Next Article in Journal
Multi-Objective Optimization Design of Foamed Cement Mix Proportion Based on Response Surface Methodology
Previous Article in Journal
Investigation of Fresh, Mechanical, and Durability Properties of Rubberized Fibre-Reinforced Concrete Containing Macro-Synthetic Fibres and Tyre Waste Rubber
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Investigation of Wind Pressure Dynamics on Low-Rise Buildings in Sand-Laden Wind Environments

1
College of Hydraulic and Civil Engineering, Xinjiang Agricultural University, Urumqi 830052, China
2
Xinjiang Key Laboratory of Hydraulic Engineering Security and Water Disasters Prevention, Urumqi 830052, China
*
Author to whom correspondence should be addressed.
Buildings 2025, 15(15), 2779; https://doi.org/10.3390/buildings15152779
Submission received: 30 June 2025 / Revised: 22 July 2025 / Accepted: 2 August 2025 / Published: 6 August 2025
(This article belongs to the Section Building Structures)

Abstract

To enhance the structural safety in wind-sand regions, this study employs the Euler-Lagrange numerical method to investigate the wind pressure characteristics of typical low-rise auxiliary buildings in a strong wind-blown sand environment. The results reveal that sand particle motion dissipates wind energy, leading to a slight reduction in average wind speed, while the increase in small-scale vortex energy enhances fluctuating wind speed. In the sand-laden wind field, the average wind pressure coefficient shows no significant change, whereas the fluctuating wind pressure coefficient increases markedly, particularly in the windward region of the building. Analysis of the skewness and kurtosis of wind pressure reveals that the non-Gaussian characteristics of wind pressure are amplified in the sand-laden wind, thereby elevating the risk of damage to the building envelope. Consequently, it is recommended that the design fluctuating wind load for envelopes and components of low-rise buildings in wind-sand regions be increased by 10% to enhance structural resilience.

1. Introduction

With the continuous advancement of China’s Belt and Road Initiative, the locational disadvantages of the arid and semi-arid regions in the western part of the country are gradually being alleviated. The extensive areas of desert and Gobi, coupled with abundant solar and wind energy resources, provide a unique geographical advantage for the development of the new energy sector, including photovoltaic and wind power industries [1]. In 2022, the National Development and Reform Commission and the National Energy Administration in China jointly issued the “Planning and Layout Scheme for Large-scale Wind Power and Photovoltaic Bases Focused on Desert, Gobi, and Arid Regions” to promote the construction of large-scale new energy base projects in the “Deserts-gobi-desert” region [2]. The structural safety of numerous substations, distribution rooms, equipment rooms, and low-rise buildings for offices and residences, which are constructed to support transportation, energy, and other foundational and emerging industrial projects, is a prerequisite for ensuring the safe and stable operation of the project systems.
In recent years, global warming has led to an increase in both the frequency and intensity of wind disasters. Numerous investigations into wind disasters have shown that low-rise buildings are among the most vulnerable types of structures to damage during such events [3,4]. Wind engineering researchers have conducted extensive and productive studies on the wind field characteristics around low-rise buildings in various extreme wind environments, including the seasonal monsoon atmospheric boundary layer [5], typhoons [6], tornadoes [7,8], and downbursts [9,10], as well as the wind pressure characteristics on building surfaces. However, there is a limited amount of literature addressing the load characteristics of building structures in wind-sand environments [11,12,13,14,15,16].
The process of strong wind-sand transport, such as sandstorm, is a catastrophic weather phenomenon that occurs uniquely in arid and semi-arid regions. This process poses substantial risks to civil engineering, transportation, and municipal infrastructure through several mechanisms: wind-sand erosion undermines the efficacy or integrity of protective systems [17]; sand accumulation on roadways precipitates vehicle overturning [18,19]; sand burial of railway tracks leads to train derailments, while high-velocity sand particle impacts shatter train windows, disrupting rail operations [20]; and sand particle erosion damages pipelines [21], buildings [22], photovoltaic modules [23], and concentrated solar power equipment [24], resulting in structural damage or abrasive wear on material surfaces. In 2024, for instance, Xinjiang, China, experienced two severe sandstorms with intensities reaching level 15, resulting in significant damage to renewable energy infrastructure [25,26]. Consequently, wind-sand disasters represent a risk factor that cannot be overlooked during the design, construction, and operational phases of foundational engineering facilities in wind-sand areas.
In strong wind-blown sand environments, the wind speed gradient is large, and turbulence intensity is high, accompanied by intense turbulent mixing phenomena [27]. According to Wang et al. [28], based on field multi-point spatial synchronous observations conducted at the Qingtu Lake Observation Array (QLOA) in Gansu, China, multiple sets of results for sand-laden and impurity-free wind flows under near-neutral conditions were selected. The statistical results of turbulent kinetic energy in wind fields within similar Reynolds number ranges indicate that the streamwise turbulent kinetic energy in sand-laden wind (field with sand particles) is significantly higher than that in impurity-free wind (field without sand particles), and the same conclusion applies to vertical turbulent kinetic energy. This indicates that the presence of sand particles significantly enhances the turbulent kinetic energy of the flow field, resulting in widespread damage to building envelopes in strong wind-blown sand environments [29].
Wind load, a pivotal factor in structural design, has consistently attracted significant attention and is typically categorized into mean and fluctuating wind load. The mean wind load, induced by the average wind speed, can be treated as a static load in simplified engineering analyses. Conversely, fluctuating wind load stems from instantaneous variations in wind speed, namely turbulence effects, which trigger dynamic structural responses. The repetitive action of these loads can cause structural vibrations and fatigue damage, posing a threat to structural safety. With the continuous advancement of computational fluid dynamics (CFD) technology and the rapid development of computer hardware and software, CFD has emerged as an essential tool for simulating wind and wind-sand effects on structures [30]. Sandstorms are fundamentally gas-solid two-phase flows occurring within the atmospheric boundary layer (ABL). Balachandar and Eaton [31] provided a comprehensive review of the applicability of various numerical methods in multiphase flow computations, correlating them with particle Stokes number (St) and volume fraction. In the wind-sand two-phase flow problem investigated in this study, the St of sand particles is generally greater than 0.2 [28]. Both the Euler-Euler (E-E) and Euler-Lagrange (E-L) methods can be employed to describe wind-sand motion. The E-E method treats both airflow and sand particles as continuous phases, enabling accurate depiction of the motion of small particles with good flow-following behavior. However, its accuracy diminishes for larger particles with poor flow-following characteristics, potentially compromising the precise representation of the actual force state on engineering structures in wind-sand environments. In contrast, the E-L method solves airflow motion using the Navier-Stokes (N-S) equations and tracks particle motion individually, thereby providing more comprehensive information on particle and fluid dynamics, albeit at the cost of greater computational resources [32].
Previous numerical studies on the impact of wind-sand interactions on buildings [11,12,13,14] have primarily focused on the effects of sand particle collisions on the average load-bearing capacity of engineering structures. Huang et al. [15,16] conducted wind-sand tunnel experiments to investigate the influence of wind speed and sand particle concentration on the overall pulsating load effects on building models. However, these studies did not account for the fluctuating wind loads induced by wind disasters that cause damage to envelope structures and connecting components. In reality, compared to the failure of a building’s primary structure, damage to enclosure structures-such as roofs, walls, and insulation layers-caused by fluctuating load effects is far more prevalent in wind disasters, as illustrated in Figure 1. Furthermore, localized openings resulting from enclosure structure failures (Figure 1a) may exacerbate the risk of overall building collapse under the combined influence of internal and external pressures [33]. Consequently, the fluctuating load effects on engineering structures in wind-sand environments warrant greater emphasis.
Overall, research on the combined effects of wind and sand on engineering structures remains significantly limited, with notable deficiencies in understanding the force characteristics of structures in wind-sand environments and the theoretical framework for wind-sand-resistant structural design. In light of this, this study focuses on low-rise buildings prone to damage in intense wind-sand environments. By integrating E-L two-phase flow numerical simulations with theoretical analysis, this study conducts a detailed investigation into the wind pressure distribution characteristics of engineering structures under sand-laden wind. A key innovation of this work lies in providing a mechanistic explanation of the variations in wind pressure characteristics in wind-sand environments from the perspectives of flow field energy changes and sand particle-structure interaction mechanisms. Furthermore, the study proposes design amplification factors for fluctuating wind load effects in wind-sand regions. The findings of this work hold significant practical importance for deepening the understanding of load characteristics of engineering structures under combined wind-sand effects, and enhancing structural safety in intense wind-sand environments.

