A novel method for the estimation of wind pressure acting on the surface of buildings is proposed. The proposed method focuses on the response of double glazing to wind pressure. It takes its basis in a physical modeling of load-response relationship considering the behavior of the air between the two glasses consisting of double glazing. It also utilizes the data collected on-site under no wind condition in order to estimate the unknown parameter values required for the wind pressure estimation. The possibility and the validity of the proposed method were investigated with experiments using a pressure control box. It was demonstrated that the wind pressure estimation is possible and works well for cases. However, in several cases the errors of the estimation were found to be significant presumably associated with biases in estimating temperature of the air between the glasses and due to lack of sufficient data for unknown parameter estimations.

View full abstract

The three-dimensional turbulent flow was calculated using the large eddy simulation (LES) to solve the flow around a polymer insulator with/without a Core and its aerodynamic noise by OpenFOAM with Curle's equation. The resonance of aerodynamic noise also was measured using a low-noise wind tunnel. The diameter of the insulator Shed was varied at D = 129, 155, and 178 mm. The incline angle of the test model was varied from 0º to 30º. Vortices were generated independently from the Core (cylindrical shape) and the Shed (disk shape). The separated vortices from the Shed have a higher frequency than the wake vortices of the Core. The sound pressure level of a high frequency increases with increasing the incline angle. The frequency of the separated vortices from the Shed depends on the thickness at the tip. It was found that the resonance of high-frequency aerodynamic noise generated by polymer insulators has a source in the separated vortices from the Shed.

View full abstract

In this study, an experiment was conducted to examine the impact of opening and closing doors on the risk of preventing airborne droplet leakage when a home treatment room is set up in a house. A home treatment room and a hallway were reproduced in an artificial climate room, and a PIV analysis was conducted on the air outflow characteristics when the door is opened, assuming a home. The experiment targeted two types of doors, and experiments were conducted on cases in which the pressure difference in front of and behind the door and the temperature difference between the rooms were changed. Through the experiment, the amount of air outflow from the door was calculated for each case examined, and the air outflow characteristics from the door were clarified.

View full abstract

In this study, we performed a numerical study (LES) on the influence of separated flows from the wind turbine nacelle and tower on the wind turbine near wake region. As a result, when the tower part was made into a cylinder, a downward flow was induced behind the tower. It has become clear that as a result of this phenomenon, the wake region behind the wind turbine blades is also entrained toward the ground. In the future, we plan to change the size of the wind turbine tower and examine the above flow mechanism in detail.

View full abstract

Hu and Yoshie (2020) proposed SER_C* (Stability Effect Ratio for Normalized pollutant concentration C*) as an index that expresses the stability effect on pollutant concentration. SER_C* represents the ratio of C* under non-neutral atmospheric stability conditions to that under neutral condition. They showed that the SER_C* increases with the Bulk Richardson number (R_b) in wind tunnel experiments. However, it is not clear to what extent the range of R_b in the wind tunnel experiments covers the range of R_b in real atmosphere. And the correspondence between R_b and the Pasquill atmospheric stability class is also unclear. To clarify these issues, in this study, we first calculated R_b in the real atmosphere using observed wind speed and air temperature data and ground surface temperature obtained by WRF analysis, and clarified the range and occurrence frequency of R_b, and correspondence between R_b and the Pasquill Atmospheric Stability Class. Next, we investigated the relationship between SER_C* and R_b in the real atmosphere by classifying NO_x concentrations observed at air monitoring stations in Tokyo by R_b.

View full abstract

The objective of this study is to develop an acoustic anemometer for dense wind speed field measurements in urban spaces. The advantage of this anemometer is its lower cost and ease of installation compared to existing measurement devices. The proposed anemometer calculates wind speed based on the sound incurred by the vibration of a string excited by Karman vortex street. Our previous report confirmed that wind speed can be estimated when the string is perpendicular to the wind direction. In this study, we experimentally verify that wind speed estimation is possible even when the string is inclined relative to the wind direction. Using a string with a diameter of D = 0.20mm, it was found that the wind speed component perpendicular to the string can be estimated within a relative error of 5% in a uniform flow.

View full abstract

We examined the accuracy of a finite element code that enables us to analyze wind in the atmospheric boundary layer in a typhoon using wind observation data. As wind observation data, we used wind profiler and AMeDAS observation results during the period when Haishen 2020 passed near Minami Daito Island. Two gradient wind models, the Kepert model and the Georgiou model, were selected as upper-air boundary conditions to investigate their effects. As a result, it was found that the type of gradient wind model has a substantial effect on the consistency with observation results. In addition, in the numerical analysis, the drag coefficient of the sea surface is given over the whole lowest boundary without considering topographical and ground roughness effects. It was confirmed that such idealization leads to conservative wind speed in comparison with observation.

