Current Research Projects
As high technology applications increase the demand for micro sized parts in the field of polymers, there is also an increasing demand for micropellets as raw material. Those micropelletized thermoplastics can be used for accurate sintering processes and also for enhancing common processes of thermoplastics like extruding and injection molding. So far underwater pelletizing processes are most commonly used to manufacture micropellets. However, they are limited in pellet size and quality. The goal of this research work is to optimize a technique to manufacture uniform shaped spherical micropellets of a unit size that uses the physical method of Rayleigh Disturbances. The phenomenon of Rayleigh Disturbances describes the breakup of a liquid stream into droplets if jetted into another fluid of different density and viscosity. The research work includes design improvements of the process, experimental work with different resins and simulation of the phenomena observed during processing.
Rayleigh disturbances on PE-HD particles.
To produce a high performance long-fiber reinforced composite sailboat from bio-based materials via vacuum resin infusion aided by finite computational fluid dynamics. This testing would aid in the proper determination of key properties needed for simulation purposes. Composites are combinations of fibers and resins in a certain proportion. The fibers act as reinforcement to enhance the strengthening of the composite part, while the resin matrix gives the final shape of the composite part. Composites have many advantages over metals and other traditional materials, like wood, in the manufacturing of boats. Since the traditional composites have poor recyclability and especially are subjected to pyrolysis at the end-of-life, the biodegradability and recyclability of composites need to be enhanced. The aim of the project is to investigate the possibility of replacing the traditional synthetic materials used in boat manufacturing with bio-based resins and natural fibers by analyzing their rheological, mechanical and thermal behavior.
Commonly, a perfectly random initial orientation is assumed for fiber orientation simulations for LFT compression. However, due to the plastification process in the extruder, a distinctive non-random fiber orientation emerges within the strand. The fiber orientation distribution within a D-LFT strand was measured using micro-computerized tomography (µCT). With a µCT data processing software package (Volume Graphics StuidoMAX 2.0), the local fiber orientation could be quantified. A 3D mold filling simulation (Moldex3DTM) that considers the initial orientation of the D-LFT strand is used in this project. The results of the fiber orientation distribution of the finished part are compared with results based on the assumption of a perfectly random initial fiber orientation, so the impact of the initial orientation distribution within the D-LFT strand can be evaluated. Current sub-projects are:
- Measuring rheological properties of LFT-Materials and evaluate the impact of fiber orientation and fiber length distribution on viscosity using the sliding plate rheometer.
- Developing and improving a mapping tool in order to implement the fiber orientation distribution into Moldex3D.
- Design a die for a mini-extruder to produce LFT charges with a unidirectional orientated and perfectly dispersed fibers.
- Processing µCT of experiments to validate fiber orientation simulations.
- Using Moldex3D to predict fiber orientation within different parts.
Cross-sectional view of the scanned D-LFT strand. Shows that the assumption of a perfectly random orientation does not hold.
Simulation results for fiber orientation. Initial fiber orientation within the strand leaves an obvious footprint.
Warping of Fiber-Reinforced Polymeric Matrix Composites. Case Study: Spring-Forward of Carbon Fiber-Epoxy Prepregs Processed via Oven Curing in Vacuum Bagging
The increase of residual stresses during the phase change of thermoset fiber-reinforced composites causes unintended distortions. A specific situation of interest is the angular distortion in curved fiber-reinforced thermoset parts, also referred to as the spring-forward effect or anisotropy induced curvature change. It is our aim to predict and quantify the effect of thermal expansion mismatch, phase changes and cure shrinkage in the distortion of fiber-reinforced composites through numerical technics. This prediction will allow us to do reverse engineering by quantifying how much a mold must be distorted to obtain the expected profile.
Some of the current numerical and analytical models neglect the effect of phase change in the distortion. This study extends the current models to involve the cure time-temperature-diagram (TTT) of thermoset resins. An extensive thermal and rheological experimental study accompanies and validates the simulation technique.
Simulation and experimental result of the case where the fibers are oriented alternating with the inner most fibers oriented in the direction 1. This produced the most complex geometry with a spring-forward effect along with a saddle effect that forces the profile outward.
The drapability of a fabric has to do with its capacity to be deformed from a flat stage to the shape of a three dimensional mold without the formation of wrinkles or creases. In order to analyze the draping of a textile it is important to characterize its behavior when subjected to in-plane shear, out of plane bending and tension.
