Options
2011
Conference Paper
Titel
Numerical sensitivity analysis of a complex glass forming process by means of local perturbations
Abstract
In many industrial processes the spatial distribution of the process has to be considered explicitely. Examples are forming processes of the steel and glass industry. The underlying equations of such processes are e.g. Navier Stokes equations and heat transfer with nonlinear material properties. Hence the model of such flow and forming processes is in general a set of coupled, nonlinear partial differential equations (PDEs). The spatio-temporal solution of defined scenarios can be obtained by numerical solution of the PDEs. But unfortunately it is difficult to extract relevant generic knowledge from simulation scenarios. As in reality the processes are disturbed (e.g. caused by inhomogenities in the material or instationary boundary conditions), for process optimization it is important to analyze the impact of spatio-temporal disturbances to relevant process "output" parameters (e.g. geometric parameters of the product). Unfortunately most analytical approaches of sensitivity analysis [3, 4] or perturbation theory [5] can not be applied straightforward to such spatial distributed and complex nonlinear processes. Therefore we propose a concept in which the model with varying definite locally perturbed shape functions is calculated numerically. The tendency and relation between the small perturbation (e.g. furnace temperature variations) and variable field (e.g. geometry and velocity of the produced glass tube) will be estimated from the numerical simulated data. In our study a glass forming process is considered [1, 2]. We investigate the effect of local perturbations (e.g. material inhomogenity or local temperature increasing) to the global variable fields such as temperature, velocity and geometric properties.