Polydomain model predictions of molecular orientation in isothermal channel flows of thermotropic liquid crystalline polymers

We report process simulations of molecular orientation of liquid crystalline polymers for isothermal channel flows in extrusion. The simulations use a “polydomain” model due to Larson and Doi (Larson and Doi, J. Rheol., 35, 539 (1991)), which is implemented by exploiting a nearly exact analogy with a fiber orientation model that is widely used for analysis of composites processing. Simulation results are compared to experimental data previously obtained using a customized X‐ray capable extrusion die that allows a wide range of channel flow geometries to be explored. Competition between inhomogeneous shear and extension in these kinematically complex flows has a profound effect on the resulting molecular orientation distributions. We find that the Larson–Doi model successfully predicts most aspects of the experimental observations, demonstrating its utility for process modeling. POLYM. ENG. SCI., 2008. © 2008 Society of Plastics Engineers     

Bulk and surface molecular orientation distribution in injection‐molded liquid crystalline polymers: Experiment and simulation

Bulk and surface distributions of molecular orientation in injection‐molded plaques of thermotropic liquid crystalline polymers (TLCPs) have been studied using a combination of techniques, coordinated with process simulations using the Larson‐Doi “polydomain” model. Wide‐angle X‐ray scattering was used to map out the bulk orientation distribution. Fourier Transform Infrared Attenuated Total Reflectance (FTIR‐ATR) and Near‐Edge X‐ray Absorption Fine Structure (NEXAFS) were utilized to probe the molecular orientation states to within about ～5 μm and ～2 nm, respectively, of the sample surface. These noninvasive, surface‐sensitive techniques yield reasonable self‐consistency, providing complementary validation of the robustness of these methods. An analogy between Larson‐Doi and fiber orientation models has allowed the first simulations of TLCP injection molding. The simulations capture many fine details in the bulk orientation distribution across the sample plaque. Direct simulation of surface orientation at the level probed by FTIR‐ATR and NEXAFS was not possible due to the limited spatial resolution of the simulations. However, simulation results extracted from the shear‐dominant skin region are found to provide a qualitatively accurate indicator of surface orientation. Finally, simulations capture the relation between bulk and surface orientation states across the different regions of the sample plaque. POLYM. ENG. SCI., 50:1864–1877, 2010. © 2010 Society of Plastics Engineers     

Mechanical properties of glass fiber reinforced polypropylene injection molded with a rotation mold

A special mold (Rotation, Compression, and Expansion Mold) was used to impose a controlled shear action during injection molding of short glass fiber reinforced polypropylene discs. This was achieved by superimposing an external rotation to the pressure‐driven advancing flow front during the mold filling stage. Central gated discs were molded with different cavity rotation velocities, inducing distinct levels of fiber orientation through the thickness. The mechanical behavior of the moldings was assessed, in tensile and flexural modes on specimens cut at different locations along the flow path. Complete discs were also tested in four‐point flexural and in impact tests. The respective results are analyzed and discussed in terms of relationships between the developed fiber orientation level and the mechanical properties. The experimental results confirm that mechanical properties of the moldings depend strongly on fiber orientation and can thus be tailored by the imposed rotation during molding. POLYM. ENG. SCI. 46:1598–1607, 2006. © 2006 Society of Plastics Engineers.     

Prediction of fiber orientation in a rotating compressing and expanding mold

In the rotating/compressing/expanding mold (RCEM), one mold wall can expand, compress, and rotate during injection molding, thus offering opportunities to control the thermomechanical history of a polymer and its microstructure. A computer simulation of flow and fiber orientation in RCEM was developed. The predictive model extends the generalized Hele‐Shaw formulation to account for compression/expansion and rotation of the mold wall, and uses the Folgar–Tucker model for fiber orientation predictions. A 20% GF polypropylene was molded under various molding conditions. The predicted fiber orientation distributions were compared with experiments. The model compares favorably with experiments, provided that the fiber orientation equation is modified by a strain‐reduction factor that slows the transient development of fiber alignment. The effect of fountain flow on orientation must also be included to correctly predict fiber orientation near the mold walls, mainly for the case of stationary and linear motions of the mold surface. Compression or expansion of the mold has only a small effect on fiber orientation, but rotation of the mold dramatically changes the orientation, causing fibers to align in the tangential direction across the entire thickness of the molding. This rotation action perturbs the fountain flow and becomes the dominant factor affecting fiber alignment across the entire cavity thickness. POLYM. ENG. SCI., 2008. © 2008 Society of Plastics Engineers.     

