Search

Metamodels are developed and used for aerodynamic optimization of a ducted opening integrated into a high-rise building to maximize the amplification factor within the duct. The duct consists of a nozzle, a throat, and a diffuser. 211 high-resolution 3D RANS CFD simulations are performed to generate training and testing datasets. The space-filling design and Genetic algorithm are used for data sampling and optimization, respectively. The performance of five commonly-used metamodels is systematically investigated: Response Surface Methodology (RSM), Kriging (KG), Neural Network (NN), Support Vector Regression (SVR), and Genetic Aggregation Response Surface (GARS). The investigation is based on (i) detailed in-sample and out-of-sample evaluations of the metamodels, (ii) annual available power in the wind (Pavailable), and (iii) annual energy production (AEP) for a 3-bladed horizontal-axis wind turbine (HAWT) installed in the mid-throat for the optimum designs obtained by the metamodels. The results show that converging-diverging ducted openings can magnify the experienced wind speed by the turbine and enhance the available wind power. In addition, the use of different metamodels can lead to a variation of up to 153% in the estimated Pavailable. For a small dataset, crude yet still acceptable accuracy can be achieved for Genetic Aggregation Response Surface and Kriging at a very low computational time.

Scale-adaptive simulation (SAS) approach is employed to investigate the complex dynamic stall phenomena occurring on a wind turbine blade. The results are com-pared with the more popular less computationally-expensive unsteady Reynolds-averaged Navier-Stokes (URANS) approach where the latter is validated using three sets of experimental data. The comparison reveals that the two approaches have similar predictions of the instant of the formation/bursting/shedding of the laminar separation bubble (LSB) and dynamic stall vortex (DSV), the size of the LSB and aerodynamic loads during the upstroke. This is while the two approach-es exhibit dissimilar predictions of the trailing-edge vortex characteristics, its in-teraction with the DSV, number of secondary vortices and aerodynamic loads during the downstroke.

The current study systematically analyzes the impact of number of blades (n) on the aerodynamic performance of 2-, 3- and 4-bladed Darrieus H-type vertical axis wind turbines (VAWTs). A large number of operational parameters, i.e., tip speed ratio (λ), Reynolds number (Re), turbulence inten-sity and reduced frequency (K) are investigated to provide a deeper insight into the impact of n on the dynamic loads on the blades, the turbine performance and the wake. High-fidelity unsteady Reynolds-averaged Navier-Stokes (URANS) simulations, extensively validated with experiments, are employed. The results show that (i) within the turbine optimal operational range, the turbine power coefficient (CP) is almost independent of n; (ii) when dynamic stall is present, CP values are dependent on n due to the impact of K; and (iii) decreasing n leads to an increase in the maximum lift coefficient, while the drag coefficient of the blade(s) reduces due to the higher K. The present findings support the optimal aerodynamic design of small- to large-scale VAWTs.

This chapter discusses the simulation of urban thermal microclimate with a focus on heat waves in urban areas, the simulation of overheating of buildings and the effects of adaptation measures to limit temperatures in buildings and urban areas during heat waves. The spatial scales are the meteorological microscale (neighborhood scale) and the building scale; the methods are computational fluid dynamics and building energy simulation. Adaptation measures investigated at the neighborhood scale are avenue trees, green facades and green roofs; adaptation measures at the building scale are increased thermal resistance, increased thermal mass, increased short-wave reflectivity of facades and roofs, peak ventilation, vegetated roofs and exterior solar shading.

Vertical axis wind turbines (VAWTs) are promising candidates for wind energy harvesting in the urban environment. However, their aerodynamic performance still falls behind of their horizontal axis counterparts. This could be associated to the comparatively small research they have received in the past decades as well as their complex unsteady aerodynamics. Computational Fluid Dynamics (CFD) has been widely used to evaluate and improve the aerodynamic performance of VAWTs. An extensive literature study reveals that the 2D unsteady Reynolds-Averaged Navier-Stokes (URANS) approach has been used in the majority of the CFD studies on VAWTs. The current study intends to evaluate the aerodynamic performance of a VAWT, calculated using 2D URANS, and compare it with that of 2.5D URANS and 2.5D scale-adaptive simulation (SAS). SAS is a hybrid RANS-LES model developed by Menter and Egorov [1]. The four-equation transition SST turbulence model is employed in the URANS simulations as well as in the RANS region of the hybrid RANS-LES simulation. The studied turbine is a one-bladed Darrieus H-type VAWT with a solidity of 0.125 operating at a low tip speed ratio of 2.0, which corresponds to the most complex case for VAWTs where dynamic stall is dominant. The reduced frequency is 0.125 representing the high unsteadiness in the flow. Significant benefits of the one-bladed turbine are: (i) less blade-wake interactions while the essential flow features, such as dynamic stall, are still present, (ii) reduced computational costs due to the smaller number of cells. The turbine characteristics is based on the experiment by Simão Fereira et al. [2]. Validation studies for the one-bladed turbine as well as the other turbines have been performed [3-5]. A comparative analysis of the instantaneous tangential and normal loads on the turbine (see Fig. 1), spatiotemporal distribution of pressure coefficient (see Figs. 2a-c) and skin friction coefficient (see Fig. d-f) on the blade suction side, the evolution of the shed vorticity by the blade, dynamic loads on the blade and the turbine wake are employed to evaluate the performance of URANS modeling in comparison to the SAS model. The instantaneous turbine loads calculated using the 2D and the 2.5D URANS, shown in Fig. 1, are in line with minor differences in the downwind side. Despite the 180 times higher number of cells and 10 times finer time step of the SAS modeling, an overall good agreement exists between the 2D URANS and the SAS results. The predicted thrust coefficients for 2D and 2.5D URANS and SAS are 0.422, 0.424 and 0.430, respectively. Nevertheless, there exist noticeable differences between the URANS and SAS results in the bursting location of the laminar separation bubble (LSB), the evolution of the dynamic stall vortex (DSV), the leading-edge secondary and tertiary vortices and the trailing-edge separation. The findings of the present study help to highlight the deficiencies of URANS modeling of VAWTs in dynamic stall.

Vertical axis wind turbines (VAWTs) are promising candidates for wind energy harvesting in the urban environment. However, their aerodynamic performance still falls behind of their horizontal axis counterparts. This could be associated with the comparatively small research they have received in the past decades as well as their complex unsteady aerodynamics. Computational Fluid Dynamics (CFD) has been widely used to evaluate and improve the aerodynamic performance of VAWTs. An extensive literature study reveals that the 2D unsteady Reynolds-Averaged Navier-Stokes (URANS) approach has been used in the majority of CFD studies on VAWTs while the sensitivity of the CFD results to the choice of the turbulence model has not yet been comprehensively investigated in the literature. Therefore, the current study intends to evaluate the aerodynamic performance of three different VAWTs, calculated using 2D URANS with seven different one- to four-equation turbulence models, namely Spalart-Allmaras, RNG and realizable k-ε, k-ω SST, kωSSTi (with the intermittency model), k-kl-ω and transition SST. The results are also compared against inviscid simulations. The turbines are selected based on the availability of the experimental data for the comparison in addition due to their geometrical and operational differences which would help to further generalize the conclusions. A comparative analysis of the turbine power performance and wake velocities for the turbines indicates that the 3-equation k-kl-ω and the 4-equation transition SST models are the two best-performing models while the inviscid modeling and the ε-based models exhibit the worst performance for CP predictions in comparison to the experiment. The findings support more accurate CFD simulations of VAWTs.