View full abstract

In this study, the pseudo global warming experiment of Typhoon Hagibis (2019) was conducted to investigate the impact of global warming on the development and weakening of the typhoon. The air temperature, geopotential height and sea surface temperature for the future climate condition were set by adding the global warming components of these meteorological elements to the NCEP FNL. The global warming components of the air temperature, geopotential height and sea surface temperature were calculated from the differences of October monthly mean values between in the 2010s and 2090s, which were predicted by MRI-CGCM3 under the RCP8.5 scenario. Under the future climate condition, the typhoon maintained its intensity up to higher latitudes than under the current climate condition, and the typhoon intensity at the landfall latitude increased significantly. The reason for the maintained increase in the typhoon intensity was that the rising of the air temperature and sea surface temperature due to global warming increased the amount of water vapor in the lower atmosphere at higher latitudes and made the atmospheric condition unstable.

View full abstract

The effect of global warming on local-scale severe wind hazards caused by an intense tropical cyclone (TC) was investigated using a regional high-resolution atmospheric model. Pseudoglobal warming approaches with various warming scenarios were applied to Typhoon Jebi (2018) that passed through the Kinki district of Japan and induced devastating wind disasters. A higher rise in temperature under warmer climates resulted in a stronger TC just before the landfall on the Pacific coast of western Japan. The enhanced TC significantly increased maximum wind speed around the coastal and/or lower-elevation regions in Kinki. In the Kobe- Osaka metropolitan areas, for example, the areal-mean maximum surface wind speed showed a robust linear trend with the temperature rise at the height of 2 m (ΔT_2m), increasing by 1.16 m s^{-1 °}C^-1. The exposed area of violent winds rapidly expanded in the future warming climates. The area size approximately doubled per unit rise in ΔT_2m. The results suggested that global warming increases the risks of severe wind disasters in the Kinki district in terms of increased winds over wider areas.

View full abstract

In this study, meteorological observations were conducted over a wide area and at multiple locations in northern Kyoto Prefecture from late October 2023, and the wind speed, duration, and frequency of Yuragawa arashi were clarified. The results showed that after sunset of the Yuragawa arashi, a clear pressure difference was observed at each location, with a maximum pressure of 1.7 hPa higher at Fukuchiyama than at the mouth of the river around 8:00 a.m. After sunset, the wind speed at the mouth of the river changed to the southerly direction and reached more than 5m/s from around 19:00, with a maximum of 6.6 m/s. Yuragawa arashi when the pressure difference between the Fukuchiyama Basin and the mouth of the Yuragawa is large.

View full abstract

Numerical experiments using a regional meteorological model with sensible heat flux (SHF) from the ground being varied were conducted for summertime local circulations over Osaka to investigate the effects of SHF on the variability of local-scale winds. It was shown that the decrease in SHF increased the surface level pressure, resulting in smaller pressure gradient. This led to weaker wind speeds at 10 m and 850 hPa, in particular reducing the frequency of strong winds. The weakening of wind speeds at 10 m contributed to the reduction in the intensity of the horizontal convergence at 10 m. Consequently, the wind speed maxima of the mean vertical winds and their height also decreased, causing decrease in the intensity and height of the horizontal divergence at 850 hPa. Therefore, these results indicated that the SHF modification affected the three-dimensional atmospheric circulation.

View full abstract

This study aims to investigate the applicability of the unsteady Reynolds averaged Navier–Stokes equations model (URANS) simulation with the proposed modification that reproduces periodic fluctuations induced by bluff bodies. The modification introduces large scale influence by adding the time scale of the production term of ε–equation in the standard k-ε model. Airflow within a simplified 9 × 9 cubic array with point-source dispersion of air pollutants was tested. We compared the URANS results with those from a large-eddy simulation (LES) and a wind tunnel experiment (WTE). Although the original model dropped the advection term in the time scale for total derivative and only considered the partial differential respect to time for flow around 2D square cylinder, but the advection term has great influence on reproducibility of periodic fluctuations of flow within the array in this study.

