M. Cengiz Altan School of Aerospace and Mechanical Engineering The University of Oklahoma NSF/DOE/APC Workshop Future of Modeling in Composites Molding Processes Materials and Measurements Group June 2004 Current Research and Future Challenges: Materials, Mechanics and Manufacturing Issues in Composites Molding

Current Research Three-Dimensional Features of Void Morphology in Resin Transfer Molded Composites Most void studies characterize void morphology from two-dimensional microscopic images. But different planes show different features of voids in molded composites. We investigate three-dimensional features of void morphology by image analysis of both through-the-thickness and planar surfaces. Relatively large cylindrical voids, observed in cigar shapes on a planar surface, appear only as small irregular or elliptical voids on a through-the-thickness surface. We characterize size, shape, and location of distribution as well as spatial variations in void density. Through-the-thicknessPlanar 100 μm

Research Challenges Three-Dimensional Features of Void Morphology in Resin Transfer Molded Composites How do we get rid of voids? What and how much are we really loosing by the presence of voids? What are the effective void removal or reduction methods? Are sealing the mold, bleeding, or packing good strategies of void removal? Do we need to worry about the voids we cannot see? Should we look for voids under 1  m in size? We need to integrate the models for void transport, deformation, break up, dissolution into micromechanics models to predict composites performance under static and dynamics loading. void size distribution unsealed sealed packed

Current Research Molding of Composites with Nanoclay We investigate the effects of various nanoclay types, nanoclay content, and mixing temperature on nanoclay dispersion in molded epoxy at multiple scales. Methods for dispersion characterization involve image analysis of SEM micrographs (above 1.5  m cluster size) and using Wavelength Dispersive Spectrometry by an Electron Microbeam Analyzer (below 1.5  m cluster size). Nanoclay clusters experience breakdown to smaller sizes possibly due to viscous stresses as verified by the increase in the number of clusters away from the inlet. Significant cluster filtration by glass fiber preform is also observed. This yields nonhomogeneous nanoclay content. We investigate effects of nanoclay content on void morphology throughout the molded composites. Significantly higher void content is observed by increasing nanoclay up to 10%. 2% nanoclay 10% nanoclay Nanoclay powder 500 nm 400 μm

Research Challenges Molding of Composites with Nanoclay Several nanoclay types are commercially available with different surface chemistry and treatment. All have different dispersion characteristics. Nanoclays are relatively cheap (at least compared to Carbon Nanotubes) and are being used in commercial products, primarily in thermoplastics. Can we also use them in RTM composites? What are the benefits? (a) Improved thermal response, higher Tg, etc. (b) Improved stiffness and strength, etc. Obtaining meaningful improvements has been problematic. We need to achieve improvement over the composite properties not just those of epoxy. Exfoliation of clay platelets seems to be needed. What are the effective dispersion and exfoliation strategies? (a) mechanical mixing, high shear rates help but may not be adequate; (b) acoustic methods, sonification; (c) processing at elevated temperatures. Different rheological behavior during mixing and molding is expected for different nanoclay content. Models for cluster break up, rheology of the nanoclay mixture and constitutive models for the molded nanoclay composite will be helpful.

Current Research Design of Molds for Reduced Molding Pressure & Increased Fiber Content How can bigger, better, and cheaper RTM composites be fabricated? One aspect towards that goal is the reduction of molding pressure. We study the effect of flow channels machined on the top mold surface. Complex fill patterns and significant reduction in inlet pressure are observed. Mold surfaces with 2 or 4 channels (4 channels are shown on the mold above) are studied. Channel length is also varied up to the total disk radius (shown as 100% in the figure). Figure above shows 70% reduction in the inlet pressure if the flow channels are as long as the disk radius. Top surface Bottom surface Inlet Pressure Measurements E-Glass/Epoxy

Research Challenges Design of Molds for Reduced Molding Pressure & Increased Fiber Content Pressure is significantly reduced with flow channels but the fill pattern becomes more complicated. Formation of large dry spots must be avoided. Yet, various combinations of channel length and fiber volume fraction yield such defects even in a geometrically simple center-gated composite disk. Impregnation through-the-thickness becomes important, which requires more sophisticated flow simulation models than commonly used today. Accurate characterization of the preform permeability tensor is needed to develop effective predictive tools for the mold design. Feasibility of using flow channels must be assessed for larger parts with more complicated shapes. Top surface Bottom surface

Current Research Moisture Absorption and Desorption of Molded Composites Effects of Temperature, Pressure, Cyclic Loading, Fiber Content We investigate moisture absorption and desorption behavior of molded composites. Current methods of measuring diffusivity are based on the initial slope of the mass gain curve and the maximum moisture absorbed. Such methods yield inaccurate diffusivity estimates for anisotropic composites. Better methods are needed. Effects of fiber content, thermal environment, and loading history on the moisture uptake are studied. Anisotropic moisture absorption model for a three-dimensional composite sample

Research Challenges Moisture Absorption and Desorption of Molded Composites Effects of Temperature, Pressure, Cyclic Loading, Fiber Content Methods for accurate diffusivity measurements accounting for edge effects and anisotropy need to be developed and verified. Unfortunately such experimental verification takes very long times (i.e., experiments lasting 1-2 years). Non-Fickian models for moisture diffusion are needed. Two or three-parameter models might be necessary. Thermal and mechanical cycling effects moisture absorption considerably. Coupled effects of thermal and mechanical cycling during and prior to moisture absorption are not understood. Significant market opportunities exist for composites in off-shore applications. Modeling and experimental verification of long term performance of molded composites under high pressure (10,000 psi and above) and high temperature ( °F) is needed to gain acceptance in such markets.