These behaviours can be measured by means of diffferent methods:
Picture Frame test:
The picture frame test is used to determine the textile shear behavior. In order to accomplish this purpose, the sample is pinned to a frame which prescribes a uniform shear deformation field. The whole fixture is loaded in a tensile test machine allowing an accurate measurement of force and displacement.
Cantilver Bending Test:
The cantilever bending test is used to determine the bending behavior of a textile. In order to accomplish this purpose, a sample is clamped in a cantilever configuration and allowed to deform as a result of its own weight. The fabric shape is recorded via photography for different lengths.
Selected Former Research Projects
In the human body, if a bone is damaged and a critical amount of bone is lost, the body cannot repair itself, and normally a transplant is needed. A three-dimensional scaffold can provide the structural support and mechanical strength necessary to encourage cell growth, particularly in terms of bone tissue engineering. Synthetic, biodegradable polymers are preferred materials because they have minimal foreign body reaction, good processability, and superior mechanical properties. Polylactide (PLA) and its copolymers have been used extensively in research on bone repair as well as many other tissues, and have already been approved for use in humans by the Food and Drug Administration (FDA).
Many methods for the fabrication of PLA scaffolds have been investigated, such as solvent casting/particulate leaching, supercritical-gas injection molding, fiber bonding, phase separation, and melt molding. These methods may use harsh organic solvents that are detrimental to cell growth, or have only limited control over the final porosity and interconnectivity of the scaffold. Recently, however, there have been promising developments in the use of sintering. In the sintering process, polymer particles are heated to temperatures slightly above the melting point, and they fuse together slowly, maintaining a porous structure. PLA scaffolds of sufficient mechanical strength and porosity have been fabricated through this method.
Our work seeks to improve the processing of PLA scaffolds to optimize strength, stiffness, and pore size and interconnectivity, which all play important roles in encouraging bone growth. The bulk scaffold stiffness is measured with a precision compression testing instrument. Recent work seeks to use micro-Computed Tomography (μ-CT) testing to evaluate porosity, pore size, and pore interconnectivity. The μ-CT instrument uses X-rays to construct a three-dimensional image of the sample. Future work will use μ-CT data to create finite element models of the scaffolds. We hope to connect heat transfer models and particle size with the final properties of the scaffolds through simulation.
researcher: Tom Mulholland, Katerina Sanchez
The objective of this project is the development and validation of computational models to simulate the behavior of fiber reinforced composite parts during the manufacturing process. Current approaches are based on simulation as well as trial and error techniques. However, in order to properly deal with fiber damage, fiber jamming and fiber-matrix separation, a comprehensive understanding of the physics behind fiber motion is required. Within this project, the fibers are modeled using a mechanistic approach, where their structure is represented as flexible chains composed of a combination of springs and beads or cylinders.
Simulation results of the project will include final fiber orientations, fiber attrition and fiber density distributions within a molded part. This will shed light on phenomena that are not well understood, such as fiber matrix separation (characterized by ribs and features with low fiber content) and the existence of a fiber free skin region observed in some injection molded parts. In the future, the tools developed in this project will help the process engineer to predict potential defects and optimize the properties of a part before a mold is actually made. With a higher level of understanding of fiber motion phenomena during molding, it will eventually be possible to mass produce polymer composite parts with higher quality and controlled properties.
Additionally, an experimental set-up capable of visualizing the fountain flow effect has been developed. The device consists of a transparent sleeve filled with silicon oil and a stationary plunger. A cluster of fibers is positioned inside the sleeve near the plunger at the beginning of the experiment. As the sleeve moves downwards, a fountain flow develops near the free surface. In this way, the trajectory of the fibers as they move through the front region can be observed and studied. This experimental setup (and others in development) in conjunction with the study of real parts will help to explain the phenomena of fiber matrix separation and fiber attrition and allow the validation of the computational model.
