Reliable validation data are essential for ensuring the accuracy and applicability of computational fluid dynamics (CFD) simulations used to analyze urban wind environments. This data paper presents a comprehensive experimental database curated by a working group of the Architectural Institute of Japan (AIJ) to support CFD validation. The database includes 13 benchmark cases, ranging from simplified building geometries to complex urban configurations, with measurements of wind velocity, turbulence, pollutant (tracer) concentration, and temperature under various flow and thermal conditions. The experimental data were compiled from both existing and newly conducted wind-tunnel and field measurements, collected through collaborations among universities, research institutes, and industry. These cases provide a standardized and versatile basis for evaluating CFD model performance, facilitating uncertainty assessment, inter-model comparison, and best-practice development. Each case includes detailed information on experimental setups and measurement techniques to aid reproducibility and application. All datasets are publicly available through open data repositories and have already been widely used in numerous validation studies worldwide. This data paper aims to encourage continued use and broader dissemination of these benchmark cases, thereby enhancing the reliability and credibility of CFD predictions in urban wind and pollutant dispersion analyses.

A numerical water tank is developed and the predicted floater motions in irregular waves agree well with measurements. The floater motion in combined wave and current conditions is investigated by the developed numerical water tank, and it is clarified that the mean surge motion of the floater increases, and the oscillation amplitude decreases at the surge resonant period. The Quadratic Transfer Functions (QTFs) of potential theory are then modified based on regular and bichromatic wave simulations using the developed numerical water tank. The dynamic analysis is conducted using an engineering model, and the prediction accuracy of floater motion in irregular waves is improved with the modified QTFs by the proposed method. Finally, the drag coefficient in the combined oscillatory and steady flows is evaluated through forced oscillation simulations in steady flow using the developed numerical water tank, and the drag coefficient in the coupled flow model is augmented by considering the representative velocity ratios for oscillatory and steady flows. The resonant surge motion in the combined irregular wave and current predicted with the augmented drag coefficient model shows good agreement with that predicted by the numerical water tank.

In this study, the effect of wind turbine wakes on fatigue loads is evaluated using wake flow fields simulated from on-site measurements and large eddy simulations (LES). An observation-fed flow field reconstruction method based on multiple measured wind speed is proposed to predict fatigue loads, and its accuracy is validated against those measured at a utility-scale wind turbine and estimated using wake fields simulated by LES. First, the influence of the inflow generation method on the fatigue load of a 2.4 MW wind turbine installed at the Choshi offshore wind farm is investigated using time-series data from LiDAR, SCADA, and strain gauges. Subsequently, the fatigue loads of the wind turbine tower at different downstream locations are analyzed using wake flows simulated by LES. The results indicate that the primary cause of increased fatigue loading is the reduction in rotor speed within the wake, which brings the 3P frequency close to the first natural frequency of the tower. Furthermore, the effect of the minimum rotational speed limit in torque control is examined, showing that enforcing this limit prevents the 3P frequency from coinciding with the tower's first mode and effectively mitigates the fatigue load. Finally, a LiDAR-based wake reconstruction method is proposed to estimate fatigue loading for wind turbines operating in wakes. The proposed method demonstrates substantial improvement over existing approaches, achieving the highest accuracy with an 8-point ring-shaped LiDAR scan configuration of radius r = 1.0R.