The process of cavity filling including the fountain flow in injection molding

This work deals with the simulation of the filling of a cavity utilizing the Marker-and-Cell numerical technique in solving the transient problem involved. The cavity is confined by two parallel plates, and is “end fed.” The flow was assumed isothermal and the fluid incompressible, obeying the power law model. Special attention was given to the flow region near the advancing melt front, in order to obtain a better insight of the “fountain effect,” during which the fluid flows from the center to the walls of the cavity. The results of the simulation of the front flow region are supported by and in qualitative agreement with experimental results involving “tracer resins” during cavity filling. Although the flows considered were slow and isothermal, this study has significant practical ramifications on industrial mold filling during injection molding.     

Modeling and simulation of fiber‐reinforced polymer mold‐filling with phase change

A gas‐solid‐liquid three‐phase model for the simulation of fiber‐reinforced composites mold‐filling with phase change is established. The influence of fluid flow on the fibers is described by Newton's law of motion, and the influence of fibers on fluid flow is described by the momentum exchange source term in the model. A revised enthalpy method that can be used for both the melt and air in the mold cavity is proposed to describe the phase change during the mold‐filling. The finite‐volume method on a non‐staggered grid coupled with a level set method for viscoelastic‐Newtonian fluid flow is used to solve the model. The “frozen skin” layers are simulated successfully. Information regarding the fiber transformation and orientation is obtained in the mold‐filling process. The results show that fibers in the cavity are divided into five layers during the mold‐filling process, which is in accordance with experimental studies. Fibers have disturbance on these physical quantities, and the disturbance increases as the slenderness ratio increases. During mold‐filling process with two injection inlets, fiber orientation around the weld line area is in accordance with the experimental results. At the same time, single fiber's trajectory in the cavity, and physical quantities such as velocity, pressure, temperature, and stresses distributions in the cavity at end of mold‐filling process are also obtained. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2016 , 133, 42881.     

Particle image velocimetry in molten plastic

Injection molding is a standard manufacturing process for plastic parts. The most important process step, mold filling, involves unsteady nonisothermal flow of a non‐Newtonian molten plastic. Mold‐filling flow largely determines molecular orientation within the final part and thereby influences final part properties. This article describes techniques for successfully applying particle image velocimetry (PIV) to molten plastic flow during injection molding. The primary experimental challenges include the following: engineering optical access to the molten plastic flow at elevated temperatures (230–245°C) and pressures (～20 MPa), finding particles that survive the thermal‐mechanical environment that melts the plastic, and developing experimental and data‐reduction techniques that allow multiple imperfect planar PIV measurements to be combined. Here, a custom optical‐access mold allowed mold center‐plane PIV to be performed in molten polystyrene. Simple statistical assessments of the velocimetry data and scaled residuals of the continuity equation suggest that the PIV can be conducted in molten plastics with an uncertainty of ±2%. POLYM. ENG. SCI., 2011. © 2011 Society of Plastics Engineers     

Modeling fountain flow and filling front shape in reaction injection molding

A numerical model of the reaction injection molding process was developed to test front shape and flow approximations employed in previous models. The model was two-dimensional and simulated the flow, reaction, and heat transfer in the typically long axial dimension and the typically small thickness dimension of a mold. The filling front shape and the velocity profiles in the filling fluid were determined by numerical solution of the momentum equation with the appropriate stress boundary conditions using the method of Patankar (1980). The predicted temperature and conversion results agreed with calculations assuming that the front was flat perpendicular to the flow and that a parabolic velocity profile existed behind the fountain flow region at the front. Thus, simple assumptions about front shape and velocity in the thin dimension of a reaction injection mold can be employed without significant loss of accuracy in modeling reaction injection molding.     

