No description available.

No / The study is based on a previously proposed methodology for multiaxial fatigue life assessment of injection-molded components (called ‘Through Process Modeling’ (TPM)). The present contribution focuses on stress concentration effects induced in notched samples. Purely macroscopic approaches are unable to capture the different mechanical responses of variably injected parts with the same shape. A high interest of the present method is to take into account the difference of fiber orientation resulting from the process. After briefly reminding the TPM method, it will be shown that good lifetime estimations are obtained for laterally injected samples, from a fatigue criterion identification based on longitudinally injected ones.

No description available.

No description available.

With this research, we work towards the overall objective of being able to accurately simulate fiber orientation in complex flow geometries of composite fluids of industrial significance. The focus of this work is to understand the rheological behavior of these materials and its connection to fiber orientation as determined in simple shear flow. The work includes the development of a novel approach to characterizing the transient rheology; an experimental study of the relationship between the stress growth functions in startup of flow and the fiber orientation; a critical assessment of the limitations of current fiber suspension theory; and an approach to determining unambiguous model parameters by fitting. A key difference between the rheological studies performed in this work and others is the use of a cone-and-plate device combined with "donut" shaped samples (CP-D) to prevent boundary effects on the measurement. The conventional method for obtaining transient rheological data is to use parallel disk (PP) geometry set at a gap where the measurements are independent of disk spacing. However, this work suggests that the inhomogeneous velocity gradient imposed by the PP geometry induces excessive fiber-fiber contact contributing to exaggerated measurements of the stress growth functions. An experimental study of the transient rheological behavior of a 30 wt% short glass fiber-filled polybutylene terephthalate was performed using the CP-D. Stress growth measurements during startup of flow were performed in combination with direct measurement of the fiber orientation to determine the relationship between the transient rheology and the fiber microstructure. The well defined fiber orientation and rheological experiments allowed for a quantitative assessment of current fiber suspension theory. Comparison between the experimental fiber orientation and predictions based on Jeffery's equation and the Folgar-Tucker model show that the fiber orientation evolves much slower than predicted. In addition, the addition of a "slip" term improved the agreement between the predictions and experimental results. Predictions using the Lipscomb model coupled with the Folgar-Tucker model, with slip, were fit to the transient stresses to determine the feasibility of fitting unambiguous model parameters for a specific composite fluid. Model parameters determined by fitting at a shear rate of 6 s-1 allowed for reasonable predictions of the transient stresses in flow reversal experiments at all the shear rates tested. / Ph. D.

