• Energy Research

Thermo-mechanical process simulation of an industrially pultruded rectangular hollow profile is presented. Glass/polyester is used for the continuous filament mat (CFM) and the uni-directional (UD) layers. The process induced residual distortions together with the temperature and degree of cure are predicted using a three dimensional (3D) thermo-chemical model sequentially coupled with a 2D quasi-static generalized plane strain mechanical model. The predicted deformation pattern at the end of the process is found to agree quite well with the one observed for the real pultruded parts in a commercial pultrusion company. In addition, the predicted warpage behaviour is further analysed by adjusting the mandrel length as well as including the mandrel heating. Using the proposed process model, the effect of the mandrel configurations on the quality of the pultrusion is investigated in terms of temperature, degree of cure and distortions.These unwanted residual distortions may lead to not meeting the desired geometrical tolerances e.g. warpage of pultruded window frames and hollow profiles as well as spring-in of L-shaped profiles, etc.

A novel three dimensional thermo-chemical simulation of the pultrusion process is presented. A simulation is performed for the pultrusion of a NACA0018 blade profile having a curved geometry, as a part of the DeepWind project. The finite element/nodal control volume (FE/NCV) technique is used. First, a pultrusion simulation of a U-shaped composite profile is performed to validate the model and it is found that the obtained cure degree profiles match with those given in the literature. Subsequently, the pultrusion process simulation of the NACA0018 profile is performed. The evolutions of the temperature and cure degree distributions are predicted inside the heating die and in the post-die region where convective cooling prevails. The effects of varying process conditions on the part quality are investigated for two different heater configurations and with three different pulling speeds. Larger through-thickness gradients are obtained for the temperature and degree of cure as the pulling speed increases. This will affect the process induced residual stresses and distortions during manufacturing.

This paper in particular deals with the integrated modeling of a pultruded NACA0018 blade profile being a part of EU funded DeepWind project. The manufacturing aspects of the pultrusion process are associated with the preliminary subsequent service loading scenario. A 3D thermo-chemical analysis of the pultrusion process is sequentially coupled with a 2D quasi-static mechanical analysis in which the process induced residual stresses and distortions are predicted using the generalized plane stain elements in a commercial finite element software ABAQUS. The temperature- and cure-dependent resin modulus is implemented by employing the cure hardening instantaneous linear elastic (CHILE) model in the process simulation. The subsequent bent-in place simulation of the pultruded blade profile is performed taking the residual stresses into account. The integrated numerical simulation tool predicts the internal stress levels of the profile at the end of the bending analysis. It is found that the process induced residual stresses have the potential to influence the internal stresses arise in the structural analysis.

Pultrusion is a continuous manufacturing process used to produce high strength composite profiles with constant cross section. The mutual interactions between heat transfer, resin flow and cure reaction, variation in the material properties, and stress/distortion evolutions strongly affect the process dynamics together with the mechanical properties and the geometrical precision of the final product. In the present work, pultrusion process simulations are performed for a unidirectional (UD) graphite/epoxy composite rod including several processing physics, such as fluid flow, heat transfer, chemical reaction, and solid mechanics. The pressure increase and the resin flow at the tapered inlet of the die are calculated by means of a computational fluid dynamics (CFD) finite volume model. Several models, based on different homogenization levels and solution schemes, are proposed and compared for the evaluation of the temperature and the degree of cure distributions inside the heating die and at the postdie region. The transient stresses, distortions, and pull force are predicted using a sequentially coupled three-dimensional (3D) thermochemical analysis together with a 2D plane strain mechanical analysis using the finite element method and compared with results obtained from a semianalytical approach.

This paper investigates the integrated modeling of a pultruded NACA0018 blade profile which is a part of the FP7 EU project DeepWind. The pultrusion process simulation is combined with the preliminary subsequent in-service load scenario. In particular, the process induced residual stresses and distortions are predicted by using a new approach combining a 3D Eulerian thermo-chemical analysis, in which the temperature and the cure degree distributions are obtained, and a 2D quasi-static plane strain mechanical analysis. The post-die region where convective cooling prevails is also included in the process model. The bending into shape of the pultruded blade profile is simulated with and without taking the residual stresses into account. The internal stress distribution in the profile is evaluated after the bending analysis and it is found that the process induced residual stresses have the potential to promote or to demote the internal stresses in the structural analysis.