2. Numerical Simulation Methods

2.1. Governing Equations

The wind field is simulated by solving the three-dimensional incompressible N-S equations, coupled with the Lagrangian particle tracking method to model interactions between the wind field and moving particles [34]. The continuity and momentum equations, accounting for particle reaction forces, are expressed in a Cartesian coordinate system as follows
u i x i = 0
u i t + u j u i x j = ρ p x i + ν 2 u i x j x j + F i
where ui (i = 1, 2, 3) represents the velocity component in the streamwise, lateral, and vertical directions, respectively; p, ρ = 1.225 kg/m3 and v = 1.5 × 10−5 m2/s denote the pressure, density and kinematic viscosity of the air, respectively; and Fi is the additional body force. Since the effects of heat on airflow and particle motion are not considered, the energy equation is omitted.
Assuming the sand particles are spherical, they exert a reaction force Fi on the airflow, equal in magnitude and opposite in direction to the drag force exerted by the airflow on the particles. This force is included as a source term in the Equation (2) to model two-way coupling between the solid particles and the airflow [35]. The expression for Fi is given below
F i = - V cell 1 n p = 1 n pcell 0.5 C dp A p u - u p ( u i - u p i )
where Vcell is the volume of the finite control body; npcell is the number of particles within the control body; Ap is the projected area of the particle in the flow direction; and u and up are the velocity of the airflow and the particles, respectively.
Sand particle motion is influenced by gravity, buoyancy, drag, Saffman lift, Magnus force, and Basset force. However, the magnitudes of the Saffman lift, Magnus, and Basset forces are significantly smaller than those of the gravity and drag forces acting on the sand particles [36]. Therefore, this study considers only the effects of gravity, buoyancy, and drag forces. According to Balachandar and Eaton [31], thewind-sand flow is considered a dilute gas-solid two-phase flow (with a particle volume fraction significantly less than 10−3), and particle-particle collisions in the airflow are neglected in the numerical calculations. Sand particle motion follows Newton’s second law, with the equations of motion expressed as follows
m p d u p i d t = π ( ρ p ρ ) d p 3 g 6 + 0.5 C dp A p u - u p ( u i - u p i )
where mp is the mass of a sand particle; ρ p = 2650 kg/m3 is the density of the sand particles; g = 9.81m/s2 is the gravitational acceleration; dp is the particle diameter; and π ( ρ p ρ ) d p 3 g / 6 is the sum of the gravitational force Fg and buoyant force Fb acting on each particle. The drag coefficient of the particle,
C dp = 24 Re p + 6 ( 1 + Re p ) + 0.4   ,      Re p 1000 0.424   ,            Re p > 1000
is defined according to Clift [37], and Rep is the particle Reynolds number,
Re p = u - u p d p / ν .

2.2. Computational Domain, Boundary Conditions, and Meshing

Owing to the significant computational resource requirements of numerical simulations for aeolian sand transport in gas-solid two-phase flows at the scale of the real ABL, studies on aeolian sand issues are commonly conducted using scaled-down models [35,38]. To obtain more refined analytical outcomes, this research focuses on mechanistic elucidation and is therefore carried out based on scaled-down models.
As shown in Figure 2, the study examines a typical low-rise building with a 15° roof slope. The model is scaled at a 1:10 ratio, with dimensions of 500 mm in length (L), 300 mm in width (W), and 440 mm in height (H). Due to its bi-axial symmetry (symmetric along both x and y axes), the incoming wind direction varies counter clockwise from 0° to 90° in 15° increments, resulting in seven conditions (0°, 15°, 30°, 45°, 60°, 75°, 90°). The 0° angle is perpendicular to the longer edge of the roof, as illustrated in Figure 3. The computational domain measures 36L (x) × 7L (y) × 5L (z), maintaining a blockage ratio below 3.0% for all wind directions, as recommended for numerical simulations in Computational Wind Engineering (CWE) [39]. The model is positioned 26L from the inlet to ensure adequate mixing of sand particles and the wind field, forming a uniform wind-sand flow field. The wake region extends more than 10H downstream to allow the flow to reach a fully developed state.
To numerically investigate wind or wind-sand interaction effects on engineering structures, accurate definition of the turbulent inlet boundary conditions for the ABL is essential [40,41,42]. This study employs the Divergence-Free Spectral Representation (DFSR) method [43] to generate fluctuating wind velocity time series at the inlet plane nodes, superimposed on the mean wind speed profile to define the inlet boundary conditions. The outlet is set as a free outflow boundary, assuming that the normal gradients of flow field variables (e.g., velocity and pressure) are zero at the outlet, and symmetry boundaries are applied to the lateral and top surfaces. The bottom of the computational domain and the model surface use a no-slip wall. Wall analysis is conducted using the nutURoughWallFunction in OpenFOAM [44], which employs a velocity-corrected wall model for turbulent viscosity coefficients, suitable for modeling rough wall surfaces [45]. Sand particles are uniformly released in the horizontal direction at the inlet, with rebound conditions applied for particles interacting with the no-slip wall, while all other boundaries are designated as escape boundaries.
The computational domain grid was generated using BlockMesh for structured meshing and SnappyHexMesh for hexahedral mesh refinement. To ensure good self-sustainability of the turbulent characteristics within the computational domain, uniform grid resolution was applied from the inlet to the model. Local grid refinement was implemented in the near-wall region, with the first boundary layer mesh height approximately 0.0014 m (y+ ≈ 30), meeting the wall function requirements [46]. The boundary layer mesh growth ratio is set to 1.05. In regions farther from the model, a coarser mesh is used. The total number of grid cells for the 0°, 15°, 30°, 45°, 60°, 75°, and 90° wind direction models ranges from approximately 3.054 million to 3.069 million. Figure 3 shows the overall and local mesh distribution for the 0° wind direction model, with consistent mesh partitioning applied to all other wind direction models.

