Glass molding has become a key replication-based technology to satisfy intensively growing demands of complex precision optics in the today's photonic market. However, the state-of-The-Art replicative technologies are still limited, mainly due to their insufficiency to meet the requirements of mass production. This paper introduces a newly developed nonisothermal glass molding in which a complex-shaped optic is produced in a very short process cycle. The innovative molding technology promises a cost-efficient production because of increased mold lifetime, less energy consumption, and high throughput from a fast process chain. At the early stage of the process development, the research focuses on an integration of finite element simulation into the process chain to reduce time and labor-intensive cost. By virtue of numerical modeling, defects including chill ripples and glass sticking in the nonisothermal molding process can be predicted and the consequent effects are avoided

During fabrication of glass lens by precision glass molding (PGM), residual stresses are setup, which adversely affect the optical performance of lens. Residual stresses can be obtained by measuring the residual birefringence. Numerical simulation is used in the industry to optimize the manufacturing process. Material properties of glass, contact conductance and friction coefficient at the glass-mold interface are important parameters needed for simulations. In literature, these values are usually assumed without enough experimental justifications. Here, the viscoelastic thermo-rheological simple (TRS) behavior of glass is experimentally characterized by the four-point bending test. Contact conductance and friction coefficient at P-SK57™ glass and Pt-Ir coated WC mold interface are experimentally measured. A plano-convex lens of P-SK57™ glass is fabricated by PGM for two different cooling rates and whole field birefringence of the finished lens is measured by digital photoelasticity. The fabrication process is simulated using finite element method. The simulation is validated, for different stages of PGM process, by comparing the load acting on the mold and displacement of the molds. At the end of the process, the birefringence distribution is compared with the experimental data. A novel plotting scheme is developed for computing birefringence from FE simulation for any shape of lens

Precision molding is a replicative production method for the mass production of complex glass optics in high precision. In contrast to the traditional material removal process, such as grinding and polishing, the surface as well as the entire shape of the optical component is created by deforming glass at elevated temperatures using precise molding tools with optical surfaces. The molded glass components present high shape accuracy and surface finish after the molding process, therefore no further processing is required. During the molding process, the glass is heated in the molding tool up to above the transition temperature Tg, then pressed into desired shape and cooled down to approximately 200 °C. The precision glass molding is therefore a complex thermo-mechanical process, in which the glass lens undergoes uneven cooling speed and stress distribution. These lead to several drawbacks on the molded glass optics, such as form deviation, index change and fracture. In this study, FEM simulation was employed in order to achieve preliminary understanding of the molding process. The FEM model included viscoelasticity behavior of glass material (stress-relaxation, structure-relaxation and thermos-rheological simplicity), as well as thermodynamics model of the molding machine. In the form of a case study of a real molding example, the form deviation, index change and fracture of the molded glass optic were predicted in advance of the molding experiment by means of the numerical calculation of thermal shrinkage, volume change and stress distribution respectively. The good agreement between simulation results and molding experiment results proves the accuracy of the developed FEM model.

No abstract

Intensively growing demands on complex yet low-cost precision glass optics from the todays photonic market motivate the development of an efficient and economically viable manufacturing technology for complex shaped optics. Against the state-of-the-art replication-based methods, Non-isothermal Glass Molding turns out to be a promising innovative technology for cost-efficient manufacturing because of increased mold lifetime, less energy consumption and high throughput from a fast process chain. However, the selection of parameters for the molding process usually requires a huge effort to satisfy precious requirements of the molded optics and to avoid negative effects on the expensive tool molds. Therefore, to reduce experimental work at the beginning, a coupling CFD/FEM numerical modeling was developed to study the molding process. This research focuses on the development of a hybrid optimization approach in Non-isothermal glass molding. To this end, an optimal configuration with two optimization stages for multiple quality characteristics of the glass optics is addressed. The hybrid Back-Propagation Neural Network (BPNN)-Genetic Algorithm (GA) is first carried out to realize the optimal process parameters and the stability of the process. The second stage continues with the optimization of glass preform using those optimal parameters to guarantee the accuracy of the molded optics. Experiments are performed to evaluate the effectiveness and feasibility of the model for the process development in Non-isothermal glass molding