The capping temperature inversion is a common atmospheric phenomenon that strongly influences the mean flow and turbulence structures within the atmospheric boundary layer (ABL). In this study, full-scale large eddy simulations are utilized to shed light on the characteristics of inversion-capped ABL flows over steep hilly terrain. As atmospheric stratification increases, the vertical wind veer becomes stronger, creating asymmetric flow patterns in the hill wake, and the buoyancy force acts to resist turbulent wake motions. In contrast to the conventionally neutral boundary layer (CNBL) and the convective boundary layer (CBL) cases, the stable boundary layer (SBL) exhibits pronounced flow acceleration at the hilltop and a faster wake recovery on the lee side of the hill. Based on the quadrant analysis, flow separation and vortex shedding are found to enhance organized motions downstream of the hill crest. In the wake region, sweep and ejection motions are identified at different heights. Furthermore, a wind speed prediction approach is developed for inversion-capped ABL flows over steep hilly terrain under both stable and unstable stratifications. The proposed approach incorporates the effect of the free-atmosphere lapse rate. Overall, it shows satisfactory agreement in predicting mean wind speed profiles over steep hills under both CBL and SBL conditions. The overestimation of mean wind speed at the hilltop in the SBL case can be corrected using the sigma coordinate transformation technique.

Fatigue of main bearing significantly affects the reliability of drivetrain system (DTS) and strongly depends on the DTS configuration. Main bearing failure rate is high for Three-Point Mount wind turbine within the designed life. In this study, a three-dimensional elastic DTS model is established and validated with field measurements. The predicted shaft moments and torque arm motions show good agreements with observations. A numerical pounding model of main bearing is then proposed to evaluate the load factor due to pounding. Finally, the formulas in ISO 281 for predicting main bearing rating life L10 are modified by introducing two new parameters: life ratio and load factor. The predicted rating life L10 matches well with the onsite records.

In this study, an accident at Taikoyama wind farm is investigated by the fatigue analysis of high-tension bolts and tower. Firstly, a sophisticated aeroelastic model is proposed by identifying the structural and control parameters and validated with the measured tower base bending moment. The predicted bending moment at the tower top shows that the tensile stress occurs at the downwind side of tower due to the eccentricity of the centre of gravity of rotor and nacelle. The bolt axial force is then predicted using the finite element model of the tower top with consideration of effects of ball bearings, yaw breaks and pinion gear. The predicted bolt fatigue life is about three months when the residual bolt axial force is less than 30 %, which matches the maintenance record. Finally, the stress of the tower shell is investigated by a sophisticated FEM model. It is found that the tensile stress is generated inside of the tower shell due to the leverage effect. The relationship between the local stress and the nominal stress shows the nonlinearity and the local stress in the case with damaged bolts is three times larger than that in the case with intact bolts. The predicted fatigue life of the tower favourably agrees with the observation.

In this study, large-eddy simulations are performed to elucidate the spatiotemporal characteristics and physical mechanisms of turbulent boundary layers over hilly terrain under stable, neutral, and unstable stratification. The impact of thermal stratification on turbulent flows over a steep three-dimensional hill is clarified through flow patterns and statistical characteristics. Compared to neutral stratification, the separation bubble downstream of the hill crest is reduced under unstable stratification, while it is enlarged under stable stratification. In addition, turbulent eddy motions in the wake region are enhanced in the unstable condition but are suppressed in the stable condition. Both mean velocities and turbulence fluctuations over steep hilly terrain are amplified by unstable stratification and attenuated by stable stratification. The flow characteristics on the hill crest are comprehensively determined by the topography and thermal stratification, whereas the flow dynamics in the hill wake are predominantly influenced by terrain-induced turbulence. Moreover, the mechanisms driving the formation of flow fields over steep hilly topography under different thermal stratification are investigated through force balance analysis using the time-averaged Navier-Stokes equations. The results indicate that turbulence plays a negligible role in the force balance upstream of the hill, while it becomes the dominant factor for the force balances downstream of the hill.

In this study, a novel unsteady Reynolds-Averaged Navier-Stokes (URANS) model is proposed in conjunction with a prespecified averaging time and turbulent inflow, and a new peak factor that considers the effect of averaging time used in URANS is derived to calculate gust wind speed. Firstly, the URANS incorporated with a prespecified averaging time and turbulent inflow is proposed to predict the time series of wind speed over flat terrain and investigate the variation of higher-order moments and zero-crossing rate at which the fluctuating wind speed changes algebraic sign with the averaging time. The predicted higher-order moments are less sensitive to the averaging time, but the predicted zero-crossing rate slightly decrease with increasing the averaging time due to the moving average effect. Secondly, a new peak factor is proposed to consider the averaging time. Finally, the gust wind speeds over flat terrain and around a single building predicted using the proposed URANS and the new peak factor based on the Hermite model are found to be in good agreement with the LES results, while those predicted by the conventional models deviate more from the LES results.

