Composite manufacturing by Liquid Composite Molding (LCM) processes such as Resin Transfer Molding involve the impregnation of a net‐shape fiber reinforcing perform a mold cavity by a polymeric resin. The success of the process and part manufacture depends on the complete impregnation of the dry fiber preform. Race tracking refers to the common phenomenon occurring near corners, bends, airgaps and other geometrical complexities involving sharp curvatures within a mold cavity creating fiber free and highly porous regions. These regions provide paths of low flow resistance to the resin filling the mold, and may drastically affect flow front advancement, injection and mold pressures. While racetracking has traditionally been viewed as an unwanted effect, pre‐determined racetracking due to flow channels can be used to enhance the mold filling process. Advantages obtained through controlled use of racetracking include, reduction of injection and mold pressures required to fill a mold, for constant flow rate injection, or shorter mold filling times for constant pressure injection. Flow channels may also allow for the resin to be channeled to areas of the mold that need to be filled early in the process. Modeling and integration of the flow channel effects in the available LCM flow simulations based on Darcian flow equations require the determination of equivalent permeabilities to define the resistance to flow through well‐defined flow channels. These permeabilities can then be applied directly within existing LCM flow simulations. The present work experimentally investigates mold filling during resin transfer molding in the presence of flow channels within a simple mold configuration. Experimental flow frot and pressure data measurements are employed to experimentally validate and demonstrate the positive effect of flow channels. Transient flow progression and pressure data obtained during the experiments are employed to investigate and validate the analytical predictions of equivalent permeability for a rectangular flow channel. Both experimental data and numerical simulations are presented to validate and characterize the equivalent permeability model and approach, while demonstrating the role of flow channels in reducing the injection and mold pressures and redistributing the flow.

The mathematical and associated computational modeling and analysis of mold filling, heat transfer, and polymerization reaction kinetics in Resin Transfer Molding (RTM) are quite complex and not only require accurate computational approaches to capture the process physics during the simulations, but also must permit complex geometric configurations to be effectively analyzed. The process simulations at a macroscopic level require the representative macroscopic constitutive behavior which can be predicted from a microscopic analysis of the representative volume element (RVE) of the fiber preform configurations. This is first presented here for purposes of illustration in reference to determination of the preform flow permeabilities. Next, an effective integrated micro/macro approach and developments including a viable flow solution modeling and analysis methodology with emphasis on providing improved physical accuracy of solutions and computational advantages are described for the transient flow progression inside a mold cavity filled with a fiber preform under isothennal and non-isothermal flow conditions. The improved physical accuracy and the overall effectiveness of the new computational developments for realistic process modeling simulations are first demonstrated for isothermal conditions. Subsequently, the new integrated flow/ thermal methodology and developments are extended for non-isothermal conditions. The highly advective nature of the non-isothermal conditions involving thermal and polymerization reactions also require special numerical considerations and stabilization techniques and are also addressed here. Finally, validations and comparisons are presented with available analytical and experimental results whenever feasible. Emphasis is also placed upon demonstrations for practical engineering problems.

The success of resin transfer molding (RTM) depends upon the complete wetting of the fiber preform. Effective mold designs and process modifications facilitating the improved impregnation of the preform have direct impact on the successful manufacturing of parts. Race tracking caused by variations in permeabilities around bends, corners in liquid composite molding (LCM) processes such as RTM have been traditionally considered undesirable, while related processes such as vacuum assisted RTM (VARTM) and injection molding have employed flow channels to improve the resin distribution. In this paper, studies on the effect of flow channels are explored for RTM through process simulation studies involving flow analysis of resin, when channels are involved. The flow in channels has been modeled and characterized based on equivalent permeabilities. The flow in the channels is taken to be Darcian as in the fiber preform, and process modeling and simulation tools for RTM have been employed to study the flow and pressure behavior when channels are involved. Simulation studies based on a flat plate indicated that the pressures in the mold are reduced with channels, and have been compared with experimental results and equivalent permeability models. Experimental comparisons validate the reduction in pressures with channels and validate the use of equivalent permeability models. Numerical simulation studies show the positive effect of the channels to improve flow impregnation and reduce the mold pressures. Studies also include geometrically complex parts to demonstrate the positive advantages of flow channels in RTM.

The Adapteva Epiphany many-core architecture comprises a 2D tiled mesh Network-on-Chip (NoC) of low-power RISC cores with minimal uncore functionality. It offers high computational energy efficiency for both integer and floating point calculations as well as parallel scalability. Yet despite the interesting architectural features, a compelling programming model has not been presented to date. This paper demonstrates an efficient parallel programming model for the Epiphany architecture based on the Message Passing Interface (MPI) standard. Using MPI exploits the similarities between the Epiphany architecture and a conventional parallel distributed cluster of serial cores. Our approach enables MPI codes to execute on the RISC array processor with little modification and achieve high performance. We report benchmark results for the threaded MPI implementation of four algorithms (dense matrix-matrix multiplication, N-body particle interaction, a five-point 2D stencil update, and 2D FFT) and highlight the importance of fast intercore communication for the architecture.

No abstract