Kisa Matsushima

A Numerical Approach for Injection Molding of Short-Fiber-Reinforced Plastics Using a Particle Method

This study proposes a numerical approach for predicting the injection molding process of short-fiber-reinforced plastics using the moving particle semi-implicit (MPS) method, which is a particle-simulation method. Unlike conventional methods using orientation tensors, this approach represents all fibers and resin as an assembly of particles, and automatically analyzes the interaction between fiber and resin and between fibers. In addition, this method can follow the motion of a specific fiber, which is a significant advantage over orientation tensors. This study simulated the injection molding of short-fiber-reinforced plastics; the thermoplastic resin was considered as an incompressible viscous fluid and the fibers were modeled as rigid bodies. The numerical result illustrated that the molding material was unidirectionally reinforced by short fibers since the fibers rotated and were aligned parallel to the flow direction due to the velocity gradient near the wall boundary. Moreover, the stagnation of resin at a corner was predicted. The results agreed well with previous studies, and the present approach was confirmed. Beyond this, we predicted the accumulation of fibers near the wall due to the velocity gradient, which could not be represented by conventional simulations based on orientation tensors.

Numerical Simulation to Detect the Change of Airfoil Stalling Characteristics Dependent on Reynolds number

The compact scheme with a new method to detect the transition point is applied to low-Reynolds number (Re = 3.43 x 10?, 2.51 x 10?) over the NACA9324 airfoil. RANS simulations have difficultly to predict the stalling characteristics because the large separation and transition occur in flow at low-Reynolds number region, NACA9324 shows peculiar aerodynamic characteristics. Its stall type change from leading-edge stall to trailing-edge stall when the Reynolds number increase over 3.43 x 10?. In preliminary study, fully laminar computations using conventional 3 rd -order TVD and 2 nd -order central differencing could not predict stalling characteristics of the NACA9324. Because that computations couldn’t resolve a laminar bubble and a transition from laminar to turbulent flow. Thus, compact high-order scheme with the new transition method is applied to the flow around the NACA9324 airfoil. As a current result, the compact scheme succeeds in predicting the trailing-edge stalling aerodynamic characteristics. The new transition method is being examined.

Supersonic Wing Design Method Using an Inverse Problem for Practical Application

Supersonic wing design system using an inverse problem realizes high-fidelity geometric control with much smaller number of flow simulations than generic design system. It was used in Japanese SST project “NEXST-1” successfully to design the NLF wing. However, there is still need for improvement to apply this method to critical or severe design problems. One of the required improvements is reduction of load in CAD process and another is accuracy in determining geometric correction values. In this research, aiming to reduce the load in CAD process, geometric representation by analytical functions was proposed on the basis of the PARSEC representation. On the other hand, to enhance the accuracy in determining geometric correction values, the basis equation used in inverse problem solver was modified. The redesign of NLF wing of “NEXST-1” was conducted to validate the modified design system. Then it is confirmed that the modified one has more ability than original one.