Long fiber reinforced thermoplastic composites are used for creating lightweight, but mechanically sound, automotive components. Injection molding is a manufacturing technique commonly used for traditional thermoplastics due to its efficiency and ability to create complex geometries. Injection molding feedstock is often in the form of pellets. Therefore, fiber composites must be chopped for use in this manufacturing method. The fibers are cut to a length of 13 mm and then fiber attrition occurs during processing. The combination of chopping the fibers into pellets and fiber breakage creates a distribution of mostly short fiber lengths, with some longer fibers remaining. Discontinuous fiber reinforcements are classified as long for aspect ratios greater than 100. For glass fibers, that distinction occurs at a length of 1 mm, and for carbon fibers 0.5 mm. Traditional composite materials and manufacturing processes utilize continuous fibers with a controlled orientation and length. The use of chopped discontinuous fibers requires a method to predict the orientation of the fibers in the final molded piece because mechanical properties are dependent on fiber length and orientation. The properties and behavior of the flow of a fiber reinforced polymer composite during molding are directly related to the mechanical properties of the completed part. Flow affects the orientation of the fibers within the polymer matrix and at locations within the mold cavity. The ability to predict, and ultimately control, flow properties allows for the efficient design of safe parts for industrial uses, such as vehicle parts in the automotive industry. The goal of this work is to test material characterization techniques developed for measuring and predicting the orientation of fiber reinforced injection molded thermoplastics using commercial grade long carbon fiber (LCF) reinforced nylon 6,6 (PA 6,6). Forty weight percent LCF/PA 6,6 with a weight averaged fiber length of 1.242 mm was injection molded into center gated disks and the orientation was measured experimentally. A Linux based Numlab flow simulation process that utilizes the finite element method to model the flow and orientation of fiber reinforced materials was tested and modified to accurately predict the orientation for this composite and geometry. Fiber orientation models used for prediction require the use of empirical parameters. A method of using non-lubricated squeeze flow as an efficient way to determine the strain reduction factor, , and Brownian motion like factor, CI, parameters for short glass fiber polypropylene orientation predictions using the strain reduction factor (SRF) model was extended to use with the LCF/PA 6,6 composite. The 40 weight percent LCF/PA 6,6 material was compression molded and underwent non-lubricated squeeze flow testing. The flow was simulated using finite element analysis to predict the fiber orientation using the SRF model. The empirical parameters were fit by comparing the simulated orientation to experimentally measured orientation. This is a successful method for predicting orientation parameters that is significantly more efficient than optimizing the parameters based on fitting orientation generated in injection molded pieces. The determined orientation parameters were then used to reasonably predict the fiber orientation for the injection molded parts. The authors proved that the experimental and simulation techniques developed for the glass fiber reinforced polypropylene material are valid for use with a different, more complex material. / Doctor of Philosophy / Fibers reinforce thermoplastic polymers to create lightweight, but mechanically sound, automotive parts. Thermoplastics flow when heated and harden when cooled. This work compares two of the commonly used thermoplastics, polypropylene (plastic grocery bags, food storage containers) with a glass fiber reinforcement and a form of nylon called PA 6,6 with a carbon fiber reinforcement. Injection molding is a manufacturing technique commonly used for un-reinforced thermoplastics due to its efficiency and ability to create complicated shapes. Injection molding feedstock is often in the form of pellets. Therefore, fiber composites must be chopped for use in this manufacturing method. The fibers are cut to a length of 13 mm and then fiber breakage occurs in the injection molder. The combination of chopping the fibers into pellets and fiber breakage creates a range of lengths. This distribution consists of mostly short fiber lengths, with some longer fibers remaining. Discontinuous fiber reinforcements are classified as long for aspect ratios (the ratio of length over diameter) greater than 100. For glass fibers, that distinction occurs at a length of 1 mm, and for carbon fibers 0.5 mm. Traditional composite materials and manufacturing processes utilize continuous fibers with a controlled orientation and length, such as the weave pattern one might see in a carbon fiber hood. The use of chopped fibers requires a method to predict the orientation of the fibers in the final molded piece because mechanical properties are dependent on fiber length and orientation. The way that the plastic flows during molding is directly related to the mechanical properties of the completed part because flow affects the way that the fibers arrange. The ability to predict, and ultimately control, flow properties allows for the efficient design of safe parts for industrial uses, such as vehicle parts in the automotive industry. The goal of this work is to test the techniques developed for measuring and predicting the orientation of fiber reinforced injection molded thermoplastics using commercial grade long carbon fiber (LCF) reinforced nylon 6,6 (PA 6,6). LCF/PA 6,6 with an average fiber length of 1.242 mm was injection molded into a disk and the orientation was measured experimentally. A computer flow simulation process that utilizes the finite element method to model the flow and orientation of fiber reinforced materials was tested and modified to accurately predict the orientation for this composite and geometry. Fiber orientation models used for prediction require the use of parameters. There is no universal method for determining these parameters. A method of using non-lubricated squeeze flow as an efficient way to determine the parameters for short glass fiber polypropylene orientation predictions was extended to use with the LCF/PA 6,6 composite. The LCF/PA 6,6 material was compression molded and underwent non-lubricated squeeze flow testing. The flow was modeled to predict the fiber orientation. The empirical parameters were fit by comparing the simulated orientation to experimentally measured orientation. This is a successful method for predicting orientation parameters. The determined orientation parameters were then used to reasonably predict the fiber orientation for the injection molded parts. The authors proved that the experimental and simulation techniques developed for the glass fiber reinforced polypropylene material are valid for use with a different, more complex material.