Effect of Melt‐Memory on the Crystal Polymorphism in Molded Isotactic Polypropylene

The influence of γ‐quinacridone as a β‐crystal nucleating agent in injection molded isotactic polypropylene (iPP) is discussed. Samples are injection molded and characterized via polarized‐light optical microscopy and X‐ray diffraction. Mold‐filling simulation is used to understand the shear and cooling processes during sample preparation. The cooling rate associated with the quench near the mold wall is estimated to be greater than 600 K s^?1 using simulation, confirming previous studies that β‐crystal growth is not supported at that cooling rate. The non‐nucleated samples form β‐crystals at a distance of 100–300 µm from the skin and in the core of the sample, which is not expected based on quiescent cooling data. Since the mold‐filling simulation does not predict shear in the core, the formation of the β‐crystals formed in this region is attributed to shear‐induced crystallization effects in the injection unit of the molding machine that are not modeled in flow simulation, as they are typically excluded from any molding simulation analysis. This “melt‐memory” effect has shown to be significant, and it is suggested that the prediction of final properties of injection moldings requires understanding and knowledge of the entire shear history of the material including that of the injection unit.     

Nonisothermal transient filling simulation of fiber suspended viscoelastic liquid in a center‐gated disk

The present study numerically investigates a fiber orientation in injection‐molded short fiber reinforced thermoplastic composite by using a rheological model, which includes the nonlinear viscoelasticity of polymer and the anisotropic effect of fiber in the total stress. A nonisothermal transient‐filling process for a center‐gated disk geometry is analyzed by a finite element method using a discrete‐elastic‐viscous split stress formulation with a matrix logarithm for the viscoelastic fluid flow and a streamline upwind Petrov–Galerkin method for convection‐dominated problems. The numerical analysis result is compared to the experimental data available in the literature in terms of the fiber orientation in center‐gated disk. The effects of the fiber coupling and the slow‐orientation kinetics of the fiber are discussed. Also, the effect of the injection‐molding processing condition is discussed by varying the filling time and the mold temperature. POLYM. COMPOS., 2011. © 2010 Society of Plastics Engineers     

Some relationships between injection molding conditions and the characteristics of vitrified molded parts

The various mold filling phenomena influencing the characteristics of fabricated parts are surveyed. The phenomena leading to jetting in injection mold filling are considered. These are associated with the magnitude of swell by the melt as it exits the gate into the mold. Special attention is given to the influence of non-isothermal runner flow. A theory of extrudate swell of polymer melts with temperature profiles is developed using Tanner's unconstrained recovery theory. In the. absence of jetting, mold filling by a simple advancing front takes place. The hydrodynamics of the advancing front and the stress fields in the flowing melt are determined. Analysis and modeling are presented based on the use of hydrodynamic lubrication theory involving a solid layer along the mold wall and a hot isothermal melt core. This theory is compared with experimental measurements of pressure losses in mold filling. The development of birefringence in injection molding processes is analyzed. Birefringence distributions are due to frozen-in flow birefringence. A new experimental study is presented and its results compared with theoretical predictions. The problem of thermal stresses in injection molded parts is considered.     

Effects of nozzle length on mold filling simulation accuracy

A common assumption in mold filling analysis is that the molding machine is capable of providing a fully pressurized uniform temperature melt to the rear orifice of the sprue. Based upon this assumption, the finite element model then only represents the geometry of the mold itself from the sprue rear orifice forward. This assumption, however, is acceptable only when the injection modeling machine nozzie extension is of such a length that the pressure drop in the nozzle is negligible. For large injection molded parts the exclusion of the molding machine nozzle in the design phase of the mold filling analysis may produce overly optimistic results. This study of an actual large (7.9 kg) injection molded thin wall (4 mm) vinyl part will show the effects of nozzle length on the accuracy of mold filling analysis software predictions. Analytical results with and without representation of the molding machine nozzle in the finite element model will be compared to production trial results.     

