Numerical Modelling of the Filament Winding Proces (original) (raw)

Numerical Modelling of the Filament Winding Process

A framework composed of several building blocks that allows the simulation of the filament winding process is described. The physical and thermo-chemical phenomena interacting at the layer/laminate level together with analytical description of the compaction and consolidation mechanisms were modelled and incorporated in finite element software.

Finite element modeling of the filament winding process

Composite Structures, 2001

A ®nite element model of the wet ®lament winding process was developed. In particular, a general purpose software for ®nite element analysis was used to calculate the ®ber volume fraction under dierent process conditions. Several unique user de®ned subroutines were developed to modify the commercial code for this speci®c application, and the numerical result was compared with experimental data for validation. In order to predict the radial distribution of the ®ber volume fraction within a wet wound cylinder, three unique user de®ned subroutines were incorporated into the commercial ®nite element code: a ®ber consolidation/compaction model, a thermochemical model of the resin and a resin mixing model. The ®ber consolidation model describes the in¯uence of the external radial compaction pressure of a new layer as it is wound onto the surface of existing layers. The thermochemical model includes both the cure kinetics and viscosity of the resin. This model analyzes the composite properties and tracks the viscosity of the resin, which is a function of the degree of cure of the resin. The resin mixing model describes the mixing of``old'' and``new'' resin as plies are compacted. Validations were made by comparing image analysis data of ®ber volume fraction in each ply for ®lament wound cylinders with the FEM results. The good agreement of these comparisons demonstrated that the FEM approach has can predict ®ber volume fraction over a range of winding conditions. This approach, then, is an invaluable tool for predicting the eects of winding parameters on cylinder structural quality.

Fiber-waviness Model in Filament Winding Process

Journal of Solid Mechanics and Materials Engineering, 2010

Fiber waviness is one of the initial defects in the filament winding process, and causes reduction of compressive strength of the composite structure. The mechanism of growth of fiber waviness is, however, not completely clear. In the present study, a model for generating fiber waviness is proposed. It is assumed to be due to local fiber micro-buckling arising from the compression load caused by shrinkage of a metal jig. Three faults are considered as causes of micro-buckling: bonding between metal jig and composite, insufficient cure of the resin, and initial deflection of fibers. Analysis and experiments based on this model have been carried out.

Weighted Residuals Method Applied to the Filament Winding Process

Simultaneous process of filament winding and curing a reinforced polymer is widely used in composite industries. This process is normally used for manufacturing cylindrical structures of the composite materials, which have been used in many fields including aerospace and chemical industries, sports and military. The process consists in winding fibres passing through a resin bath on a mandrel at room temperature so that the resin is crosslinked via chemical reaction. In the chemical step, physical properties of the composite material obtained depend on temperature and cure behaviour; therefore, modelling and simulation of this process are important to investigate the quality of the final process product. The numerical solution of the mathematical model of the filament winding process is the main objective of this contribution. In this work, the modelling consists of the mass and thermal balances by using the autocatalytic model for describing the kinetic reaction rate. The model equation system was solved through the method of lines applying the weighted residuals method to discretize the radial direction and integrating the resulted differentialalgebraic system using the DASSL code. The temperature and degree of cure behaviours were obtained and validated with experimental data. A good agreement between experimental data and simulated results was verified associated with low computational cost. Based on simulated results, the process variables can be optimised aiming to the quality of final product.

SELECTION OF OPTIMAL PROCESSING PARAMETERS IN FILAMENT WINDING

Wet filament winding is widely used as a production method for fibre reinforced plastic tubular products. The resulting quality of the product can be characterised by parameters like fibre volume fraction, void content and mechanical properties under certain loading conditions. These quantities are functions of the initial characteristics of resin and fibres, processing parameters (e.g. fibre tension, resin temperature, number of fibre tows and winding speeds) and geometric parameters such as mandrel diameter and winding angle. These initial parameters determine other important secondary processing parameters, namely the layer thickness and the resulting bandwidth. Generally these secondary processing parameters are, at the same time, design requirements. It is therefore important to predetermine the proper material and processing parameters in order to produce parts with the desired characteristics and quality. The results of a study are presented in which the influence of processing parameters on the final part characteristics is investigated using video analysis and other techniques. The aim of this study is to contribute towards a more rational winding process optimisation method and to reduce the reliance on the current empirical approach.