2.3. Turbulence Modeling and Numerical Configuration

This study utilizes the Improved Delayed Detached-Eddy Simulation (IDDES) method, built upon the Spalart-Allmaras (S-A) turbulence model [47]. This hybrid approach integrates Reynolds-Averaged Navier-Stokes (RANS) modeling near the wall to capture small-scale turbulent motions with Large Eddy Simulation (LES) in flow separation regions to resolve large-scale turbulent structures, thereby ensuring both accuracy and computational efficiency in simulating complex turbulent flows [48]. The method effectively balances computational cost and precision, making it well-suited for applications in turbulent flow research, such as building wind environment simulations [49], near-field pollutant dispersion studies [50], and wind-sand movement [34].
In the S-A model, only one additional equation for eddy diffusivity transported quantity ν ¯ , replaces the length scale term d in the transport equation with the IDDES length scale l IDDES = f ˜ d ( 1 + f e ) l RANS + ( 1 f ˜ d ) l LES , where l RANS = d , l LES = C DES Δ . d is the distance to the nearest no-slip wall. Empirical constant C DES = 0.65 . Δ = max { Δ x , Δ y , Δ z } is the needed sub-grid length scale, Δ x , Δ y , and Δ z are the local streamwise, lateral, and vertical cell sizes, respectively. The function f ˜ d serves as a blending mechanism between RANS and LES modes. To mitigate the reduction in turbulence at the interface within the boundary layer separating these flow regimes, a boosting function f e is incorporated into the length scale definition. This adjustment increases modeled turbulence to compensate for the reduced analytical turbulence.
Numerical computations employed a second-order implicit Backward scheme for temporal discretization. The convective and gradient terms were discretized using a second-order Gauss linear scheme. The diffusion term was discretized with the Gauss linear scheme and modified to enhance stability. The PISO (Pressure-Implicit with Splitting of Operators) algorithm was utilized to achieve pressure-velocity coupling, with convergence determined by a residual threshold of 10−6. An unsteady wind field was simulated using a time step of 0.001 s, ensuring that the average Courant-Friedrichs-Lewy number remained below 1 across all grid cells. The total simulation duration was 23 s, with data from the final 20 s used for statistical analysis.

2.4. Wind Field and Sand Particle Configuration

This study establishes a turbulent wind field based on the field-measured wind characteristics in desert regions by Huang et al. [16]. According to the monitoring results of all suspended sand particles during storms in QLOA, as reported by Liu et al. [51], within the height range of 0 to 30 m near the ground, the average particle size of sand particles ranges between 70 µm and 105 µm (as shown in the shaded area of Figure 4, with particle size decreasing as height increases). Since this study focuses on low-rise buildings, the average particle size of sand particles is set to 105 µm, with a standard deviation of 30 µm, and the particle size distribution follows a log-normal distribution [52], closely aligning with the measured particle size distribution results presented in Figure 4. The monitoring results for sand particle mass concentration during storms range from 10−7 to 10−1 kg/m3, and within the height range of low-rise buildings (<5.0 m), the sand particle concentration is between 10−4 and 10−3 kg/m3 [51]. Therefore, the sand particle mass concentration in this paper is set to 1.0 g/m3.

2.5. Data Processing

A total of 648 measurement points was arranged on the model surface to extract the wind pressure time series data. The distribution of measurement points is relatively dense in areas where the edge wind pressure varies significantly, while the spacing between measurement points increases in the central region of the model, ensuring a uniform distribution. The arrangement and numbering of the measurement points on the model surface are depicted in Figure 5. The model defines five surfaces: surface A (roof) with 156 measurement points, the windward (and leeward) surfaces B (and D) with 143 measurement points, and the right (and left) side surfaces C (and E) with 103 measurement points. The measurement point numbering follows the format of “letter + number”; for instance, A1 denotes the first measurement point on the roof. Additionally, two characteristic cross-sections have been defined: the mid-symmetrical plane of the building is designated as section M (1–2–3–4), and the cross-section at a height of 0.6L is designated as section N (5–6–7–8), which will be used for further analysis.
Wind pressure is typically a random process that varies with time, and its characteristics can be described using statistical measures. This study focuses on the mean and fluctuating wind pressures in both clear and sand-laden wind fields, as well as the third-order (skewness) and fourth-order (kurtosis) moments of wind pressure.
(1)
Wind Pressure Coefficients. Wind pressure is characterized by the dimensionless wind pressure coefficient [5], defined as the ratio of the pressure induced by airflow on the building surface to the dynamic pressure of the undisturbed incoming wind speed (Equation (7)). The average wind pressure coefficient Cp,mean (Equation (8)) represents the static effect of wind loads and serves as the basis for the overall structural design; conversely, the fluctuating wind pressure coefficient Cp,rms (Equation (9)) represents the dynamic effects of wind loads, causing structural vibrations and potentially leading to fatigue failure of components. In this study, the eave height is utilized as the normalization height for wind pressure.
C p ( t ) = [ p ( t ) p 0 ] / 0.5 ρ U H 2
C p , mean = t = 1 M C p ( t ) / M
C p , rms = t = 1 M ( C p ( t ) C p , mean ) 2 / ( M 1 )
where p(t) and Cp(t) represent the wind pressure and pressure coefficient at time t, respectively; p0 is the static pressure at the reference point; UH is the average wind speed (time-averaged) at the reference point; and M is the number of extracted time steps. The convention is that positive wind pressure indicates a compressive force on the surface, while negative wind pressure indicates a suction force. Any reference to negative wind pressure values in this paper refers to their absolute values for comparison.
(2)
Skewness and Kurtosis of Wind Pressure. Existing studies have indicated the presence of large-scale vortices-characteristic turbulence (Building Induced Turbulence) –in areas such as the rooftop, side, and leeward surfaces of structures, influenced by the shape of the structure [53]. In regions affected by characteristic turbulence, the assumption of quasi-stationarity [54] is no longer valid, leading to significant non-Gaussian characteristics in the fluctuating wind pressure. The time series of wind pressure is characterized by asymmetric pressure distribution accompanied by large-magnitude wind pressure pulses. These pulses can generate substantial instantaneous suction forces in localized areas of the structural surface, and their repeated action can easily result in structural damage to the enclosing components and their connections, as illustrated in Figure 1. Therefore, it is crucial to pay special attention to regions of wind pressure that exhibit non-Gaussian characteristics.
For a standard Gaussian signal, the skewness is 0 and the kurtosis is 3, and its probability density function can be fully described by the first two statistical moments (mean and variance). The statistical characteristics of non-Gaussian signals cannot be completely described using only the first two moments. The third moment (skewness, Csk) and fourth moment (kurtosis, Cku) of the wind pressure coefficient can reflect the degree of skewness and peakedness of the probability distribution of the random process, respectively. The expressions for Csk and Cku are given as follows
C sk = t = 1 M [ ( C p ( t ) C p , mean ) / C p , rms ] 3 / M
C ku = t = 1 M [ ( C p ( t ) C p , mean ) / C p , rms ] 4 / M .
The length of the sampling time is a critical parameter that affects the statistical results of unsteady turbulent flow simulations. To ensure the reliability of the statistical results, the influence of different sampling durations (ranging from 1 to 30 s) on the statistical outcomes was investigated. Wind pressure time series data were extracted from measurement points A7, A33, and A85 on the roof (2–3) of the model at Section M for the 0° wind direction model under both the clean and the sand-laden wind. Using the wind pressure statistical results from a 30 s duration as a reference, the deviations ε of the average wind pressure and fluctuating values for different statistical durations relative to the 30 s wind pressure results were defined as follows
ε = ( p T p 30 ) / p 30 × 100 %
where pT and p30 represent the wind pressure statistical results for the sampling durations of T seconds and 30 s, respectively, including both the average values and the fluctuating values.
Figure 6 shows the deviation ε of the Cp,mean and Cp,rms at measurement points A7, A33, and A85 for sampling duration T s, relative to the 30 s statistical results. It can be observed that the statistical results of wind pressure in both the clean wind and the sand-laden wind gradually stabilize with the increase in sampling time. A comparison between Figure 6a,b indicates that the Cp,rms results in the sand-laden wind converge more slowly compared to those in the clean wind. This suggests that particle motion further influences the disturbances in airflow, making the fluctuations of the wind pressure pulsation values more pronounced. However, in both conditions-the clean wind and the sand-laden wind-the maximum relative errors for all parameters occur at statistical durations of 16 s and 13 s, respectively, and are –5.67% and 6.67%. Thereafter, the ε of the Cp,mean and Cp,rms remain within 5%. Therefore, a statistical time of 20 s in this study ensures the reliability of the statistical results.