In the following video, the movement of a group of fibers in a fountain flow velocity field can be visualized. In this simulation, the observer moves with the average velocity of the flow and the mold walls move downwards. Therefore, the front of the flow remains stationary. So far, the simulations developed in this project predict the existence of fiber free regions at the mold surface showing good agreement with the experimental observations reported in the literature.
researcher: Daniel Ramirez, Sebastian Kollert
The properties of thermoplastic parts and especially the shrinkage are determined by the material as well as the processing parameters and conditions. The parameter temperature and pressure play a central role. Among other things, they govern the flow properties, the cooling conditions and especially the solidification behavior. However, the pressure dependence as well as the effect on the process are seldom investigated. Due to the lack of experimental data, they are mostly neglected in simulations. Therefore, a significant deviation between experimental and predicted dimensions are found, especially for thick parts.
In this project, the known models should be broadened by the factor pressure. Thereto new insights on the pvT behavior of amorphous plastics are used and implemented in the injection molding process for the first time. The major goal is the improvement of the shrinkage prediction of thick plastic parts, like e. g. optical lenses. Hereby, the manufacturer should be enabled to adapt the part geometry during the design process and produce high presicion parts in the subsequent production process.
researcher: Dr.-Ing. Natalie Rudolph
During single-screw extrusion, shear rates may reach 200 [s-1] in a screw channel near a barrel wall, where the values are higher between flight tips and the barrel. At too high of a shear rate, instabilities such as melt fracture can occur. Two numerical techniques were separately and successfully applied to capture viscoelastic flows and each were used to model flows during extrusion. The meshless Radial Functions Method (RFM) was implemented to simulate flow patterns in two and a half dimensions (2.5D), and correctly predicts secondary flows in fully developed non-circular ducts. Viscoelastic secondary flows in full 3D non-circular ducts were then simulated using a Finite Volume Method (FVM) approach with single and multi-mode viscoelastic models. Results are in excellent agreement with industry experiments as well as numerical results using both the RFM and a finite element method.
Numerical simulations of full 3D viscoelastic entry flows were performed using FVM with a stress-splitting technique through a planar abrupt contraction test geometry. Both single and multi-mode Phan-Thien Tanner and Giesekus shear-thinning models were implemented to reproduce full 2D and 3D flows through a 4:1 contraction, where considerably high Weissenberg numbers were successfully simulated. Results obtained within this work show excellent qualitative agreement with experimental observations and simulation results found in literature.
Exploratory FVM simulations are being carried out which began from an unwrapped screw channel geometry and has now developed into modeling a full 3D rotating single screw extruder under isothermal conditions. The highly non-linear single-mode and multi-mode viscoelastic Phan-Thien Tanner and Giesekus shear-thinning models are able to capture the complex nature of polymer relaxation time under high Weissenberg conditions. This is an important beginning toward more accurate modeling and simulating flow for "process improvement through design" of an industrial extruder.
researcher: Lori Holmes
Fiber reinforced thermosetting materials are often used to manufacture automotive body panels as well as for under-the-hood automotive applications, household goods, breaker switch boxes in the electrical industry, etc. These materials are chosen because of their high strength, light-weight, heat resistance and electric properties. There are various projects we are currently working on which deal with these types of materials. Currently, of interest to us are fiber matrix separation during processing and the curing of thick thermosetting parts.
During processing of fiber reinforced polymer parts a main assumption has always been that there is a constant fiber density throughout the part. However, burn-out tests on various parts have revealed that there is a fiber density distribution throughout the part. In fact, some regions contain half the fibers as other regions in the same part. This is especially true for parts that are ribbed, or have regions that are difficult to access by the reinforcing fibers. A varying fiber density distribution throughout the part not only leads to anisotropies but also to surface waviness in large thin parts such as automotive body panels. Within this research project we are studying the causes of fiber matrix separation through modeling and experimental work. The ultimate goal of this project is to predict and control fiber density distributions throughout a part.
For many thermosetting articles the thickness can be large enough that the curing process during manufacturing takes place in a non-uniform fashion, leading to severe residual stresses and in many cases under-cured regions in the final parts.
This portion of the project addresses four major issues:
- Material characterization using differential scanning calorimetry tests
- Cure kinetic model development and fitting using the experimental DSC data
- Implementation of the curing models into our own 3D finite element heat transfer programs
- Couple 3D heat transfer curing program with a stress-strain FEM program, to predict residual stresses and properties within the finished product.