Several active methods, which requires external control systems and moving parts, have been developed to control the fiber orientation during 3D printing. Active mechanisms like rotating nozzle, impeller, and magnetic field have been integrated to realize complex internal fiber structures. In this study, instead of using active methods, I investigate a passive method for controlling the fiber orientation without any moving parts or additional mechatronics added in the printing process. Composites of polydimethylsiloxane (PDMS) and glass fibers (GF) are 3D printed. Channels, such as helicoid, are designed and integrated to guide the ink flow and passively result in different pre-alignment of fibers before the ink flow into narrow nozzle space. While passing through the designed channels, the fibers orient due to the shear between channel walls and the ink. The effect of helicoids with different pitch sizes are investigated via mechanical experiments, microstructural analysis, and numerical simulations. The results show that both surface to volume ratio and helix angle of the channel affect pre-alignment of fiber orientation at the entry of nozzle. The internal fiber structures lead to enhanced and tunable mechanical properties of printed composites. Pitch size 7-9 mm (helix angle of 7.92- 10.15o) is found to be optimal for the 3D printed PDMS-GF composites. Stiffness and strength can be tuned up to 77.6% and 47.8%, respectively, compared with the case without helicoid channel. Channels of parallel holes, parallel holes with taper end and gradually changing pitch helicoids are experimentally tested, showing further enhancement in mechanical properties.

<p>Extrusion Deposition Additive Manufacturing is a process by which short fiber-reinforced polymers are extruded in a screw and deposited onto a build platform using a set of instructions specified in the form of a machine code. The highly non-isothermal process can lead to undesired effects in the form of residual deformation and part delamination. Process simulations that can predict residual deformation and part delamination have been a thrust area of research to prevent the repeated trial and error process before a useful part has been produced. However, populating the material properties required for the process simulations require extensive characterization efforts. Tackling this experimental bottleneck is the focus of the first half of this research.</p><p>The first contribution is a method to infer the fiber orientation state from only tensile tests. While measuring fiber orientation state using computed tomography and optical microscopy is possible, they are often time-consuming, and limited to measuring fibers with circular cross-sections. The knowledge of the fiber orientation is extremely useful in populating material properties using micromechanics models. To that end, two methods to infer the fiber orientation state are proposed. The first is Bayesian methodology which accounts for aleatoric and epistemic uncertainty. The second method is a deterministic method that returns an average value of the fiber orientation state and polymer properties. The inferred orientation state is validated by performing process simulations using material properties populated using the inferred orientation state. A different challenge arises when dealing with multiple extrusion systems. Considering even the same material printed on different extrusion systems requires an engineer to redo the material characterization efforts (due to changes in microstructure). This, in turn, makes characterization efforts expensive and time-consuming. Therefore, the objective of the second contribution is to address this experimental bottleneck and use prior information about the material manufactured in one extrusion system to predict its properties when manufactured in another system. A framework that can transfer thermal conductivity data while accounting for uncertainties arising from different sources is presented. The predicted properties are compared to experimental measurements and are found to be in good agreement.</p><p>While the process simulations using finite element methods provide a reliable framework for the prediction of residual deformation and part delamination, they are often computationally expensive. Tackling the fundamental challenges regarding this computational bottleneck is the focus of the second half of this dissertation. To that end, as the third contribution, a neural network based solver is developed that can solve parametric partial differential equations. This is attained by deriving the weak form of the governing partial differential equation. Using this variational form, a novel loss function is proposed that does not require the evaluation of the integrals arising out of the weak form using Gauss quadrature methods. Rather, the integrals are identified to be expectation values for which an unbiased estimator is developed. The method is tested for parabolic and elliptical partial differential equations and the results compare well with conventional solvers. Finally, the fourth contribution of this dissertation involves using the new solver to solve heat transfer problems in additive manufacturing, without the need for discretizing the time domain. A neural network is used to solve the governing equations in the evolving geometry. The weak form based loss is altered to account for the evolving geometry by using a novel sequential collocation sampling method. This work forms the foundational work to solve parametric problems in additive manufacturing.</p>

