a steady state mode and computational costs make it prohibitive to run the simulations in a transient mode. A solution methodology was developed to model these transient furnaces in steady state mode. This solution methodology was used to model a small, industrial furnace on which a relatively comprehensive set of data was taken. This paper presents the solution methodology in detail along with some of the qualitative validation results indicating the validity of the modeling approximation.Copyright

Modeling of Radiation Heat Transfer in Glass Melts

Glass furnaces combust fuels and oxidizers so that the batch will be melted and the molten glass refined. The heat generated from the combustion space is mainly transmitted to the glass melt by radiation. This radiation is strongly wavelength dependent due to the nature of soot and carbon dioxide/water emission. Conventional wisdom indicates that the transport of radiation in molten glass can be modeled using a diffusion approximation. However, for many glass compositions, this approximation may not be justified since there is appreciable transmission of thermal energy at lower wavelengths. To account for the detailed transmission of radiation in the glass melt, a recently developed radiation model was incorporated into a glass melt simulation. This spectral model rigorously conserves emitted and absorbed energy throughout the melt and the surrounding walls. The glass melt flow simulation is a multi-phase reacting flow computational fluid dynamic code that accounts for the solid batch material, liquid glass, and gas bubble. This paper presents details of the radiation model along with results depicting the effect of the detailed radiation transmission on the glass melt flow field.Copyright