2.6. Validation and Grid Independence Testing

To validate the effectiveness of the numerical computation method proposed in this study, wind pressure data from field experiments on the Texas Tech University Building Model (TTU), obtained by Levitan [5], were utilized. Figure 7 illustrates the comparison between the mean Cp,mean and fluctuating Cp,rms wind pressures coefficients derived from the numerical method and the experimental measurements. The simulated wind pressure results show strong agreement with the experimental data, confirming the reliability of the numerical method for calculating wind pressures on low-rise building surfaces. Additionally, based on wind-tunnel experimental data from Huang et al. [15] under varying wind speeds and sand concentration conditions, comparisons were conducted regarding the concentration distribution of sand particles in the flow field, the wind loads on the model, and the interaction between sand particles and the scaled building model. The results demonstrate good overall consistency between the numerical method and experimental data. The spatial distribution of sand particle concentration obtained from numerical simulations aligns with experimental findings, and the computed interaction between sand particles and the model deviates by 13% from the experimental results, an error deemed acceptable in engineering research. A more comprehensive description of the numerical method validation can be found in Hu et al. [14].
In prior work, to assess the influence of grid resolution on numerical results, a grid independence study was performed using a model with a 0° wind angle as a case study. Three grid configurations with different refinement levels were employed: a coarse grid (1.72 million cells), a basic grid (3.07 million cells), and a fine grid (6.10 million cells). The relative error in wind pressure between the basic and fine grid models was found to be within 3% [14]. Consequently, the wind pressure results obtained using the basic grid scheme in this study are considered converged.

3. Results and Discussion

The reference wind speed calculated in this study is 7.98 m/s, defined as the time-averaged wind speed at a height of 2L in the clear wind. For the seven wind direction models, “Angle + I” and “Angle + S” denote the computational results for the respective direction conditions in the clear wind and the sand-laden wind, respectively. For instance, “0–S” represents the computational results for the 0° wind direction model in the sand-laden wind.

3.1. Wind Field Characteristics

The characteristics of the turbulent wind field and the mass concentration distribution of sand particles are crucial for understanding the sand-laden wind field. The monitoring location is situated 2L in front of the model (as shown in Figure 2). The monitored mass concentration of sand particles at the eave height (0.8L) in the empty wind field is 1.097 g/m3, which satisfies the measured range of sand particle concentration (10−4 to 10−3 kg/m3) for low-rise buildings (<5.0 m) [51]. Figure 8 presents a comparative analysis of the characteristics of the clean wind and the sand-laden wind. The average wind speed profile indicates that the wind speed in the sand-laden wind decreases to some extent compared to that in the clean wind, with a more pronounced reduction at lower heights, as shown in Figure 8a. Under the influence of gravity, the concentration of sand particles in the lower part of the wind-sand flow field is relatively higher, leading to a more significant attenuation of wind speed. However, the overall sand mass concentration in the flow field remains low, resulting in a non-significant absolute change. Here, only the average wind speed values at the eave height (0.8L) are provided: 6.14 m/s for the clean wind and 6.00 m/s for the sand-laden wind.
The results of the streamwise fluctuating wind velocity presented in Figure 8a indicate that within the height range of 0 to L, the root mean square (RMS) of fluctuating wind velocity in the sand-laden wind is significantly greater than that in the clean wind at the same height. This suggests that in regions of higher sand particle concentration at low altitudes, the movement of sand particles interferes with the airflow, with the enhancement of fluctuations being more pronounced at lower heights. However, there is no significant impact on the wind field above height L.
Figure 8b shows the power spectrum of the fluctuating wind at the eave height (0.8L). The results indicate that in the low-frequency range (less than 10 Hz), the power spectrum curves of the sand-laden wind and the clean wind do not exhibit significant differences. However, in the high-frequency range (greater than 10 Hz), the power spectrum values of the sand-laden wind increase to a certain extent. This indicates that the increase in fluctuating wind velocity is a result of the enhanced disturbance caused by sand particles, which amplifies the energy of small-scale vortices, leading to an increase in high-frequency fluctuating components within the flow field. Wang et al. [28] based on extensive long-term monitoring results from sandstorm sites, found that the enhancement of small-scale kinetic energy in wind-sand flows is more pronounced than that of large-scale vortices. This is qualitatively consistent with the findings of the present study.
The vortex structures of the flow fields in the clean wind and the sand-laden wind are identified based on the Q criterion. Q is defined as a secondary invariant of the velocity gradient tensor, represented as follows: Q = 0.5 ( Ω i j Ω i j S i j S i j ) . When Q > 0, it indicates that the rotation rate is greater than the strain rate of the fluid, suggesting that vortex structures dominate the flow field. Comparing the vortex structure images in Figure 9 for the clean wind and the sand-laden wind (where Q = 170 s−2 and colored by dimensionless instantaneous wind speed), it can be observed that the vortex structures in the clean wind (Figure 9a) are distributed relatively uniformly along the flow direction x/L, with no significant changes in the turbulent structure. In the sand-laden wind, the results in Figure 9b indicate that with the addition of sand particles, the vortex structures in both the streamwise x/L and vertical directions z/L gradually intensify (stabilizing at the 20L streamwise position), leading to increased turbulent fluctuations, as shown in Figure 8a.

3.2. Wind Pressure Characteristics

This section presents results related to wind pressure in both clear and sand-laden wind fields. Due to space constraints, Section 3.2.1 provides only the average and fluctuating wind pressure results for two representative characteristic sections (M and N in Figure 5). In Section 3.2.2, only the skewness and kurtosis cloud maps of wind pressure for the 0° (head-on wind) and 45° (oblique wind) models are presented, while the results for other wind angles are detailed in Table 1.

