Lightweight Composite Structure for Solar Central Receiver Heliostats

Fadlalah, Sulaiman
Anderson, Timothy
Nates, Roy
Item type
Degree name
Doctor of Philosophy
Journal Title
Journal ISSN
Volume Title
Auckland University of Technology

Of all Concentrating Solar Power (CSP) technologies available today, central tower CSP systems are moving to the forefront as they have the capability to become the technology of choice for the generation of renewable electricity. The potential of central tower systems to achieve high temperatures offers a path to higher efficiencies, thereby providing an inherent advantage versus the other CSP systems. Achieving these high temperatures requires a large number of heliostats, and therefore the heliostats are considered the most crucial cost element of central tower CSP systems amounting up to 50% of a plant’s total cost. To address this issue and in order for the cost of energy from central tower plants to be competitive with that of other energy systems, there is a need for innovative heliostat designs that can reduce the heliostats’ cost without affecting its performance. One way of reducing this cost is by utilizing lightweight honeycomb sandwich composites in the heliostat structure, reducing the size of the drive units and their energy consumption. However, one of the challenges faced in implementing such systems is ensuring that they are able to cope with the aerodynamic forces imposed upon them during operation. Despite the progress in heliostat development, a comprehensive review of literature revealed a lack of work undertaken on investigating the suitability of honeycomb sandwich composites for use as a heliostat mirror structure. This gap indicated a need to deliver a better understanding on the interaction between the wind and honeycomb sandwich composites employed as a heliostat mirror support structure by investigating their aero-structural robustness and behaviour characteristics. The research first studied the flow behaviour and aerodynamic loads on a stand-alone heliostat using computational fluid dynamics (CFD), with particular emphasis on the effect of wind direction and its impact on the aerodynamic loading of a heliostat. This aspect of loading had not previously been explored in any detail. The model was validated by comparing the computation predictions of the heliostat’s aerodynamic coefficients with both experimental measurements and numerical results from previously published work. The study showed that, for a 0° wind incidence angle, the drag and base overturning moment coefficients decrease as the tilt angle alters from vertical to horizontal. The lift and hinge moment coefficients, on the other hand, showed an asymmetric behaviour about the 0° tilt angle with maximum values occurring at tilt angles of 30° and -30°. Increasing wind incidence angle affected the wind loading coefficients (drag, lift, base overturning moment and hinge moment coefficients) by decreasing their magnitudes at different rates. A subsequent non-linear regression analysis delivered a correlation for each of the coefficients based on the heliostat’s tilt and wind incidence angle was developed. These formulations provide a useful analytical tool for heliostat designers to determine the wind loads on heliostats and to assess structural forces and moments on the frame of the heliostat and its reflective surface. In summary, it was shown that wind incidence had a significant impact on the aerodynamic loads encountered by a heliostat and, therefore, needs to be accounted for when examining the structural integrity of heliostats. Secondly, the study investigated the aero-structural behaviour characteristics of a proposed honeycomb sandwich composite-based heliostat structure by performing numerical fluid-structure interaction (FSI) simulations for several loading conditions at various tilt and wind incidence angles. The structural response of the heliostat’s honeycomb sandwich panel showed markedly different behaviour characteristics at various operational conditions. From the results, it was shown that the effect of heliostat’s tilt orientation on the sandwich panel’s maximum deflection and stresses becomes more pronounced as wind velocity increases above 10 m/s. This effect becomes more vital and the difference in the maximum displacement and stress values at different tilt angles escalates to a maximum at wind velocity of 20 m/s. Moreover, the wind velocity effect on the heliostat panel for the case of 0° tilt angle was negligible. This is because of the flow uniformity (the projected area of the reflector directly facing the wind is at its minimum) that leads to a significant decrease in the wind loading effect on the panel at this tilt orientation for all wind velocities (5-20 m/s). The study showed that increasing wind incidence angle affected the recorded maximum displacement and stress results by reducing their magnitudes at different rates. This is due to the fact that the heliostat’s projected area directly facing the wind decreases with the increase in wind incidence angle. This consequently reduces the effect of the blockage, causing a decrease in the wind loading effect on the