This paper presents a high-spatiotemporal-resolution dataset of turbulent flows over two-dimensional (2D) ridges and three-dimensional (3D) hills with smooth and rough surfaces measured in a boundary-layer wind tunnel. This dataset extends the experimental measurements reported in the research article entitled “A wind tunnel study of turbulent flow over a three-dimensional steep hill” by Ishihara et al. (1999). In addition to the original 3D smooth surface case, new measurements on 2D smooth and rough ridges and a 3D rough hill are conducted to clarify the effects of surface roughness and topography on the hill-induced flow field. The three velocity components on multiple representative horizontal and vertical planes around the hilly terrains were measured under different surface roughness conditions. Mean and fluctuating velocity profiles in the wake region are presented along with details of the measurement and data acquisition procedures. This database provides methodological guidance for future wind tunnel studies over complex terrain, including probe arrangement, data acquisition, and experimental design to capture complex mountainous wind fields. Furthermore, it provides a benchmark for validating computational fluid dynamics (CFD) simulations, improving turbulence models, and assessing wind resources in mountainous regions.

In this study, mean and turbulent flow fields around a line of trees and a steep hill are investigated by unsteady Reynolds-Averaged Navier-Stokes (URANS) simulation with a new turbulent inflow generation method. First, an inflow generation method is presented for the URANS model, utilizing a prespecified averaging time to divide the turbulent flow fields into resolved and modelled parts and ensure conservation of the total turbulence kinetic energy. The turbulent flow fields reproduced by the URANS model are then evaluated over flat terrain, a line of trees and a steep hill to demonstrate the performance of the URANS model. A large averaging time can be applied to the URANS model for flat terrain and a line of trees because the turbulent flow field can be simulated well by the turbulence model, but suitable averaging times are required for the URANS simulation to predict large separation vortices behind the steep hill as resolved coherent structures. Finally, an indicator is proposed to assess the performance of turbulence models considering the prediction accuracy and computational efficiency. The proposed method exhibits the best performance in predicting the mean velocity and turbulence kinetic energy, compared to Reynolds-Averaged Navier-Stokes model and Large Eddy Simulation.

In this study, an unsteady Reynolds-Averaged Navier-Stokes (URANS) model with a prespecified averaging time and a new method for turbulent inflow generation is proposed to predict turbulent flows over complex terrain. Firstly, the effect of grid resolution on turbulent flows over complex terrain is investigated by the Reynolds-Averaged Navier-Stokes (RANS) and the URANS models. URANS improves the accuracy of the predicted mean velocity and standard deviation by using finer grids, but the improvement by RANS is limited. Furthermore, the turbulent flows over hills with different slopes predicted by URANS are examined with respect to various averaging times. An optimal averaging time based on the slope of the hills is recommended considering the prediction accuracy and computational efficiency of URANS. Finally, turbulent flows over complex terrain in coastal areas are investigated by URANS and validated by wind tunnel tests. The predicted mean and standard deviation of streamwise velocity over complex terrain by URANS are in good agreement with the experimental data, while those by RANS are overestimated or underestimated.