3.2.1. Average Wind Pressure and Fluctuating Wind Pressure

The average wind pressure coefficients on the two characteristic sections in Figure 10 vary with the wind direction angle: at a 0° wind direction, where the wind is parallel to section M, the windward region (1–2) experiences positive wind pressure, with the maximum mean negative pressure occurring at the flow separation point on the roof (2–3) near the windward edge. At a 15° wind direction, the mean wind pressure distribution characteristics are similar to those at 0°, but the absolute values are relatively reduced. When the wind direction increases to 30°, the airflow at the ridge begins to exhibit secondary separation characteristics, resulting in a more pronounced negative pressure region on the roof (2–3) behind the ridge. Consequently, the area near the ridge is identified as a high-risk zone for structural envelope damage. Under the 45°, 60°, and 75° wind direction conditions, secondary separation characteristics are consistently observed, but the absolute values of the mean wind pressure progressively decrease, indicating that the 30° wind direction represents the most critical condition for damage near the ridge. At a 90° wind direction, as shown in Figure 10c1, where the wind is perpendicular to section M, the wind pressure distribution on section M is nearly uniform and symmetric, with relatively small negative pressure values.
At a 0° wind direction, where the wind is perpendicular to the (5–6) plane of section N, the maximum mean negative pressure occurs at the windward leading edges of the side surfaces (6–7 and 5–8), where intense cylindrical vortex shedding results in significant mean wind suction forces. At a 15° wind direction, the side surface (5–8) no longer experiences strong cylindrical vortex shedding, leading to a reduction in negative pressure values. As the wind direction angle increases, the positive pressure region on the (6–7) plane gradually expands, while the positive pressure region on the (5–6) plane progressively diminishes. By a 45° wind direction, a negative pressure region begins to form on the (5–6) plane in areas farther from the incoming flow. At wind direction angles of 75° and 90°, the (6–7) plane becomes the windward face, with (6–5) and (7–8) transitioning to side surfaces. Similarly, due to cylindrical vortex shedding, the maximum mean negative pressure is generated at the leading edges of these side surfaces, as depicted in Figure 10c2.
The fluctuating wind pressure coefficients on the two characteristic sections in Figure 11 exhibit distinct variations with increasing wind direction angles. At a 0° wind direction angle, where the wind is parallel to section M, the windward region (1–2) experiences positive pressure with insignificant fluctuating wind pressure values. However, at the windward leading edge of the roof (2–3), flow separation and the influence of separation vortices result in the maximum fluctuating wind pressure coefficient at the midsection of the windward roof, where the mean negative pressure is also substantial (as shown in Figure 10a1). Consequently, the windward leading edge is identified as one of the most vulnerable areas for damage to the roof’s structural envelope, as depicted in Figure 1a. At a 15° wind direction angle, the wind pressure distribution characteristics resemble those at 0°, but the fluctuating wind pressure on the roof decreases due to reduced vortex shedding intensity compared to the 0° condition. As previously discussed, at wind direction angles of 30°, 45°, 60°, and 75°, secondary flow separation occurs at the ridge in the midsection of the roof (2–3), leading to a pronounced peak in fluctuating wind pressure in the windward leading-edge region of the leeward roof under the influence of separation vortices. At a 90° wind direction angle, where the wind is perpendicular to section M, the fluctuating wind pressure distribution on section M is relatively uniform.
At a 0° wind direction angle, where the wind is perpendicular to the (5–6) plane of section N, the maximum fluctuating wind pressure occurs in the separation vortex regions at the windward leading edges of the side surfaces (6–7 and 5–8). Concurrently, vortex shedding at these locations results in significant mean wind suction forces (as shown in Figure 10a2). The combined effect of fluctuating wind pressure and mean suction renders this region one of the most susceptible to damage on the side walls, as illustrated in Figure 1b,c. As the wind direction angle increases, the side surface (5–8) gradually moves away from the incoming flow, leading to reduced vortex shedding intensity and diminished fluctuating wind pressure. Conversely, the vortex shedding intensity at the windward leading edge of (6–7) becomes stronger compared to the 0° condition, resulting in increased fluctuating wind pressure. At a 30° wind direction angle, the location of maximum fluctuating wind pressure shifts to the flow separation point along axis 6, with similar results observed for the 45°, 60°, and 75° conditions. At 90°, the (6–7) plane becomes the windward face, while (6–5) and (7–8) transition to side surfaces. Similarly, due to cylindrical vortex shedding, the fluctuating wind pressure is elevated at the separation vortex regions formed at the leading edges of these side surfaces.
The average wind pressure coefficients obtained in the sand-laden wind, as shown in Figure 10, exhibit no significant differences compared to those in the clear wind, indicating that mean wind pressure is minimally affected by turbulent fluctuations. However, the enhanced turbulent fluctuations in the sand-laden wind (Figure 8b) result in notably higher fluctuating wind pressure coefficients in Figure 11 compared to the clear wind, particularly in the windward regions of various models. This is primarily due to the significant influence of the incoming flow’s turbulent fluctuation characteristics on the fluctuating wind pressure in the windward regions, which generally aligns with the quasi-steady assumption. As discussed in Section 3.1 regarding the characteristics of the sand-laden wind field, the motion of sand particles amplifies the fluctuating effects within the flow field, thereby increasing the fluctuating wind pressure acting on the windward regions of the structure. In contrast, the leeward roof, side walls, and leeward regions are dominated by vortex motions induced by separated flows and wakes, where fluctuating wind pressure deviates from the quasi-steady assumption. Although the enhanced turbulent fluctuations in the sand-laden wind also lead to a certain degree of increase in fluctuating wind pressure in these regions, the effect is less pronounced compared to the windward regions.
Figure 12 presents the sand concentration distribution results around the model for two typical wind directions, 0° (head-on wind) and 45° (oblique wind), in a sand-laden wind. Due to the blocking effect of the obstacle, the sand particle concentration in the windward region of the model is significantly higher than the background sand concentration of the flow field (≈1.0 g/m3). Figure 13 illustrates the instantaneous flow field velocity and particle motion velocity on the facade of the model at a 0° wind direction and on a horizontal plane at a height of 0.6L. In the vicinity of the model, the airflow undergoes deceleration and accelerated bypass flow due to the obstruction, resulting in a more pronounced velocity difference between particles and airflow near the model compared to regions farther away. This leads to more significant interference with the airflow. Additionally, the rebound of particles after colliding with the wall in the windward region further exacerbates airflow disturbances. These factors collectively contribute to the more substantial increase in fluctuating wind pressure in the windward region compared to other regions.
In a wind-sand environment, the high sand particle concentration within the near-surface region exacerbates the turbulence of the near-surface wind field, resulting in increased fluctuating wind pressure on the surfaces of engineering structures, particularly in the windward region. Fluctuating loads are a primary cause of structural vibration and fatigue damage in components. Consequently, for low-height structures such as buildings and solar photovoltaic panels constructed in wind-sand regions, greater attention should be paid to the issue of fatigue damage. A statistical analysis of fluctuating wind pressure in the windward region of buildings under sand-laden wind, based on a 95% confidence interval, revealed an 8.42% ± 1.23% increase. Therefore, it is recommended that the design value for fluctuating wind loads on engineering structures in such regions be increased by 10% to enhance structural resilience.

