In industrial glass production, annealing and tempering processes are applied to control residual stresses of products. During cooling, temporary and remaining thermal stresses build up in a glass product, influencing the risk of breakage and thus the yield of a cooling process. The physical strength of a glass product is, to a large extent, determined by residual stresses after cooling.

In annealing, the aim of controlled cooling is relaxation of stresses in the glass that have been initiated during forming and subsequent cooling. In tempering, a desired residual stress pattern (compression at the surface) is induced in a glass product by employing an appropriate heat and cool cycle.

Tools

TNO Glass Group has developed validated software tools for improving industrial glass coding processes. Lehr simulations are being used for improving performance of annealing furnaces. Detailed product simulations serve to analyse and improve the characteristics of glass products and production processes.

Lehr simulations

Cooling rate and temperature uniformity are important issues in a glass lehr. Optimum lehr performance means an optimum cooling-rate for the products while maintaining temperature differences between products below an acceptable level. Temperature uniformity is analysed by adapting a model of a 'standard' lehr to fit the actual situation. The outcomes of the model calculations are local temperatures of furnace walls, belt and products and heat fluxes between these at distinct time steps. Thus the temperatures of walls, belt and products are available at every lehr zone. Risk of breakage during cooling and final product quality are analysed by calculating thermal and residual stresses inside the glass. Variations of lehr settings and product properties can be used to determine the influence of changing process parameters and to optimise the settings for each type of product. In this way, the method is easy to apply to existing lehrs.

When designing a new lehr or other system for heating and cooling, the lehr model can also be used to define the geometry and heating/cooling system. Parameters can be altered quickly in the simulations, so that a design can be optimised with respect to cooling rate and temperature uniformity.

The lehr simulation tool is based on Finite Element software for calculation of heat transfer by convection, conduction and radiation. In the furnace, heat is transferred to the products by combined convection, conduction and radiation. Inside the products, heat transfer by conduction and radiation determines the temperature history. Using a marching procedure, quick simulations of all heat transfer phenomena in the lehr are done while maintaining accurate simulation results. Temperature history graphs of details of the products at different positions on the conveyor belt are the result of these calculations. Furthermore temperature uniformity graphs of the furnace can be shown easily with this simulation tool.

Product simulations

When more detailed phenomena on product level like a complicated (3-D) product geometry are important for process performance or product quality, a global method like this does not give the appropriate results. For this kind of problem, detailed product simulations of the cooling process are required. Accurate detailed temperature and stress fields are calculated by CFD analysis of heat transfer towards and inside the product.

Heat transfer towards a product occurs by turbulent flow and radiation from the furnace environment, and heat transfer inside a glass product occurs by conduction and radiation. Stresses in the glass are calculated from glass specific behaviour, taking into account fluidous or (visco-) elastic properties of glass dependent on its temperature. Using detailed product simulations, local temperature and stress concentrations (hot-spots) are determined, which give insight into quality and defect risks on a local product scale.

Annealing and tempering processes strongly influence the final quality of glass products. Simultaneous increase of production speed and yield are nowadays challenges in industrial glass production that also cause higher demands on annealing and tempering processes. TNO Glass Group simulation tools are powerful tools to take up these challenges. They are suitable for industrial process optimisation, as well as for glass product improvement by simulation on global or on detailed scale.

More often than not, the application of one of these tools offers a payback time of one to three months in terms of a reduction of reject.

* Maurice Limpens and Roeland Brugman, TNO-TPD, Eindhoven, The Netherlands. Tel: +31 40 265 0210. Fax: +31 40 265 0850. Email: info@tpd.tno.nl; Website: www.tpd.tno.nl/glassgroup