View full abstract

The objective of this study is to quantify the momentary pleasantness of wind, based on electroencephalography (EEG). We first conducted a chamber experiment under thermoneutral conditions, and obtained the EEG signals and subjective reports of the pleasantness level while the participants were exposed to wind at five different speeds. Then, using these data from 17 participants, we performed two analyses: 1) a t-test to identify the brain regions that showed a difference in EEG trend between pleasant and unpleasant conditions; 2) a machine learning method of Support Vector Machine (SVM), which estimates a pleasant or unpleasant condition from EEG signals. In the SVM, the brain regions that contribute to the estimation of the pleasantness were also identified. The classification accuracy with all the EEG signals, four frequency bands in 44 brain regions, 176 in total, was approximately 55%. In addition, the performance of classification, even with only 6 features of EEG signals which obtained higher values of weighting factor for SVM with all the EEG signals, was almost maintained compared with the analysis using all the EEG signals. Four of the six extracted features were also showed statistically significant difference (p < 0.05) between the pleasant and unpleasant conditions in the t-test.

View full abstract

This study analyzed the flow field within a rural village involving rows of windbreak trees called Igune on the Osaki Koudo in northern Miyagi Prefecture, Japan. Field observation was first conducted to obtain the spatial distribution of mean wind speed within the whole village area for validation data. Then, large-eddy simulations (LESs) were conducted to reproduce the flow field within the village and analyze the windbreak effect of Igune on the wind environment within the village after the validation using the observation data. The wind speed showed good agreement between the simulation and field observations. In the village, the mean wind velocity was reduced by half compared with that at the outside of the village. Additionally, a parametric study was carried out to understand the relationship between the streamwise spacing of Igune rows and their windbreak effect based on the simplified 2-dimensional flow field involving 3 Igune rows. The appropriate windbreak effect appeared when the distance between Igune rows was within 10 times the Igune height.

View full abstract

We proposed an extraction method for wind corridor by applying persistent homology (PH) to three-dimensional (3D) flow-field data. The procedure consists of three steps: 1)The wind speed ratio between local wind speed and surrounding averaged wind speed is calculated using two different size of space filters based on the result of 3D computational fluid dynamics (CFD) data. The high ratio regions expected that wind smoothly passes though. 2)Then, the three-dimensional ratio is converted into two-dimensional persistent diagram (PD) to understand the characteristics of the flow field by applying PH. 3) Finally, the long life-time regions in PD are visualized as wind corridors. In addition, we applied the proposed method to a flow field in a water front of Tokyo Metropolitan area calculated with CFD as an example, and compared the visualized wind corridors with these illustrated in an urban planning guidelines by Tokyo Metropolitan office.

View full abstract

Traditional windbreak trees called Igune, located in Osaki Koudo in northern Miyagi Prefecture, Japan, consist of taller trees and shorter shrubs. These Igune are planted on the west and north sides of residences to protect against prevailing winds in the winter season. In this study, large-eddy simulation (LES) of airflow around two-dimensional Igune rows were performed. Proper orthogonal decomposition (POD) was subsequently applied to the simulated flow fields to understand the mechanisms of the flow behind the Igune. Two Igune situations were considered: one with both tall trees and shrubs, and another with only shrubs following the removal of tall trees. The LES results revealed that when only shrubs were present, a highly turbulent wind environment with reverse flow developed behind them. Conversely, when both shrubs and trees were present, the turbulent fluctuations of the reverse flow behind the trees were relaxed. The spatial and temporal functions of the first two dominant POD modes were similar. However, in the case where only shrubs were present, the third and higher POD modes contributed to the reverse flow, whereas these modes were absent when both shrubs and trees were present.

View full abstract

The proper prediction of the wind pressure acting on building walls by LES requires high grid resolution near the walls and the difference schemes with low discretization error, which generally tends to be numerically unstable. For CFD simulations, recently, one of the widely used software is OpenFOAM. OpenFOAM has a specification that the static pressure at the outflow boundary of the computational domain must be a constant value. Our previous studies have confirmed that this specification causes excessive static pressure fluctuations that do not occur in real phenomena when the turbulent velocity fluctuations with fluctuating flow rate are imposed as the inflow boundary conditions. Based on the above, first, this study attempted to stabilize the fluctuating inflow rate. Then, the influence of the grid resolution and the difference scheme of the advection terms on the computational stability and accuracy was investigated for the LES computations of wind pressure acting on a cubic model. The above investigation provided useful information for setting computational conditions when using LES to predict the wind pressure acting on building walls with sufficient numerical stability and accuracy.