Finite element analysis was implemented to evaluate the transverse shear modulus of a unidirectional glass/epoxy fiber-matrix composite based on pure shear displacement boundary conditions. Unit cells consisting of three-dimensional glass cylinders surrounded in square-cuboid epoxy matrices were modeled to represent "Representative Volume Element" (RVE) configurations in periodic and random-periodic square cell arrangements of variable size. Three RVEs were constructed and analyzed: A single unit cell, a 9-cell (3 x 3) array, and a 25-cell (5 x 5) array. Additionally, the unit cell was modeled to include an interphase. Three sets of cell arrangements were constructed and evaluated: a periodic square array, a transversely distributed random-periodic array, and a variable angularly aligned random-periodic array. Furthermore, scale and free-edge effects of the composites were studied by evaluating the shear modulus in incrementally increasing domains, as well as by isolating finite-sized domains called windows within the multiple-cell model, whereby the window is smaller than the array. Finite element software was subsequently utilized to create a three-dimensional mesh of the composite models studied. Each simulation consisted of exposing the respective domain to pure shear boundary conditions, whereby the model was subject to uniform transverse displacement along its boundary. Subsequent volumetric averaging resulted in computation of the apparent transverse shear modulus. The resulting numerically attained elastic shear modulus was then evaluated and compared to known predictive models in literature. It was shown that that the transverse random arrangement as well the random angular alignment of fibers within the composite structure had a marginal influence on the shear modulus. For random transverse distributions, a deviation in modulus of +1.5% was observed for the 25-cell array as compared to a periodic array of equal size. Similarly, a deviation of +0.3% was predicted for 25-cell arrays subject to random angular fiber misalignments up to ±0.143°, as compared 25-cell periodic arrays. Furthermore, increasing the composite medium by systematic, incremental augmentation model domains was shown to significantly lower the shear modulus in a convergent manner as G23 values dropped 33.5% from the nonhomogeneous single cell to the 9-cell model, and 2.6% from the same 9-cell to the 25-cell model, while observing the effects of a mesoscale window displayed little variance in modulus value as compared to the larger RVE from which the window was isolated from. Lastly, the predictive potential of the model developed by Sutcu for composites with interphases, and other commonly employed models for predicting the transverse shear modulus of unidirectional composites was also evaluated. Numerical results of nonhomogeneous interphase models for both periodic and random-periodic 25-cell arrays were found to be in excellent agreement with Sutcu's approximation. The shear modulus of the 25-cell, nonhomogeneous interphase model was found to lie within 3.5% of Sutcu's prediction. Volume averages for periodic arrays with no interphase were observed to lie in close proximity to Halpin-Tsai's model, displaying a variation of 7% for a 25-cell, single fiber model.

The use of fiber-reinforced polymer composites formed via injection molding is of increasing interest due to their superior mechanical properties as compared to those of the polymer matrix alone. These mechanical properties, however, are strongly dependent on the fiber length and orientation distributions within a molded part. As such, there is a need to understand and model the orientation evolution of chopped fibers in flow in order to accurately simulate the final fiber orientation distribution within injection molded parts. As a result of this, accurate and reliable experimental measurement of fiber orientation is needed. Within this research, the application and validity of the Method of Ellipses for determining the orientation of long, semi-flexible glass fibers within injection molded composites has been investigated. A fiber suspension with an average length of approximately 3.9 mm was the focus of this study and assumed to be representative of commercial distributions. A novel method to quantify fiber curvature was developed and utilized to show that flexibility in center-gated disc and the end-gated plaque samples was minimal on average for the selected fiber length distribution. Thus, it was determined that the Method of Ellipses was applicable when utilized to obtain reliable orientation data for the selected long glass fiber suspension and within the chosen geometries that exhibit 1-, 2-, and 3-dimensional velocity fields. However, a modified image analysis width was found to be necessary in regions of highly aligned fibers, due to the increase in ellipse size and the need to reduce the number of partial objects and thus minimize error. This allowed for a direct comparison of the experimental orientation behavior of short and long glass fibers within the center-gated disc and the end-gated plaque, as well as the effect of the orientation distributions on the global modulus of the part. / Ph. D.