In the present study, three dimensional (3D) numerical modeling strategies of a thermosetting pultrusion process are investigated considering both transient and steady state approaches. For the transient solution, an unconditionally stable alternating direction implicit Douglas-Gunn (ADI-DG) scheme is implemented as a first contribution of its kind in this specific field of application. The corresponding results are compared with the results obtained from the transient fully implicit scheme, the straightforward extension of the 2D ADI and the steady state approach. The implementation of the proposed approach is described in detail. The calculated temperature and cure degree profiles at steady state are found to agree well with results obtained from similar analyses in the literature. Detailed case studies are carried out investigating the computational accuracy and the efficiency of the 3D ADI-DG solver. It is found that the steady state approach is much faster than the transient approach in terms of the computational time and the number of iteration loops to obtain converged results for reaching the steady state. Hence, it is highly suitable for automatic process optimization which often involves many design evaluations. On the other hand sometimes the transient regime may be of interest and here the proposed ADI-DG method shows to be considerably faster than the transient fully implicit method which is generally used by the general purpose commercial finite element solvers. Finally, using the proposed steady-state approach, a design of experiments is carried out for the curing characteristic of the product based on pulling speed and part thickness

Pultrusion is one of the most cost-effective manufacturing techniques for producing fiber-reinforced composites with constant cross-sectional profiles. This obviously makes it more attractive for both researchers and practitioners to investigate the optimum process parameters, i.e., pulling speed, power, and dimensions of the heating platens, length and width of the heating die, design of the resin injection chamber, etc., to provide better understanding of the process, consequently to improve the efficiency of the process as well as the product quality. Using validated computer simulations is “cheap” and therefore is an attractive and efficient tool for autonomous (numerical) optimization. Optimization problems in engineering in general comprise multiple objectives often having conflict with each other. Evolutionary multi-objective optimization (EMO) algorithms provide an ideal way of solving this type of problems without any biased treatment of objectives such as weighting constants serving as pre-assumed user preferences. In this paper, first, a thermochemical simulation of the pultrusion process has been presented considering the steady-state conditions. Following that, it is integrated with a well-known EMO algorithm, i.e., nondominated sorting genetic algorithm (NSGA-II), to simultaneously maximize the pulling speed and minimize “total energy consumption” (TEC) which is defined as a measure of total heating area(s) and associated temperature(s). Finally, the results of the evolutionary computation step is used as starting guesses for a serial application of a of gradient-based classical algorithm to improve the convergence. As a result, a set of optimal solutions are obtained for different trade-offs between the conflicting objectives. The trade-off solution, thus obtained, would remain as a valuable source for a multi-criterion decision-making task for eventually choosing a single preferred solution for the pultrusion process

Pultrusion is one of the most effective manufacturing processes for producing composites with constant cross-sectional profiles. This obviously makes it more attractive for both researchers and practitioners to investigate the optimum process parameters, i.e. pulling speed, power and dimensions of the heating platens, length and width of the heating die, design of the resin injection chamber, etc., to provide better understanding of the process, consequently to improve the efficiency of the process as well the product quality. Numerous simulation approaches have been presented until now. However, optimization studies had been limited with either experimental cases or determining only one objective to improve one aspect of the performance of the process. This objective is either augmented by other process related criteria or subjected to constraints which might have had the same importance of being treated as objectives. In essence, these approaches convert a true multi-objective optimization problem (MOP) into a single-objective optimization problem (SOP). This transformation obviously results in only one optimum solution and it does not support the efforts to get more out of an optimization study, such as relations between variables and objectives or constraints. In this study, an MOP considering thermo-chemical aspects of the pultrusion process (e.g. cure degree, temperatures), in which the pulling speed is maximized and the heating power is minimized simultaneously (without defining any preference between them), has been formulated. An evolutionary multi-objective optimization (EMO) algorithm, non-dominated sorting genetic algorithm (NSGA-II [Deb et al., 2002]), has been used to solve this MOP in an ideal way where the outcome is the set of multiple solutions (i.e. Pareto-optimal solutions) and each solution is theoretically an optimal solution corresponding to a particular trade-off among objectives. Following the solution process, in other words obtaining the Pareto-optimal front, a further postprocessing study has been performed to unveil some common principles existing between the variables, the objectives and the constraints either along the whole front or in some portion of it. These relationships will reveal a design philosophy not only for the improvement of the process efficiency, but also a methodology to design a pultrusion die for different operating conditions.