heliostat. As the wind incidence angle gradually increases from 45° to 90° (the projected area of the reflector continues to decrease) for all tilt angles, the wind incidence angle influence on the maximum displacement and stress values gradually increased and the values notably decreased thus reaching its minimum at β = 90°. This implies that the heliostat panel at 90° wind incidence angle, regardless of any tilt angle, is not significantly influenced by wind loadings at wind velocities of 20 m/s and below. The study also showed that when wind strikes the heliostat structure at 0° and 45° incidence angles, the shielding effect caused by the supporting components and torque tube was clearly noticeable. When the incoming wind acted on the reflector’s back surface, the maximum displacement and stress values were slightly lower compared to the ones recorded when the flow acted on the heliostat’s mirror surface. In all of the operational conditions studied, it was concluded that the worst case was found to be at a tilt angle of 30° under the effect of wind flow at 0° to the heliostat surface with a velocity of 20 m/s. Despite this observation, it was found that the heliostat managed to maintain its structural integrity according to relevant optical and material failure standards. Taking the worst case operational condition as a basis, and given that the mechanical properties of honeycomb core-based sandwich composites are highly dependent upon the honeycomb’s geometric configuration (e.g., cell wall angle (φ), cell wall length (a), cell wall thickness (t)) and the core thickness (D), a comprehensive parametric study was performed to investigate the effect that each of these parameters has on the aero-structural behaviour characteristics of the honeycomb sandwich composite-based heliostat. The study was carried out for three different core thicknesses (D) with various honeycomb configurations. From this the study revealed that varying the honeycomb’s cellular geometry significantly affected both the strength and stiffness properties of the sandwich composite-based heliostat structure, illustrating that it is attainable to control suitably the strength of the heliostat’s honeycomb sandwich panel to achieve superior mechanical properties by varying the cell’s configuration. These variations in the heliostat’s structural response highlighted the necessity for a generalized model that can capture the influence of each of the honeycomb core’s geometrical parameters on the heliostat structure’s performance (i.e. optical, material failure and weight reduction). Having a predictive model that estimates the heliostat’s structural performance, under the worst case operational condition and based on the desired site’s maximum recorded wind speed, eradicates the need of going through the hurdles of establishing an FSI model for each of the honeycomb core’s geometrical parameters. This, in turn, runs down the implementation time and keeps off unnecessary computations. In this sense, and given that this approach is one of the prominent tools for modelling complex non-linear relationships, particularly in situations where the development of phenomenological or conventional regression models becomes impractical or cumbersome, artificial neural network (ANN) technique was utilized to establish a novel predictive model that predicts the structural performance of the honeycomb sandwich composite-based heliostat based on its honeycomb core’s physical parameters. The results showed that the established ANN model was capable of accurately predicting the structural performance of the honeycomb sandwich composite-based heliostat. Finally, a rigorous investigation was carried out on the utilization of particle swarm optimization (PSO) algorithm to establish a novel prediction-optimization model that predicts and optimizes the structural performance of honeycomb sandwich composite-based heliostats. The model couples the ANN predictive model with the PSO algorithm for determining the optimum honeycomb core configuration leading to minimum self-weight of the heliostat’s sandwich composite panel while satisfying the structural performance requirements (i.e. optical and material failure). It was shown that the proposed integrated ANN-PSO model, which was encompassed as a user-friendly graphical user interface (GUI), delivers a useful, flexible and time-efficient tool for heliostat designers to predict and optimize the structural performance of honeycomb sandwich composite-based heliostats as per desired requirements. In summary, the work presented is a significant milestone in the quest to develop cheaper lightweight heliostats that are strong and capable of withstanding wind loads and other environmental conditions, and a major step on the way to move central tower CSP systems to the forefront to become the technology of choice for energy production.

Solar Energy , Central Tower Concentrating Solar Power (CSP) systems , Heliostat , Computational Fluid Dynamics (CFD) , Wind , Fluid-Structure Interaction (FSI) , Artificial Neural Network (ANN) , Particle Swarm Optimization (PSO) , Finite Element Analysis (FEA).
Publisher's version
Rights statement