View full abstract

The characteristics of fluctuating pressures and aerodynamic stability of a long-span flat roof have been investigated using proper orthogonal decomposition (POD) and complex proper orthogonal decomposition (CPOD) analyses. The wind pressures on rigid and vibrating roofs were computed using a CFD simulation with LES. Focus was on the distribution along the centerline of the roof. Since the CPOD analysis considers the advection phenomena of the pressure field, the results can reveal the coherent structure of fluctuating pressures. Finally, the aerodynamic stability of the long-span flat roof was examined based on the hysteresis of the dynamic displacement-wind force curve, the energy consumption during a period of the roof's vibration, and the relationship between fluctuating pressure and roof displacement.

View full abstract

This study reveals turbulent field around high-rise building in building cluster, and examines the effect of the turbulent field on fluctuation for sectional wind forces. As a result, peak of frequency appears around fB/U_ref = 0.1 corresponding to period of vortex shedding at 300m height which is not affected by windward buildings. On the other hand, at 200m height, peak of frequency exists at a frequency larger than fB/U_ref = 0.1 and fluctuation of sectional wind force with low frequency exists. The result of reduced order model constructed from the mode of proper orthogonal decomposition(POD) implies that separation shear layer from buildings with smaller width makes the peak frequency larger and the fluctuation with low frequency is induced by turbulent field around several buildings.

View full abstract

The present paper aims to clarify the range of detailed modeling of surrounding buildings that affect mean wind pressure acting on low-rise building roofs. The study takes its basis in wind tunnel experiments with wind tunnel models that are developed with digital data focusing on a 410 m by 810 m area in the southern part of Osaka prefecture. Taking advantage of the digital data, the height and footprint are considered for all wind tunnel models. In addition, the roof shape is considered for buildings in the vicinity of the target buildings. The series of wind tunnel experiments was conducted by changing the range of the detailed modeling of surrounding buildings. The results indicate that the surrounding buildings within approximately 420 m to 600 m can affect the mean wind pressure of target low-rise buildings.

View full abstract

The objective of this study is to simplify the process of generating airflow in turbulent boundary layer wind tunnels using machine learning. A total of 55 sets of roughness element configurations, combining spires, barriers, and roughness blocks, were prepared. The airflow characteristics resulting from these configurations were organized into five target parameters for prediction: the gradient of the mean wind speed, the gradient and magnitude of turbulence intensity, and the gradient and magnitude of the turbulence integral scale. The dataset used comprised airflow data collected in a single wind tunnel, created in the past to target ground surface roughness categories I to IV and model scales ranging from 1/100 to 1/500. Random Forest (RF) and Support Vector Regression (SVR) were employed as the machine learning methods. There was no significant difference in prediction accuracy between the two methods, allowing users to choose whichever is more convenient for them. While prediction accuracy decreases with extreme roughness element configurations, the target airflow characteristics were obtained within a ±20% range of the parameters in the training data, making the method practically sufficient for application.

View full abstract

The objective of this study is to estimate the maximum peak wind force coefficients on rooftop signboards more reasonably. The maximum peak wind force coefficients for I-shaped signboards installed on one side wall of a building were measured through systematic wind pressure measurements, in which parameters such as building shapes, signboard shapes, and installation positions were changed over a wide range. The most critical area for the wind resistance design is the edge of signboard, where the maximum peak wind force coefficient occurs at oblique wind directions relative to the signboard. Flow visualization focusing on these wind angles revealed that the maximum peak wind force coefficient increased as the radius of the vortex generated by the flow separation from the windward edge of signboard became smaller. An empirical formula predicting the maximum peak wind force coefficients is derived as a function of the aspect ratios of the building and signboard, the installation position of the signboard on the building's roof, and the relative size of the signboard to the building. The accuracy of this formula is within ±20% of the experimental values.

View full abstract

Wind tunnel test was examined using house models with photovoltaic panels mounted on a gable roof under different wind turbulence conditions. Negative peak wind force coefficients on the panel near the gable vary with turbulence intensities. Following the wind tunnel test, Large Eddy Simulation (LES) was conducted to investigate the flow fields around the panel when a negative peak wind force appears on the panel. The greater negative peak wind force appears for higher turbulence intensity. The minimum value of the negative peak wind force on the panel near the roof is due to generating a conical vortex from the corner of the eaves and the gable. The vortex is generated by the advection of an air parcel with rotating motion from upwind at around the average height of the roof. The minimum value of the negative peak wind force is described by an equation as the sum of an average wind force and a fluctuating wind force. Design wind loads for photovoltaic panels mounted on a gable roof under different wind turbulence conditions can be predicted using the proposed equation for the various surface roughness categories.