3.2.2. Skewness and Kurtosis Results of Wind Pressure

The interaction between airflow and bluff body structures leads to the generation of organized vortices exhibiting strong spatial correlations in the surrounding flow field, thereby causing the surface pressure on buildings to exhibit non-Gaussian characteristics. In practical applications of structural wind engineering, there is no widely accepted unified standard for distinguishing between Gaussian and non-Gaussian characteristics of wind pressures; different criteria are often employed for various types of structures [55,56]. This study adopts the evaluation criteria established by Kumar and Stathopoulos [57,58] for assessing Gaussian and non-Gaussian characteristics of low-rise buildings: a measuring point is classified as non-Gaussian if the |Csk| > 0.5 and Cku > 3.5; the corresponding area is then classified as a non-Gaussian feature region. Figure 14 and Figure 15 present the statistical results for the Csk and Cku under two wind direction conditions of 0° (head-on wind) and 45° (oblique wind), while the results for other wind angles are detailed in Table 1.
In Figure 14a, the Csk and Cku statistical results of the wind pressure time series at measuring points under a 0° wind direction model are presented, showing the percentage of measuring points exhibiting non-Gaussian characteristics. In a sand-laden wind, the percentage is 60.34%, while in a clean wind, it is 43.21%, marking an increase of 39.6%. The red lines in Figure 14b–e delineate the boundaries of regions where the |Csk| > 0.5 or the Cku > 3.5. The figures indicate a significant increase in the distribution area of non-Gaussian characteristics of wind pressure in the sand-laden wind; the area classified as non-Gaussian increased by 54.8% compared to the clean wind. The skewness and kurtosis contour maps in Figure 14b–e reveal that, with few exceptions, the absolute values of the skewness and kurtosis maxima and minima are significantly heightened in the sand-laden wind compared to the clean wind. In Figure 14b,c, a comparison of Csk shows that the maximum positive skewness increased from 0.60 to 0.93, while the minimum negative skewness decreased from −1.40 to −1.99. In Figure 14d,e, a comparison of Cku indicates that the maximum kurtosis increased from 9.58 to 11.39. This clearly demonstrates that the non-Gaussian characteristics of wind pressure in the sand-laden wind are more pronounced.
Figure 15a presents the Csk and Cku statistical results for the wind pressure at measuring points under a 45° wind direction model, depicting the percentage of measuring points exhibiting non-Gaussian characteristics. In the sand-laden wind, this percentage is 22.69%, while in the clean wind, it is only 12.19%, indicating an increase of 86.1%. The red lines in Figure 15b–e delineate the regions associated with non-Gaussian characteristics, and the results for the sand-laden wind show a similarly significant increase, with the coverage area of non-Gaussian regions rising by 149.4%. The contour maps of skewness and kurtosis in Figure 15b–e reveal that, with a few exceptions, the absolute values of the maxima and minima of skewness and kurtosis on various surfaces are significantly elevated in the sand-laden wind compared to the clean wind. In Figure 15b,c, a comparison of Csk shows that the maximum positive skewness increased from 0.64 to 0.65, while the minimum negative skewness decreased from −1.53 to −1.80. In Figure 15d,e, a comparison of Cku indicates that the maximum kurtosis increased from 6.66 to 9.81. This demonstrates that the non-Gaussian characteristics of wind pressure in the s sand-laden wind under the 45° wind direction model are also significantly enhanced.
Based on the distribution of the maxima and minima of Csk and Cku across the models’ surfaces in Figure 14 and Figure 15, regions with strong non-Gaussian characteristics of wind pressure are observed at the edges of the buildings, along ridges, and in areas influenced by cylindrical vortices, conical vortices, and wake effects, which also correspond to regions of high wind pressure extremes. The non-Gaussian areas on the building surfaces are significantly influenced by the wind direction. Under both wind direction conditions, the overall distribution characteristics of non-Gaussian wind pressure areas in the clean wind and the sand-laden wind do not show significant changes; however, the non-Gaussian characteristics of wind pressure in the sand-laden wind are enhanced, leading to a substantial increase in the distribution area of non-Gaussian features on the building surfaces.
Table 1 further summarizes the number of measurement points exhibiting non-Gaussian characteristics in wind pressure (NNG) on various surfaces of the building within clean and sand-laden wind under seven wind directions, as well as the proportion of the corresponding areas of these measurement points to the respective surface areas (NGA, %). As shown in Table 1, under different wind directions, the extent of Gaussian and non-Gaussian distribution regions on various building surfaces exhibits significant variability.
Figure 16 further illustrates the area proportion of non-Gaussian regions on the overall building surface, as well as the growth rate of non-Gaussian regions within the sand-laden wind. It can be observed that the distribution ranges of Gaussian and non-Gaussian regions on various building surfaces vary significantly under different wind directions. The model at a 0° wind direction shows the largest overall non-Gaussian distribution area, whereas the model at a 45° wind direction exhibits the smallest non-Gaussian distribution area. This difference is attributed to the facades that are directly influenced by the incoming flow, where the wind pressure generally follows the quasi-steady assumption: the probability density distributions of wind pressure and wind speed are similar, with relatively small absolute values of skewness and kurtosis, resulting in predominantly Gaussian characteristics. In contrast, surfaces affected by separation flows and wake flows (i.e., characteristic turbulence), such as the top, side, and leeward areas, predominantly exhibit non-Gaussian characteristics, characterized by negative skewness and larger kurtosis values. The 45° wind direction model has a larger windward area compared to other wind conditions, thus resulting in a relatively smaller overall proportion of non-Gaussian distribution. However, examining the number of non-Gaussian measurement points and the percentage of non-Gaussian area across surfaces, along with Figure 14 and Figure 15, indicates that the overall characteristics of non-Gaussian region distribution in clean and sand-laden wind do not change significantly with different wind directions. As shown in Table 1, with a few exceptions, the sand-laden wind generally exhibits more measurement points with wind pressure demonstrating non-Gaussian characteristics, with non-Gaussian areas increasing to varying extents. Notably, the increases in non-Gaussian areas under 45° and 60° wind directions are the most significant, at 149.4% and 168.0%, respectively. Comparatively, the total area proportions of non-Gaussian areas are largest for the 0° and 30° wind directions, and the increase in non-Gaussian areas in the sand-laden wind is also relatively significant, with rates of increase 54.8% and 27.1%, respectively.
In a strong wind-sand environment, the disturbance of airflow caused by sand particle motion intensifies, leading to an increase in the root-mean-square of fluctuating wind speed in the turbulent wind field (as shown in Figure 8a). Consequently, the non-Gaussian characteristics of wind pressure in the wind-sand environment are more pronounced compared to those in a clean wind. This observation aligns with the findings of Yang et al. [56], which indicate that the non-Gaussian characteristics of wind pressure in the separation zone of bluff body structures become more significant with increasing turbulence intensity. Therefore, in a wind-sand environment, the risk of localized tearing damage to building envelopes is considerably higher than in a clean wind environment.

4. Conclusions

This study conducts a numerical simulation investigation into the wind pressure characteristics on the surfaces of low-rise buildings in a sand-laden wind environment, with a focus on the influence of sand particle motion on turbulent wind fields and the wind pressure characteristics on low-rise building surfaces in sand-laden wind. The main conclusions are as follows:
(1)
The presence of sand particles results in a portion of the wind field’s energy being consumed to drive particle motion, leading to a slight reduction in the mean wind speed in the sand-laden wind compared to the clear wind. Enhanced disturbance of the flow field by particles increases the energy of small-scale vortices, resulting in an increase in the fluctuating wind speed.
(2)
In sand-laden wind, the average wind pressure coefficient on low-rise buildings remains comparable to that in clean wind, but the fluctuating wind pressure coefficient significantly increases, particularly in windward regions across various wind directions. Consequently, it is recommended that the design value for fluctuating wind loads on low-rise buildings in wind-sand regions be increased by 10% to improve the safety of envelope structures and their connections.
(3)
Under clean and sand-laden wind, the distribution of non-Gaussian regions on building surfaces remains similar. However, sand-laden wind significantly intensifies non-Gaussian wind pressure characteristics, with the area of non-Gaussian regions in the model increasing by up to 1.5 times at its maximum. This suggests a substantially higher risk of localized tearing damage to building envelopes in strong sand-laden wind environments.

