Numerical Simulation Of The Flow And Fiber Orientation In Reinforced Thermoplastic Injection Molded Products

Author : Jin Wang
Publisher :
ISBN 13 :
Total Pages : 290 pages
Book Rating : 4.:/5 (814 download)

DOWNLOAD NOW!

Download or read book Improved Fiber Orientation Predictions for Injection Molded Composites written by Jin Wang and published by . This book was released on 2007 with total page 290 pages. Available in PDF, EPUB and Kindle. Book excerpt: Short fibers are commonly utilized to reinforce the polymer matrix for injection molded parts. Fibers suspended in the molten polymer matrix are oriented by the flow during the mold filling process, and acquire a preferential orientation pattern in the final part. The fiber orientation introduces anisotropy to the mechanical and thermal properties of the material, including the elastic modulus, the tensile strength, and the thermal expansion. A final part is stronger and stiffer in the direction along which the most fibers align, while it is weaker and more compliant in the other directions. An accurate prediction of fiber orientation in injection molding is crucial for designing a mold and controlling part properties. The well-established Folgar, Advani, and Tucker model is widely used to predict fiber orientation. However, recent experiments indicate that this theory overestimates the change rate of the fiber orientation tensor, and therefore predicts a similar orientation pattern in injection-molded parts regardless of the flow length, the part thickness, or the mold filling speed. A strain reduction factor (SRF) was introduced by Huynh (2001) to decrease the change rate of orientation tensor. Though it produces an excellent agreement with experimental data, the SRF model is not objective and encounters difficulty in complex flows. Inspired by the idea of reducing the growth rates of the eigenvalues of the orientation tensor by a scalar factor, a new orientation model was built, in which we modified the closure and fiber-fiber interactions terms accordingly. A possible approach to reduce the rotation rates of the eigenvectors was also explored, but no successful model with this feature was found. A finite difference program was developed and used to simulate the filling process for two simple geometries: end-gated strips and center-gated disks. The program is based upon the Hele-Shaw approximation to solve the velocity field, and implements the new orientation model to predict the fiber orientation. The results using the same scalar factor as the SRF model show an excellent agreement with experimental measurements, for both strips and disks, in small thicknesses and at different filling speeds. A special treatment was proposed for thick strips, to account for the radial flow front that is observed in short-shot experiments and is different from the almost flat flow front in thin strips. The finite difference program was also extended to simulate the filling process of injection molds with rotation, compression, and expansion (RCEM). The comparison of the predicted and measured fiber orientation demonstrates again the usefulness of our new orientation model. The new orientation theory was further implemented to model complex flows where the full equations of motion must be solved. The solver of fiber orientation equations was added to FIDAP(TM), a commercial finite element software, through user subroutines. The final program is able to solve for velocity, temperature, pressure, and fiber orientation in any two- or three-dimensional geometry. Flow through the gate of our end-gated strips was simulated, and the fiber orientation was calculated. The orientation results were compared to the measured values at a region just inside the gate, and then were used as inlet conditions for the finite difference program to successfully predict the downstream orientation. Our new orientation model is aphenomenological theory, and the value of the scalar factor to reduce the orientation change rate is determined by matching experimental data. Since the rheology of a fiber suspension is affected by the fiber orientation, a rheological experiment measuring the shear viscosity and the normal stress difference is one approach to determine the phenomenological parameters. The formulations were worked out for the shear stress and the normal stress difference with respect to the fiber orientation in a parallel-disk rotational rheometer, and the model was fit to the measured values in a least square sense by adjusting the model parameters. This provides a convenient and viable route for determining the parameters of the orientation model.