In this study, a novel dynamic wake model is proposed to predict real-time wind speed and power production by incorporating a new wake propagation velocity model and a new wake deflection model and is validated by numerical simulations and wind tunnel tests. Firstly, the unsteady Reynolds-Averaging Navier-Stokes (URANS) model is used to evaluate the dynamic wake model and validated by phase-averaged LES results. The wake propagation velocity model is proposed based on the results of URANS simulations considering various operational and inflow conditions. It is found that the propagation velocity of the wind turbine wake is smaller than half of the ambient wind speed in the near wake region and asymptotically approaches approximately 0.65 times the ambient wind speed in the far wake region. The new wake deflection model for a yawed wind turbine is then derived from the momentum conservation equation and the double-Gaussian wake model and is validated by experimental results and numerical simulations. The new wake deflection model shows a significant improvement in prediction accuracy of wind speed and power production in both near and far wake regions. Finally, the proposed wake deflection and propagation velocity models are incorporated into the dynamic wake model and validated by numerical simulations with time-varying yaw angle and wind speed. The normalized root mean square error (NRMSE) of power production of downstream wind turbine at x = 7D predicted by the proposed dynamic wake model is reduced from 15.05 % to 1.89 % for the simulation of time-varying yaw angle and from 16.30 % to 7.35 % for the simulation of time-varying wind speed compared to that using the conventional dynamic wake models.

In this study, the dynamic loads on the wind turbine support structures caused by the combined wind and earthquake are investigated by means of the coupled and uncoupled analysis approaches. A series of numerical models are prepared for a 2.4 MW gravity foundation supported wind turbine including the aero-elastic model, the finite element model, the generation of earthquake inputs and the parameters in the sway-rocking model. The standard lumped mass model with a lumped dashpot at the hub height show good agreement with the aero-elastic model in terms of the modal frequencies and modal shapes. The combined seismic and aerodynamic loads are then investigated using the coupled analyses for various operational scenarios including normal operation, emergency stop and parked condition, in which effects of the wind and earthquake misalignment are considered. When the misalignment equals to 0 degree, the emergency stop yields the maximum tower base moment while when the misalignment reaches 90 degree, the normal operation outputs the maximum loads. The uncoupled analysis approach becomes appealing since it can reduce computational costs significantly and be simper for civil engineers involved in site specific permitting. A series of uncoupled analyses are then performed to investigate the rules to combine the seismic and aerodynamic loads. The vector sum method for the normal operation and parked condition and the SRSS method for the emergency stop are proposed to combine the seismic and aerodynamic loads in the uncoupled analysis approach. The proposed combination criteria show reasonable agreement with the coupled analysis approach.

The crosswind stability of railway vehicles, considering distinct national standards in both quasi-steady and unsteady wind conditions, like Chinese hat gust wind (EN 14067-6, 2010), is systematically evaluated. Initially, it is observed that the quasi-static analysis (QSA) proposed in Japan marginally underestimates the wheel unloading ratio of railway vehicles by approximately 3 %, as the external forces increase beyond a certain threshold in quasi-steady winds due to the neglect of the vertical degree of freedom and the inability to accurately evaluate the activation of vertical bump stops. Subsequently, to accurately assess the crosswind responses of railway vehicles under realistic conditions, a non-linear lateral acceleration model is proposed to account for the effects of track irregularities, validated against field test data. Finally, it is noted that the Chinese standard is more conservative, with characteristic wind speeds (CWC) approximately 2 m/s lower than those calculated by the European standard, while the Japanese guideline is more stringent at high train velocities but more lenient at low velocities. The CWC evaluated under gust wind conditions is around 2.4 m/s higher than those obtained under quasi-steady winds, due to the maximum gust wind speed is low-pass filtered by the centered moving average method.