5. Limitations and Outlooks

The wind field characteristics of sandstorm flow are influenced by multiple factors, including atmospheric thermal stratification stability, strong convective flows, and sand particle motion. This study solely considers the interaction between particle motion and the flow field, neglecting complex effects such as thermal dynamics and convection. Consequently, the simulated wind-sand flow field differs from the actual dust storm flow field.
Given the substantial computational cost of the E-L model, the simulations in this study are based on a small-scale model. As a result, the computed wind pressure distribution likely deviates from that observed in real sandstorm environments. Improving numerical methods for multiphase flow, reducing computational costs, and enhancing computational efficiency remain long-term objectives in the field of multiphase flow research.
In the future, we aim to further expand research efforts in this field by focusing on the following key areas: (1) Establishing on-site monitoring platforms in typical –wind-sand regions to collect data on flow field characteristics and sand particle distribution under extreme wind and wind-sand conditions, such as sandstorms. (2) Conducting long-term on-site monitoring of typical engineering structures in wind-sand regions, such as low-rise buildings and photovoltaic modules, to investigate the spatial and temporal variations in wind pressure distribution and peak pressure (suction) on structural surfaces. This will provide empirical wind load data for typical engineering structures in extreme wind-sand environments, supporting further wind tunnel experiments and numerical simulations. (3) Promoting the incorporation of wind-sand loads into engineering design standards and proposing structural design parameters for wind-sand regions.

Author Contributions

D.H.: Conceptualization, methodology, validation, writing-original draft, writing-review and editing. T.Z.: Writing-review and editing, formal analysis, investigation. Q.J.: Conceptualization, methodology, validation, supervision, software, funding acquisition. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Xinjiang Uygur Autonomous Region Natural Science Fund Project (grant no. 2024D01B43), the National Key Research and Development Program Project (grant no. 2023YFF130420304), and the National Natural Science Foundation-China-Macau Joint Research (grant no. 42381164666).

Data Availability Statement

The data used in this work are available at Figshare. https://doi.org/10.6084/m9.figshare.29434700.v1.