Unified approach of filament winding applied to complex shape mandrels

Composite Structures, 2014

The filament winding process faces up limiting fabrication inconveniences when designing complex geometries of composite structures. Even the complete coverage of a cylindrical mandrel requires introducing deviation from geodesic trajectories. As a consequence, models for non-geodesic paths have been developed. The present research aims to establish, to solve and to validate a generic mathematical model that contributes either to wind complex shapes, or to solve common filament winding disadvantages, on the basis of an integrated strategy. This so-called unified approach leads to benefit of composite structures made by filament winding despite the limitations of the manufacturing process. Based on the mathematical description of the mandrel geometry, the theory of surfaces leads to express the local curvatures. Considering the slippage tendency of the fiber tow over the surface, a local stability criterion involving mathematical parameters of the mandrel surface is established, and a general fiber path equation can be formulated. A numerical tool is developed and applied to predict the evolution of the filament winding angle of the fiber tow placed over the surface of two axisymmetric geometries: a convex and a concave one. Experimental validation is carried out by manufacturing these geometries using a four axis filament winding machine.

Model and experimental study of fiber motion in wet filament winding

Composites Part A: Applied Science and Manufacturing, 1998

A model for wet filament winding of thermosetting matrix cylinders was developed. The model relates the processing conditions (applied temperature, fiber tension and processing speed) to temperature, degree of cure, fiber volume fraction and stresses and strains within the composite cylinder. In this work, the modeling techniques behind predicting fiber volume fraction are described and validated. Specifically, a fiber motion model was developed which describes the motion of each layer of the cylinder during winding. This fiber motion model includes the effects of the fiber bed compacting as each new layer is wound and the resin flow through the porous fiber bed. In addition, several unique features were incorporated into the fiber motion model: fiber bed stiffness is evaluated as a function of both the fiber volume fraction and the resin cure state and a rule for the mixing of highviscosity resin with lower viscosity resin to simulate bleeding through the resin from previously wound layers. Both effects must be included if the fiber position is to be accurately predicted. Model predictions for fiber volume fraction are compared with several full-scale (1 m diameter) commercially wound cylinders. Tow tension, winding time and fiber sizing were varied. There is good agreement between model predictions and experimental data. Several important trends in both the data and model were observed: (1) low fiber volume fraction layers occur when the time between winding one layer and the next is long enough for the resin to each gelation; (2) even when the layer has not completely gelled, high-viscosity resin can bleed through to the next layer wound and Cause a low fiber volume fraction; and (3) fiber sizing can increase the overall fiber volume fraction by improving the fiber bed compaction characteristics.

Mathematical Modelling of Fibre Winding Process for Composite Frames

Communications - Scientific letters of the University of Zilina, 2016

This article describes the authors´ own mathematical modelling designed for the production process of a new type of low-weight composite frame. The used real technology is based on the winding of carbon or glass filament rovings around a polyurethane core which is a frame with a circular cross section (this type of composites is used, for example, to reinforce the doors and windows of airplanes). The core is attached to the end-effector of the robot (robot-end-effector) and successively passes through the fibre-processing head during the winding process. Quality production depends primarily on the correct winding of fibres around the polyurethane core. It is especially needed to ensure the correct angles of the fibre winding around the polyurethane core and the homogeneity of individual winding layers. The numerical model described in Euclidean space E3 of the manufacturing process is used when the fibre-processing frame is passing through the fibre-processing head. We use the described mathematical model and matrix calculus to enumerate the trajectory of the robot-end-effector to determine the desired passage of the core through the fibre-processing head. The calculated sequence of "tool-centre-point" values of the robot allows us to define the desired trajectory of the robot-end-effector and, thereby, the passage of the frame through the fibre-processing head. The calculation of the trajectory was programmed in the Delphi development environment. A practical example is analysed in the article.

Clean wet-filament winding - Part 1: design concept and simulations

Journal of Composite Materials, 2013

This is a two-part paper where part 1 presents details of a modified wet-filament winding process. Here, the resin bath was replaced with a resin injection system that impregnated the fibres prior to winding them onto a rotating mandrel. The resin and hardener were stored in separate containers and pumped on-demand via a pair of precision gear-pumps to a static mixer. The mixed resin system was then supplied to a custom-designed resin impregnation unit. The theoretical basis for the design of the resin impregnation unit is presented along with simulations of the various parameters that influence the impregnation time and the degree of impregnation. Part 2 of this series papers presents the experimental data on the performance of the resin impregnation unit and a comparison of the physical and mechanical properties of the tubes manufactured using the conventional and modified wet-filament winding techniques.