View full abstract

The present paper investigates the design wind loads on mullions installed at the corners of rectangular buildings based on a wind tunnel experiment using building models with various side ratios and aspect ratios. Because the mullion supports corner curtain walls, it is subjected to two mutually orthogonal wind forces at the same time. Building Standard Law of Japan specifies positive and negative peak pressure coefficients for designing the cladding/components of buildings based on the most critical maximum and minimum peak wind pressure coefficients irrespective of wind direction. Therefore, combinations of these peak wind pressure coefficients overestimate the design wind loads on mullions. In the present paper, discussion is made of the combination of wind pressure coefficients on two adjacent walls, which predicts the maximum load effects on the mullions appropriately.

View full abstract

This study investigates the wind flow field around the Seto-Ohashi Bridges during typhoon KROSA in 2019, using the Weather Research and Forecasting Model (WRF). Three WRF Large Eddy Simulation (WRF-LES) models, i.e., 1.5 order Turbulence Kinematic Energy (1.5TKE) closure model, Smagorinsky (SMAG) model, Nonlinear Backscatter and Anisotropy (NBA) model, and YSU Planetary Boundary Layer (PBL) scheme were tested to evaluate their performance for reproducing the wind flow filed around the Seto-Ohashi Bridges. The correlation coefficient between the simulated and observed 10-minute mean wind speed and direction showed no significant difference between the LES models and the YSU PBL scheme. Additionally, no significant variation in gust factor reproduction was observed. During typhoon KROSA passage, the maximum wind speed was observed at Hitsuishijima Viaduct with an approaching east wind, when typhoon centre started to land on Shikoku Island. Owing to the terrain blockage effect of Shikoku Island, significant variation of wind speed and direction along the Seto-Ohashi Bridges was observed, when the wind came from east and southeast. This result highlights the influence of local terrain on the wind flow field of the Seto-Ohashi Bridges.

View full abstract

There is concern about the occurrence of in-phase galloping of two parallel cylinders integrated together, such as communication cables. Therefore, it is desirable to take measures to suppress vibration from an aerodynamic point of view. The authors conducted wind tunnel experiments to evaluate the relationship between the arrangement of the two cylinders with different combinations of the diameters and the characteristics of galloping, especially when the distance between the cylinders is small. In addition, the authors calculated the response amplitude using a time response analysis based on a quasi-steady aerodynamic force model and compared it with the results of free vibration experiments to discuss the applicability of the quasi-steady theory. As a result, larger vibrations occurred when the gap between cylinders was small, and the quasi-steady theory was applied to some extent even to the complex structures consisting of two cylinders.

View full abstract

Reducing the aerodynamic forces of the cables has been a main concern for the design of long cable-stayed bridges. This research investigated the effects of surface modifications, which are the aerodynamic countermeasure against rain-wind-induced vibration, on the aerodynamic forces of and the flow field around cables through wind tunnel tests. According to the measurements of aerodynamic forces and surface pressure, the drag and fluctuating lift coefficients of the spiral protuberance and indent cables were smaller than those of the normal circular cylinder, accompanied by the delay in the flow separation and the recovery of the base pressure. According to Particle Image Velocimetry (PIV) test and Proper Orthogonal Decomposition (POD) analysis, the discretely distributed indent and spiral protuberance increased the surface roughness and suppressed the Kármán vortex shedding, subsequently reducing the wake width and the drag force.

View full abstract

Offshore wind resource assessment requires in-situ wind observation of at least one year. On the other hand, at initial stages of an offshore wind project, it would be useful if annual wind conditions could be estimated based on shorter-term data and used for initial feasibility studies as soon as observation data are collected at the site. This study tried to estimate annual wind conditions from several months of wind observations by using the MCP (Measure-Correlate-Predict) method. The MCP methods tested in this study include those based on linear regression and machine learning. As a result, it was found that the linear regression (No wind direction classification) and the Random Forest difference model have the highest estimation accuracy. For short observation periods of less than six months, it was found that the season of the observation greatly influence estimation accuracy and the accuracy can be worse than the estimation without MCP. In the estimation using the Random Forest, it was found that the estimation accuracy can be greatly improved by constructing a model that uses wind speed differences as the training data, instead of directly estimating wind speeds.

