T. Staffan Lundström

Effect of Perturbation of Fibre Architecture on Permeability Inside Fibre Tows

The influence on the permeability inside fibre tows from fibre packing arrangement and non-uniform fibre diameter is investigated. The analysis mainly consider flow transverse to the fibres and is based on the fibre geometry. Perturbations from perfect quadratic and hexagonal arrangements of fibres at constant fibre volume fraction are considered. The importance of microgeometry on permeability is pointed out and it is indicated how to deal with this problem. The analysis shows that the fibre transverse permeability is strongly dependent on the fibre size distribution. When an arrangement of equally sized fibres is continuously changed from quadratic to hexagonal it is furthermore found that the transverse permeability is highest just before the packing gets hexagonal. The analysis is also applied to a simple FORTRAN code to simulate removal of fibers from a quadratic arrangement. The simulations shows that the transverse permeability decreases with increased number of removed fibres under constant fibre volume fraction.

Measurement of void collapse during resin transfer moulding

Abstract This paper considers the dissolution of cylindrical voids trapped between fibres during the resin transfer moulding process. To study the behaviour of the tiny voids, a system is used which consists of a transparent mould loaded with a glass fibre reinforcement. The mould is placed under a microscope and a small area of the preform is monitored during the injection of a vinylester resin. The length of the voids decreases at a relatively fast rate and they disappear in the order of minutes due to diffusion. The experimental results are rationalized through a theory showing how certain parameters influence the rate of dissolution. The investigation shows the significance of degassing the resin before injection. This is especially important when low pressures are utilized during injection.

In-plane permeability measurements on fiber reinforcements by the multi-cavity parallel flow technique

This report discusses the advantages and drawbacks of the multi-cavity parallel flow technique for permeability measurements. An experimental series with repeated measurements on material from the same roll shows that the repeatability of the technique is very good considering the manufacturing variability of the fabric. The measured standard deviation in the repeatability study is about 10%. It is, however, shown that the permeability can vary considerably- between reinforcements of similar geometry. Furthermore, computer simulations were used to estimate the errors when highly anisotropic materials are oriented at an angle to the material principal direction in the parallel flow technique. The conclusion based on the simulations is that the length to width ratio of the cavity should be larger than the anisotropy of the reinforcement for an acceptable error.

A Statistical Approach to Permeability of Clustered Fibre Reinforcements

The focus is set on mesoscale modelling of permeability of real fabrics used in composite manufacturing. Of particular interest is the effect of expected perturbations from perfect geometries, such as fibre bundle crimp, on the permeability. To start with, variational methods are used to calculate the permeability of individual gaps between fibre bundles. Based on this study a network of unit cells is formed enabling studies of two-and three-dimensional flow through the structure. From such an analysis the overall permeability of an arbitrary distribution of unit cell permeabilities can be calculated. Here random and controlled distributions are simulated. The former is an approximate representation of a continuous strand mat and the latter may describe Non-Crimp Fabrics. The result is that for random distributions, the permeability decreases with the maximum variation in unit cell while for a controlled permeability distribution the overall permeability can as well increase as decrease depending on the type of perturbation. In both cases the type of flow: one-, two-or three-dimensional strongly influence on the quantitative result. Hence, for the type of fabrics studied, it is necessary to model the full 3D-flow through to get a correct permeability value.

Hydraulic Turbine Diffuser Shape Optimization by Multiple Surrogate Model Approximations of Pareto Fronts

A multiple surrogate-based optimization strategy in conjunction with an evolutionary algorithm has been employed to optimize the shape of a simplified hydraulic turbine diffuser utilizing three-dim ...

