Radyasyon baskın fırınlarda ısıl işlemlerin tersten tasarımı

This dissertation focuses on optimizing heat treatment processes, crucial for manufacturing industries, with the aim of ensuring desired treatment of workpieces while addressing energy costs. Traditional optimization approaches have primarily concentrated on uniform treatment over the workpiece surface, often neglecting internal conduction, leading to significant modeling errors when internal thermal gradients are critical in industrial scenarios. Our model employs a design point approach within the workpieces to define design conditions, overcoming the limitations of traditional methods that ignore internal conduction of workpieces. The study investigates the heat treatment of workpieces within batch furnaces, where the radiation heat transfer within the enclosure is dominant while incorporating internal conduction inside workpieces. A novel gradient-based optimization methodology has been developed, incorporating a direct differentiation approach for calculating sensitivity coefficients. Regularization techniques are employed to address the ill-posed nature of inverse radiant boundary design problems. Calculating sensitivity coefficients through direct differentiation typically requires explicit formulations of the variables, which are often unavailable due to radiation-conduction coupling in our target problems. To overcome this, floating random walk methods were utilized to derive explicit formulations for workpiece surface and internal design point temperatures, a novel application for inverse radiant boundary design problems. The research involves comprehensive analyses to verify and apply the developed methodology. The methodology is verified by reconstructing known heater temperatures for heating two glass panes. Analysis of design point configurations reveals that increased number of design point enhances the accuracy of the inverse solution by providing more precise sensitivity information. The study also examines the impact of noisy input data and regularizer weights, demonstrating the robustness and stability of the developed methodology. Comparative analysis of furnace operation using temperature-based and power-based approaches showed that both methods achieved uniform heating of six glass panes, with the Power-based Approach being more practical by avoiding negative power inputs. Analysis of different heating curves showed, smoother heating-to-soaking transition reduced power demands, providing insights into optimal heating curves. For a six-hour heating treatment of a cubic stainless-steel workpiece, the inverse solution determined required transient heater powers, with central heaters driving heating and edge heaters ensuring uniformity in soaking phase. Analysis of heater layout configurations showed that increasing the number of heater elements enhances temperature uniformity, providing more precise thermal management. The cooling phase analysis compared two approaches; determination of heater powers together alongside a predefined constant cooling and simultaneous determination of heater and chamber powers. The latter achieved high cooling uniformity but required higher energy, highlighting trade-offs between uniformity and energy demands. These analyses demonstrate the robustness and practical applicability of the developed methodologies, providing valuable insights for optimizing furnace design and operation. The study's findings are particularly relevant for industries such as glass, semiconductor, automotive and aerospace, where precise heat treatment processes are essential for producing high-quality components. By adopting these optimized practices, furnace designers and operators can gain significant understanding into optimizing furnace operating parameters, leading to improved process effectiveness, energy savings, and high-quality outcomes.