Online Monitoring Method for Mold Deformation Using Mahalanobis Distance

Based on the rapid advancement of IoT technology, it has become pervasive in various industries for promoting effective production, including the plastic injection molding industry. In this study, a fundamental investigation of mold deformation was conducted to develop a monitoring system. Mahalanobis distance (MD), which is calculated from mold strain data, was adopted in this monitoring system. We determined that the simple MD index is helpful for judging between normal and abnormal mold states. This index is expected to be a key component of future IoT applications.

3D Heat Conductivity Model of Mold Based on Node Temperature Inheritance

A novel 3D conductive heat transfer model was developed based on node temperature inheritance. Heat transfer of the mold and billet could be analyzed synchronously. In the model, heat transfer in the copper wall was in a steady state, whereas heat transfer in the billet was in a transient state. The temperature distribution indicated that the maximum temperature on the copper wall reached approximately 30 mm below the meniscus. The results were in better agreement with industrially measured data than those of traditional 2D heat transfer models. The model was applied to study the effect of water scale on heat transfer of a billet mold. When the scale thickness increased from 0 to 0.5 mm, the maximum temperature on the copper wall increased from 174 °C to 364 °C, which will lead to mold deformation and peeling of the coating. In addition, the shell thickness slightly decreased with increasing scale thickness.

Modified Shallow Water Equations With Application for Horizontal Centrifugal Casting of Rolls

A numerical model based on the shallow water equations (SWE) was proposed to simulate the two-dimensional (2D) average flow dynamics of the liquid metal spreading inside a horizontally rotating mold. The SWE were modified to account for the forces, such as the centrifugal force, Coriolis force, shear force with the mold wall, and gravity. In addition, inherent vibrations caused by a poor roundness of the mold and the mold deformation due to temperature gradients were applied explicitly by perturbing the gravity and the axis bending, respectively. Several cases were studied with the following initial conditions: a constant average height of the liquid metal (5, 10, 20, 30, and 40 mm) patched as a flat or a perturbed surface. The angular frequency Ω of the mold (∅1150–3200) was 71.2 (or 30) rad/s. Results showed various wave patterns propagating on the free surface. In early stages, a single longitudinal wave moved around the circumference. As the time proceeded, it slowly diminished and waves traveled mainly in the axial direction. It was found that the mean amplitude of the oscillations grows with the increasing liquid height.

Simulation of mold deformation and pattern interaction in nanoimprint lithography

As an emerging manufacturing technique, nanoimprint lithography (NIL) can fabricate micro and nanoscale features of microfluidic devices at very high accuracy and reliability. In high-temperature TNIL process, a polymer melt such as polymethyl-methacrylate (PMMA) is heated beyond the melting temperature so that it behaves predominantly as a fluid during the imprint process. The process parameters such as pressure, temperature, and material properties play critical roles in the NIL process. In this work, the process of thermal nanoimprint lithography (TNIL) is studied computationally with emphasis on the effect of soft-mold deformation on polymer melt flow and finished result by-way-of fluid-structure interaction (FSI) technology. Process is assumed isothermal at 180 °C. Applications of this modeling technique range from micro- and nano-patterns used in micro-channels for biomedical devices to other applications such as biological/particle sensors or super-hydrophobic surfaces. The simulation result is compared to experimental results, and traits observed in TNIL done with soft mold are supported and explained through numerical results.