Numerical study of the local permeability of noncrimp fabrics

Noncrimp stitched fabrics (NCFs) are often used as reinforcing materials in high-performance composite materials. Prediction models of the processing stage of the manufacturing are highly desirable in order to enhance the control of the process and enable the production of materials with higher quality. In NCFs, layers of parallel fiber bundles consisting of a large number of fibers are stitched together with other layers to form a network of interbundle channels in different directions. In earlier works, numerical simulations on unit cells had been performed in order to predict the global permeability of NCFs. It was shown that features like the thread influence the local permeability of the unit cells and therefore, the local permeability distribution of a fabric also. Furthermore, this influences the global permeability of the entire fabric. In the present paper, different geometrical features are therefore studied in order to investigate their influence on the local permeability within an NCF. The stitching process in addition to the interbundle channels, gives rise to two geometrical features, the thread which penetrates the channels and the crossing of fibers between two neighboring fiber bundles. The influences of these two features on the local permeability are studied together with variations of other geometrical parameters of the fabric. Computational Fluid Dynamics are used for the flow simulations in order to calculate the local permeability for the different unit cells. To ensure quality and trust, the numerical accuracy of the simulations is also studied. This work proves that the thread and the crossings, as well as the variations of the width and the height of the interbundle channels, have great influence on the local permeability. Prediction models therefore, have to take these features as well as geometry distortions, which influence the local permeability distribution, into account in order to make accurate predictions of the global permeability of a fabric.

Numerical model for vacuum infusion manufacturing of polymer composites

The focus is set on the development and evaluation of a numerical mgodel describing the impregnation stage of a method to manufacture fibre reinforced polymer composites, namely the vacuum infusion process. Examples of items made with this process are hulls to sailing yachts and containers for the transportation industry. The impregnation is characterised by a full 3D flow in a porous medium having an anisotropic, spatial‐ and time‐dependent permeability. The numerical model has been implemented in a general and commercial computational fluid dynamic software through custom written subroutines that: couple the flow equations to the equations describing the stiffness of the fibre reinforcement; modify the momentum equations to account for the porous medium flow; remesh the computational domain in each time step to account for the deformation by pressure change. The verification of the code showed excellent agreement with analytical solutions and very good agreement with experiments. The numerical model can easily be extended to more complex geometry and to other constitutive equations for the permeability and the compressibility of the reinforcement.

Respiratory Deposition of Fibers in the Non-Inertial Regime—Development and Application of a Semi-Analytical Model

A semi-analytical model describing the motion of fibrous particles ranging from nano- to micro scale was developed, and some important differences in respiratory tract transport and deposition between fibrous particles of various sizes and shapes were elucidated. The aim of this work was to gain information regarding health risks associated with inhalation exposure to small fibers such as carbon nanotubes. The model, however, is general in the sense that it can be applied to arbitrary flows and geometries at small fiber Stokes and Reynolds numbers. Deposition due to gravitational settling, Brownian motion and interception was considered, and results were presented for steady, laminar, fully developed parabolic flow in straight airways. Regarding particle size, our model shows that decrease in particle size leads to reduced efficiency of sedimentation but increased intensity of Brownian diffusion, as expected. We studied the effects due to particle shape alone by varying the aspect ratios and diameters of the microfibers simultaneously, such that the effect of particle mass does not come into play. Our model suggests that deposition both due to gravitational settling and Brownian diffusion decreases with increased fiber aspect ratio. Regarding the combined effect of fiber size and shape, our results suggest that for particles with elongated shape the probability of reaching the vulnerable gas-exchange region in the deep lung is highest for particles with diameters in the size range 10–100 nm and lengths of several micrometers. Note that the popular multi-walled carbon nanotubes fall into this size-range.

CFD-Modelling and Validation of Free Surface Flow During Spilling of Reservoir in Down-Scale Model

Abstract Fully three dimensional modelling of the spilling from a reservoir with relatively complex geometry were performed and compared to experimental results from a physical scale model with the aim to advance the science of numerical modelling of free surface flow of real reservoirs. In the set-up in focus the water was spilled from the reservoir through three gates that could be manoeuvred separately. In the first case two of the gates were closed and the third gate was partly opened. In this experimental set-up the water surface in the reservoir was close to horizontal. Therefore it was here meaningful to compare a rigid lid modelling approximation to the more computational heavy method of Volume of Fluids. In the second case, all three gates were open, resulting in a nonhorizontal varied flow surface profile in the reservoir upstream critical sections at the spillway crests. This case was simulated with Volume of Fluids and the position of the air-water interface was derived for two turbulence models, the standard k-ε and SSG. Water levels, velocities and the shape of the water surface were compared to experiments. The simulation results capture qualitative features such as a vortex near the outlet and show good quantitative agreement with the experiments regardless of method used to simulate the free surface. In general, simulations with the standard k-ε and the more advanced SSG turbulence models give the same results with respect to the averaged quantities measured.

Longitudinal dispersion coefficient : effects of particle-size distribution

The effects of particle-size distribution on the longitudinal dispersion coefficient (