A method for determining the flow front velocity of a plastic melt in an injection molding process

During the filling phase of an injection molding process, the flow front velocity of the plastics melt has a decisive influence on the form part quality. It has been believed that a constant flow front velocity of the melt leads to distortion‐free and residual stress‐free form parts. A process control strategy based on a constant flow front velocity of the melt, however, requires the full understanding of the flow front position as a function of the screw position of the injection molding machine. With current methods, this can only be achieved by direct measurements using a number of sensors inside the mold, which leads to complicated structure, great efforts, and high cost for the tooling equipment. This article proposes, designs, and develops an innovative method for determining the flow front velocity of a plastic melt in an injection molding using only one pressure sensor at the front of the screw and based on the idea of mapping a simulated filling process to a real injection molding process. The mapping ensues that the characteristic event points are identified and matched for both the simulated and real filling process. The results of the simulation analysis and experimental evaluation show that the proposed method can be used to determine the flow front position and the resulting flow front velocity of the melt within the cavity of the mold and provide evidence that the new method offers great potential to process control strategies based on machine independent parameters. POLYM. ENG. SCI., 59:1132–1145 2019. © 2019 Society of Plastics Engineers     

Observation of instabilities in flow front during micro injection molding process

This work has investigated filling characteristics during injection molding of a thin‐plate with microchannel patterns on it. To conduct molding experiments, an injection mold and mold inserts with microchannel patterns were designed and built. The experiments were conducted for three different channel sizes, and the results are analyzed. An unstable filling flow was observed in microscopic images for the smallest channel size. The front advanced with continuous creation and merge of microballs, which looked like teeth on the gum. We have analyzed this phenomenon and argued that there exists a critical channel width where such instability takes places. POLYM. ENG. SCI., 2010. © 2010 Society of Plastics Engineers     

Experimental investigation and numerical simulation of injection molding with micro‐features

Injection molding of thin plates of micro sized features was studied in order to manufacture micro‐fluidic devices for bioMEMS applications. Various types of mold inserts—CNC‐machined steel, epoxy photoresist, and photolithography and electroplating produced nickel molds—were fabricated and tested in injection molding. The feature size covers a range of 5 microns to several hundred microns. Issues such as surface roughness and sidewall draft angle of the mold insert were considered. Two optically clear thermoplastics, PMMA and optical quality polycarbonate, were processed at different mold and melt temperatures, injection speeds, shot sizes, and holding pressures. It was found that the injection speed and mold temperature in injection molding greatly affect the replication accuracy of microstructures on the metal mold inserts. The UV‐LIGA produced nickel mold with positive draft angles enabled successful demolding. Numerical simulation based on the 2D software C‐MOLD was performed on two types of cavity fillings: the radial flow and the undirectional flow. The simulation and experimental data were compared, showing correct qualitative predictions but discrepancies in the flow front profile and filled depth.     

Graphene coating assisted injection molding of ultra‐thin thermoplastics

High strength light weight parts are critical for the development of new technologies, particularly electronic devices, such as laptop computers, smart phones, and tablet devices. Injection molded plastics and composites are excellent choices for mass producing such parts. As the part thickness decreases from traditional injection molding (>2 mm thickness) to thin wall molding (～1 mm thickness), and lastly, to ultra‐thin wall molding (<0.5 mm thickness), avoiding incomplete filling (short shots) becomes more challenging. Even though, methods exist today for molding thin‐wall plastic parts (i.e., fast heating/fast cooling injection molding), they require multiple steps resulting in a noncost efficient process. In this article, we demonstrate the technical feasibility of using graphene coating to facilitate flow, by promoting slip at the mold walls. We evaluate the influence of coated and uncoated mold inserts on fiber orientation. We present experimental results using un‐reinforced polypropylene and a 40% by weight carbon fiber reinforced polycarbonate/acrylonitrile butadiene styrene. POLYM. ENG. SCI., 55:1374–1381, 2015. © 2015 Society of Plastics Engineers     