View full abstract

Various attempts and studies have been made to capture wind energy through wind-induced vibration of some structural element. We have developed a novel wind power generation system based on aerodynamic vibrations in the wake flow. In this study, we focused on the flutter phenomenon. Multiple flat plates serve as transducers in our system. The vibration characteristics of these plates under various configurations were investigated, considering the number of plates and the spacing between them. Notably, when three plates were arranged in tandem with unequal spacing, they exhibited the highest torsional amplitude. As a result, it was three times greater than the power output of a single flat plate.

View full abstract

Effective and reliable utilization of wind power requires a deep understanding of both wind conditions and wind turbine wake behavior on the turbine site. Erecting turbines on mountainous areas is expected to provide favorable wind speeds due to the acceleration effect, and to minimize disturbance to humans due to the lower population density. However, studies of wind behavior on mountain area are relatively fewer than different types of complex terrains like hills and forests¹⁾. In this paper, the wind characteristics of two 1.67MW turbines located a wide area of steep mountains are analyzed by Supervisory Control and Data Acquisition (SCADA) system. The purpose of this analysis is to examine terrain effect and turbine wake effect quantitatively in each specific wind direction region. In terrain effects, it shows that slope angle and the distance to crest are key parameters for wind characteristics. For wake effect, it is shown that the measurement in assessing wake effects (40%), and power reduction (40%) were shown. The results are compared with the flow results by a combination midfidelity flow simulation software, WAsP Computational Fluid Dynamics (CFD) for topography, and superimposed wake models for wake effects, respectively.

View full abstract

In recent years, wind power generation has been facing problems such as reduction of power generation and flow fluctuation leading to increase of fatigue loads due to turbine wake on downstream wind turbines. Simulations and other studies have been conducted both in Japan and overseas to investigate these problems. However, there have not been many studies on visualization and evaluation of wake flow for actual wind turbines. In this study, we attempted to visualize the wake of an actual wind turbine using snowfall and to evaluate the flow velocity distribution in the wake using the PIV (Particle Image Velocimetry) method, taking advantage of the geographical characteristics of Aomori Prefecture, which has one of the highest amounts of wind power generation and snowfall in Japan. As a result, the experiments with small wind turbines allowed for wake visualization to 2D (D: rotor diameter[m]) and numerical analysis to 1D.

View full abstract

Wind tunnel experiments were conducted to improve the performance of the magnetostrictive flow-induced vibrational power generator with a cantilevered V-shaped prism. Effects of the test models' cross-sectional height H and V-angle β on the response amplitude and power generation were investigated. The test model was made of PLA and CFRP. The V-shaped prism with β = 90° has a soft galloping vibration. The velocity at the onset of oscillation decreases with increasing β, but the test model for β = 150° does not vibrate. The torsional vibration of a V-shaped prism made of CFRP with a 45/-45 fiber orientation is smaller than with a 0/90 fiber orientation. Flow visualization by the smoke wire method reveals the relationship between vibration characteristics and flow structure around the test model. The power of the test model for β = 90° and H = 40 mm becomes larger at 1.6 mW compared to the other models.

View full abstract

Wind speed and wind direction data collected from an actual site were used to investigate the behavior of wind turbine wakes under real wind conditions. In the simulation, a twodimensional Porous-Disk model was used, and one-second data of wind turbines were utilized as boundary conditions for the calculation to replicate the wind turbine wakes in a real wind field. Although the calculated results showed time deviations from the observed results, they agreed very well when the deviations were corrected. In the actual wind field, the wake of the wind turbines changed significantly due to fluctuations in wind speed and direction, resulting in frequent switching of the time period during which the wakes affected the downstream wind turbines.

View full abstract

A fixed vane is introduced to a vertical axis wind turbine to accelerate the rotor and improve the turbine performance. The vane is located along the rotor circumference with the arm angle α, and α = 40° ~ 60° is the most effective in raising the power coefficient of the turbine. On the other hand, reverse rotation of the turbine was caused at α = -10° ~ -30°. Furthermore, the turbine rotation was totally suppressed at α = 0° and -40°.

View full abstract

The effects of wind speed on the kite flight control in Autonomous Kite-Sailing Power Generation are investigated by conducting computational flight simulations. The flight control method is improved to avoid azimuthal deviation in figure-eight trajectory. The optimal modification rate ensuring high stability in flight control is determined, and a combined equation of angle and angular velocity in kite rolling is derived for varying wind speed.

View full abstract