Mechanical properties of thermoplastic and thermosetting polymers are improved when they are reinforced with fibers. Fibers will increase the strength, stiffness and impact resistance. These materials can easily be processed using conventional processing techniques such as injection and compression molding, as well as extrusion. However, fiber damage or attrition will shorten the length of the reinforcing fibers during processing. Shorter fibers will diminish the mechanical effectiveness of the composite, and in some cases will result in properties that are inferior to the un-reinforced polymer's properties. Fiber damage during processing can be attributed to at least the following mechanisms:
- Deformation of the polymer melt,
- Fiber-fiber interactions during flow, and
- Fiber-equipment interactions.
Within this study, analytical and numerical models are developed and used to assess the significance of each individual mechanism. In addition, controlled experimental set-ups are built to study the various mechanisms and for model verification. Results from this study will lead to a full understanding of fiber attrition during processing. This will aid the process engineer to design a process, i.e. mixing devices, screw geometry, gates, or charge locations, that will lead to fiber composites with ideal fiber length distributions.
Within this project, a non-isothermal, non-Newtonian flow simulation for polymer melts using the boundary element method (BEM) is being developed to model systems such as mixing phenomena of polymer blends inside complex three dimensional mixing devices. In addition, a non-linear stress-strain simulation for polymer components will be developed. This includes the coupling of heat transfer and momentum equations for polymer flows well as momentum balance of solid components. The boundary element method is a numerical technique that lends itself to model flow, heat transfer and stress-strain behavior of a system with only a surface description of the 3D geometry, requiring no need of a mesh to represent the volume of that geometry. This attribute allows us to model flow, heat transfer and stress-strain behavior in complex geometries, and in systems with constantly changing boundary shapes. Such systems may include the flow inside starved or partially filled mixers, where the solid and free surfaces constantly change shape. For such systems the mesh generation of finite element method (FEM) programs, required for geometry representation during a processing cycle, would be prohibitive. Projects that are currently being worked under this general research area are:
- Drop deformation automotive painting processes,
- Metal flake orientation predictions during automotive painting processes,
- Study of fiber damage mechanisms during processing of reinforced polymer melts,
- Simulation of mixing processes during polymer processing.
Particle tracking in a twin-screw extruder with the boundary element method
Based on a recent literature survey conducted by us, existing experimental apparati and techniques designed to study the melting and mixing behavior of polymers inside the screw extruder suffer from limited functions or tedious procedures. This study targets an in-line, non-invasive measuring device that can gather experimental data with reasonable accuracy and at short response times with a highly instrumented single-screw extruder with built-in sensors and modular screw sections. The proposed system includes a fully instrumented 45-mm production scale single-screw extruder capable of running under industrial conditions. In addition, the modular sections of the screw will be assembled to form various screw designs and constructions for characterizing their effectiveness. In addition to the series of pressure transducers and thermocouples mounted on the barrel wall, special optical, infrared, ultrasonic, and dielectric sensors will be tested with the instrumented extruder. By sensing the difference in properties between the melt and the solid phase, this versatile apparatus will deliver real-time information revealing the polymer melting behavior without adding thermal history to the polymer as is done with the screw extraction approach. Anticipated benefits are:
- To understand the melting and mixing behavior of polymers inside the screw extruder and validate and improve the predictive tool for extrusion process.
- To establish the know-how for optimizing the screw design, controlling the process, and enhancing the part quality and throughput for extruded products.
- Current Projects
- Manufacturing of Micropellets using Rayleigh Disturbances
- Resin Infusion of Natural Fibers. Rheological and Thermal Characterization of Linseed based Epoxy
- D-LFT Compression Molding Simulation
- Warping of Fiber-Reinforced Polymeric Matrix Composites. Case Study: Spring-Forward of Carbon Fiber-Epoxy Prepregs Processed via Oven Curing in Vacuum Bagging
- Drapability of Fabrics
- Former Projects
- Sintering and Characterizing PLA Bone Tissue Engineering Scaffolds
- Study of the Fiber-Matrix Separation and Fiber Jamming During Processing of Fiber Filled Composites
- Improvement of process and shrinkage simulation with pressure dependent material data
- Modeling Three Dimensional Viscoelastic Secondary Flows in Extrusion
- Processing of Reinforced Thermosetting Polymer Parts
- Process Induced Fiber Damage
- Boundary Element Simulations in Polymer Engineering
- In-Line Measurement of Polymer Melting Behavior in Single-Screw Extruders