A model of advancing flow front in RIM

A two-dimensional model is developed for mold filling in Reaction Injection Molding using a Petrov-Galerkin finite-element method with free surface parameterization. Dependence of viscosity on the conversion and temperature is represented by the Castro-Macosko function. The model predicts the velocity, pressure, temperature, and conversion distributions with time during the filling stage of a rectangular mold. No a priori assumptions are made regarding the shape of the advancing flow front, or regarding any variables in the flow front region. The accuracy is further improved by using the Petrov-Galerkin formulation, rather than the Galerkin formulation. The results are presented for well-characterized polyurethane systems, for which reliable experimental data is available. The predictions of the model for the pressure rise are in excellent agreement with the experimental data (1) even close to the gel point. These refined predictions are expected to assist in estimating fiber orientation and bubble growth in the final RIM parts, in which the flow front region plays the most important role. Characterization of polyurethane/polyurea and polyurea materials is underway, and will be subsequently incorporated in the model.     

Effect of polymer properties on the structure of injection-molded parts

In this study, the distributions of both molecular orientation and crystallinity along the flow direction as well as across the thickness direction of injection-molded specimens of poly(ethylene terephthalate) (PET) molded at different mold temperatures were investigated. The degree of molecular orientation at the surface of the specimens was compared with that of other injected materials (polystyrene, high density polyethylene, liquid crystal polymer) showing different thermal, rheological, and crystallization characteristics. It was found that the molecular orientation at the skin layer of the molding increases with the polymer relaxation time, the rigidity of the polymer molecules, and the crystallization rate of the polymer. Moreover, in the case of PET, it was found that the crystallinity at the skin layer and in the core of the molding depends on the mold temperature. For low mold temperatures, near the gate, the maximum of crystallinity was observed at the subskin layer because of the “shear-induced crystallization” generated during the filling stage. On increasing the mold temperature, the maximum of crystallinity was found to shift to the skin layer as a result of the decrease of the thickness of this layer. For low mold temperatures, the variation of the molecular orientation in the thickness direction was found to be much the same as for the crystallinity of the polymer. This result indicates that the shear-induced crystallization process improves the degree of molecular orientation in the flow direction since it inhibits the relaxation process of the polymer molecules.     

Prediction of flow–induced process variables based on similarity analysis in the liquid molding process

In general, a numerical scheme is a widely accepted technique for estimating resin flow in the liquid molding process. A numerical mold filling analysis is essential to optimize the manufacturing process of a composite. However, finding an optimal condition from the numerical analysis requires many numerical calculations. The efforts can be greatly reduced if a similarity solution replaces the repeated numerical calculations. In this study, similarity relations are proposed to predict the flow‐induced process variables. such as resin pressure, resin velocity, and flow front evolution time, during mold filling. Numerical simulations are performed for two cases where a material property, an injection condition or a part shape is different. The model is verified by applying the similarity relation for two numerical results obtained from the thin shell structure.     

Glass fiber-filled thermoplastics. II. Cavity filling and fiber orientation in injection molding

Mold filling of a rectangular cavity of three different thick nesses fed from a reservoir is studied for unfilled and glass fiber-filled polypropylene and polystyrene. The shapes of flow fronts studied by short-shots are affected predominantly by the thickness of the cavity with other parameters playing a less important role. Pressure drop versus volumetric flow rate inside the thinnest cavity is studied experimentally and predictions are made from a computer simulation of mold filling. The orientation of fibers in the cavity is examined using a reflect-type microscope and the orientation is found to depend on cavity thickness, melt temperature, fiber content, and to a lesser extent, on volumetric flow rate. In the thinnest cavity, where the flow is quasi-unidirectional, the fibers remain in the plane of flow oriented either along the flow direction or perpendicular to it, except in the region near the flow front, where they follow a “fountain” flow behavior.