The Kugino wind farm at Japan was seriously damaged in the severe Kumamoto earthquake, characterizing as all three pile group cracks but only one tower buckling. This study aims to reveal the failure mechanism underlying such damage pattern through the Beam on Nonlinear Winkler Foundation (BNWF) analyses, where the soilfooting interaction is considered with a new q-z model (QzSimple6). It identifies three parameters in a hyperbolic function to match any desired modulus reduction curve, whereas adjusts the unloading-reloading curves iteratively with the Ishihara-Yoshida rule to achieve site-specific soil damping curve. The QzSimple6-based BNWF analyses quantitatively reproduces centrifuge test results of a pile group foundation system, and newly reveals the soil-footing interaction does not influence pile bending moments but reduces the point mass acceleration. A parametric study is conducted on the full BNWF model with identifying pile group supported wind turbine, but with scaling soil stiffness and strength. The thrust force is attracted from the aero-elastic analysis in OpenFAST and the free-filed seismic displacement are calculated with the site response analysis in OpenSees. The simulation shows consistency with site observations that the No.2 wind turbine tower is destined to buckling at the height of around 13.9 m due to the sudden reduction of tower thickness, while No.1 and No.3 towers could remain safe potentially because soil properties under them are softer than that under the No.2 tower. In contrast, all three pile groups are found to be cracked under the Kumamoto earthquake intensity since the pile bending moment relies on the footing rigidity rather than the footing-soil interaction.

This study proposes a microscale flow model to estimate mean wind speed, fluctuating wind speed and wind direction over complex terrain considering the effects of topography, atmospheric stability, and turbine wakes. Firstly, the effect of topography is considered using Computational Fluid Dynamics (CFD). Next, a mesoscale model is presented to account for the effect of atmospheric stability. The effect of turbine wakes on the mean and fluctuating wind speeds are then represented by an advanced wake model. The model is validated using the measurement data of a wind farm located in the North of Japan. The measured wind data by Lidar at a reference height are horizontally extrapolated to a nearby met mast hub height and validated by a cup anemometer. Moreover, a novel averaging method is proposed to calculate a directional equivalent Monin–Obukhov length scale to account for the effect of atmospheric stability. Finally, the measured wind data at the reference height are vertically extrapolated and validated at the lidar location. The predicted mean and fluctuating wind speeds show good agreement with the measurements.

The capital expenditure (CAPEX) for the fixed-bottom offshore wind farm is assessed using a probabilistic en-gineering cost model and the cost reduction scenarios in Japan are analyzed. Firstly, the engineering cost model is described to assess the capital expenditure. A new export cable length model is also proposed considering the landing point distance and the vessel size model is proposed as the function of turbine rated power. The proposed engineering cost model succeeds in explaining the mechanism of the increase and decrease of CAPEX experienced in the UK. The uncertainties of model parameters are identified from the reported data and modeled by the normal distribution function. The workability is predicted using the discrete event simulation. The predicted CAPEX is then compared with the existing 30 fixed-bottom offshore wind farms in the United Kingdom. The predicted mean and standard deviation values of CAPEX show good agreement with the reported ones, while the conventional parametric model underestimates the mean value and cannot predict the standard deviation. Finally, the cost reduction scenarios and their uncertainties of offshore wind farms in Japan are analyzed using the proposed probabilistic engineering cost model. The levelized cost of wind energy reduced from 20.0 JPY/ kWh to 17.0 JPY/kWh, 13.6 JPY/kWh and 10.1 JPY/kWh by the reduction of installation days using the specific installation vessel, the turbine enlargement and the improvement of operation and maintenance efficiency. The predicted supply prices for each cost reduction scenario agree well with those reported at the first auction conducted in 2021 in Japan.

In this study, the flow-induced vibration of two-degree-of-freedom staggered circular cylinders at subcritical Reynolds numbers is studied numerically with overset dynamic mesh. The streamwise spacing ratio Sx/D of 4 and the transverse spacing ratio Sy/D ranging from 0 to 2 are employed to explore the influence of staggered arrangement on the vibrations and fluid forces of cylinders. It is found the upstream cylinder mainly experiences the vortex-induced vibration while vortex-induced and wake-induced vibrations happen for the downstream one, which may be interfering or independent depending on the transverse spacing of cylinders. The phase lag between the displacement and fluid force of the downstream cylinder changes from 0° to 180° with the increasing of inflow velocity, which leads negative add mass and serves as the reason of wake-induced vibration. The theoretical formulae to estimate the critical velocity, fluid coefficient and vibration amplitude of wake-induced vibration are proposed and validated by the numerical simulations.