Acknowledgments

The authors would like to express their sincere gratitude to Ning Huang and Pei Binbin of Lanzhou University for their insightful suggestions during the development of this work.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Saracoglu, B.O.; Ohunakin, O.S.; Adelekan, D.S.; Gill, J.; Atiba, O.E.; Okokpujie, I.P.; Atayero, A.A. A framework for selecting the location of very large photovoltaic solar power plants on a global/supergrid. Energy Rep. 2018, 4, 586–602. [Google Scholar] [CrossRef]
  2. National Development and Reform Commission, National Energy Administration. Planning and Layout Scheme for Large-scale Wind and Photovoltaic Bases Focusing on Desert, Gobi, and Desert Areas [EB/OL]. Available online: http://www.spic.com.cn/spicm/xwzx/hyqy/202203/t20220307_318670.html (accessed on 7 March 2022).
  3. Cao, S.Y.; Wang, J. Statistical summary and case studies of strong wind damage in China. J. Disaster Res. 2013, 8, 1096–1102. [Google Scholar] [CrossRef]
  4. Taleb, R.; Ramanantoa, H.; Reynolds, T.; Beckett, C.T.; Huang, Y.; Rakotoarivony, M.; Gagnon, A.S.; Andriamaro, L. Fragility assessment of traditional wooden houses in Madagascar subjected to extreme wind loads. Eng. Struct. 2023, 289, 116220. [Google Scholar] [CrossRef]
  5. Levitan, M.L.; Mehta, K.C.; Vann, W.P.; Holmes, J.D. Field measurements of pressures on the Texas Tech building. J. Wind. Eng. Ind. Aerodyn. 1991, 38, 227–234. [Google Scholar] [CrossRef]
  6. Pita, G.; Pinelli, J.P.; Cocke, S.; Gurley, K.; Mitrani-Reiser, J.; Weekes, J.; Hamid, S. Assessment of hurricane-induced internal damage to low-rise buildings in the Florida Public Hurricane Loss Model. J. Wind. Eng. Ind. Aerodyn. 2012, 104, 76–87. [Google Scholar] [CrossRef]
  7. Liu, Z.; Ishihara, T. Study of the effects of translation and roughness on tornado-like vortices by large-eddy simulations. J. Wind. Eng. Ind. Aerodyn. 2016, 151, 1–24. [Google Scholar] [CrossRef]
  8. Razavi, A.; Sarkar, P.P. Tornado-induced wind loads on a low-rise building: Influence of swirl ratio, translation speed and building parameters. Eng. Struct. 2018, 167, 1–12. [Google Scholar] [CrossRef]
  9. Jesson, M.; Sterling, M.; Letchford, C.; Haines, M. Aerodynamic forces on generic buildings subject to transient, downburst-type winds. J. Wind. Eng. Ind. Aerodyn. 2015, 137, 58–68. [Google Scholar] [CrossRef]
  10. Haines, M.; Taylor, I. Numerical investigation of the flow field around low-rise buildings due to a downburst event using large eddy simulation. J. Wind. Eng. Ind. Aerodyn. 2018, 172, 12–30. [Google Scholar] [CrossRef]
  11. Zhang, M. Numerical Simulation of Wind-Blown-Sand Two Phase Flow Field Around the Building Based on Fluent. Master’s thesis, Harbin Institute of Technology, Harbin, China, 2008. [Google Scholar]
  12. Ma, Q. Numerical Simulation of Wind-Blown Sand Flow Around a Couple of Buildings. Master’s Thesis, Lanzhou University, Lanzhou, China, 2020. [Google Scholar]
  13. Bai, J. Numerical Simulation of Wind-Blown Sand Flow Around Cylindrical Engineering Structure. Master’s Thesis, Lanzhou University, Lanzhou, China, 2021. [Google Scholar]
  14. Hu, D.; Zhang, J.; Pakzad, R.; Huang, N. A numerical study on load effects of low-rise buildings in a wind-blown sand environment. J. Build. Eng. 2024, 84, 108610. [Google Scholar] [CrossRef]
  15. Huang, B.; Li, Z.; Gong, B.; Zhang, Z.; Shan, B.; Pu, O. Study on the sandstorm load of low-rise buildings via wind tunnel testing. J. Build. Eng. 2023, 65, 105821. [Google Scholar] [CrossRef]
  16. Huang, B.; Li, Z.; Zhao, Z.; Wu, H.; Zhou, H.; Cong, S. Near-ground impurity-free wind and wind-driven sand of photovoltaic power stations in a desert area. J. Wind. Eng. Ind. Aerodyn. 2018, 179, 483–502. [Google Scholar] [CrossRef]
  17. Xin, G.; Huang, N.; Zhang, J.; Dun, H. Investigations into the design of sand control fence for Gobi buildings. Aeolian Res. 2021, 49, 100662. [Google Scholar] [CrossRef]
  18. Yao, Z.; Xiao, J.; Jiang, F. Characteristics of daily extreme-wind gusts along the Lanxin Railway in Xinjiang, China. Aeolian Res. 2012, 6, 31–40. [Google Scholar] [CrossRef]
  19. Deng, E.; Yue, H.; Liu, X.–Y.; Ni, Y.–Q. Aerodynamic impact of wind-sand flow on moving trains in tunnel-embankment transition section: From field testing to CFD modeling. Eng. Appl. Comput. Fluid Mech. 2023, 17, 2279993. [Google Scholar] [CrossRef]
  20. Cheng, J.–J.; Jiang, F.-Q.; Xue, C.-X.; Xin, G.-W.; Li, K.-C.; Yang, Y.-H. Characteristics of the disastrous wind-sand environment along railways in the Gobi area of Xinjiang, China. Atmos. Environ. 2015, 102, 344–354. [Google Scholar] [CrossRef]
  21. Alghamdi, A.A.A.; Al-Kahtani, N.S. Sand control measures and sand drift fences. J. Perform. Constr. Facil. 2005, 19, 295–299. [Google Scholar] [CrossRef]
  22. Li, Z.; Pu, O.; Gong, B.; Zhao, Z.; Huang, B.; Wu, H. A new method of measuring sand impact force using piezoelectric ceramics. Measurement 2021, 179, 109390. [Google Scholar] [CrossRef]
  23. Zhang, K.; Zhang, H.; Liu, B.; Wang, T.; Wang, Z.; Tian, J. Numerical simulation study on the impact of wind-blown sand action on the loading of photovoltaic systems. Phys. Fluids 2024, 36, 075166. [Google Scholar] [CrossRef]
  24. Wiesinger, F.; Sutter, F.; Fernández-García, A.; Wette, J.; Hanrieder, N. Sandstorm erosion on solar reflectors: A field study on height and orientation dependence. Energy 2021, 217, 119351. [Google Scholar] [CrossRef]
  25. Photovoltaic Industry Network. Severe Losses! Sandstorm Hits, Solar Power Plants Fly Away–‘Photovoltaic Desertification Control’ Once Again Becomes a Hot Topic [EB/OL]. Available online: https://mp.weixin.qq.com/s/WETbWQ9R2HGGpsTPSYy41w (accessed on 20 April 2024).
  26. Qinhan Engineering. “Multiple Solar Power Stations Attacked! ‘Photovoltaic Desertification Control’ Once Again Becomes a Hot Topic” [EB/OL]. Available online: https://www.163.com/dy/article/IRIJU2AE05566LXQ.html (accessed on 22 February 2024).
  27. Hughes, G.O. Inside the head and tail of a turbulent gravity current. J. Fluid Mech. 2016, 790, 1–4. [Google Scholar] [CrossRef]
  28. Wang, G.; Gu, H.; Zheng, X. Large scale structures of turbulent flows in the atmospheric surface layer with and without sand. Phys. Fluids 2020, 32, 106604. [Google Scholar] [CrossRef]
  29. China Central Television News. Mongolia Has Issued a Severe Weather Warning for Intense Blizzards and Sandstorms. [EB/OL]. Available online: http://m.news.cctv.com/2021/05/21/ARTIZYVDvAKX9aIBSSCHPTIM210521.shtml (accessed on 20 May 2021).
  30. Ong, R.H.; Patruno, L.; Yeo, D.; He, Y.; Kwok, K.C.S. Numerical simulation of wind-induced mean and peak pressures around a low-rise structure. Eng. Struct. 2020, 214, 110583. [Google Scholar] [CrossRef]
  31. Balachandar, S.; Eaton, J.K. Turbulent dispersed multiphase flow. Annu. Rev. Fluid Mech. 2010, 42, 111–133. [Google Scholar] [CrossRef]
  32. Zhang, Z.; Chen, Q. Comparison of the Eulerian and Lagrangian methods for predicting particle transport in enclosed spaces. Atmos. Environ. 2007, 41, 5236–5248. [Google Scholar] [CrossRef]
  33. Shanmugasundaram, J.; Arunachalam, S.; Gomathinayagam, S.; Lakshmanan, N.; Harikrishna, P. Cyclone damage to buildings and structures-A case study. J. Wind. Eng. Ind. Aerodyn. 2000, 84, 369–380. [Google Scholar] [CrossRef]
  34. Xu, B.; Zhang, J.; Huang, N.; Gong, K.; Liu, Y. Characteristics of turbulent aeolian sand movement over straw checkerboard barriers and formation mechanisms of their internal erosion form. J. Geophys. Res. Atmos. 2018, 123, 6907–6919. [Google Scholar] [CrossRef]
  35. Yamamoto, Y.; Potthoff, M.; Tanaka, T.; Kajishima, T.; Tsuji, Y. Large-eddy simulation of turbulent gas-particle flow in a vertical channel: Effect of considering inter-particle collisions. J. Fluid Mech. 2001, 442, 303–334. [Google Scholar] [CrossRef]
  36. Kok, J.F.; Renno, N.O. A comprehensive numerical model of steady state saltation (COMSALT). J. Geophys. Res. Atmos. 2009, 114, 204. [Google Scholar] [CrossRef]
  37. Clift, R.; Grace, J.R.; Weber, M.E. Bubbles, Drops, and Particles; Academic Press: Cambridge, MA, USA, 1978. [Google Scholar]
  38. Zheng, X.; Feng, S.; Wang, P. Modulation of turbulence by saltating particles on erodible bed surface. J. Fluid Mech. 2021, 918, A16. [Google Scholar] [CrossRef]
  39. Tominaga, Y.; Mochida, A.; Yoshie, R.; Kataoka, H.; Nozu, T.; Yoshikawa, M.; Shirasawa, T. AIJ guidelines for practical applications of CFD to pedestrian wind environment around buildings. J. Wind. Eng. Ind. Aerodyn. 2008, 96, 1749–1761. [Google Scholar] [CrossRef]
  40. Ricci, M.; Patruno, L.; De Miranda, S. Wind loads and structural response: Benchmarking LES on a low-rise building. Eng. Struct. 2017, 144, 26–42. [Google Scholar] [CrossRef]
  41. Wang, X.; Cai, C.; Yuan, P.; Xu, G.; Sun, C. An efficient and accurate DSRFG method via nonuniform energy spectra discretization. Eng. Struct. 2024, 298, 117014. [Google Scholar] [CrossRef]
  42. Cai, Y.; Wan, J.; Kareem, A. A new divergence-free synthetic eddy method for generating homogeneous isotropic turbulence with a prescribed energy spectrum. Comput. Fluids 2023, 253, 105788. [Google Scholar] [CrossRef]
  43. Melaku, A.F.; Bitsuamlak, G.T. A divergence-free inflow turbulence generator using spectral representation method for large-eddy simulation of ABL flows. J. Wind. Eng. Ind. Aerodyn. 2021, 212, 104580. [Google Scholar] [CrossRef]
  44. Huang, X.B.; Guo, Q. OpenFOAM: From Introduction to Mastery; Water & Power Press: Beijing, China, 2021. [Google Scholar]
  45. Cabot, W.; Moin, P. Approximate wall boundary conditions in the large-eddy simulation of high Reynolds number flow. Flow Turbul. Combust. 2000, 63, 269–291. [Google Scholar] [CrossRef]
  46. Tao, W. Numerical Heat Transfer, 2nd ed.; Xi’an Jiaotong University Press: Xi’an, China, 2001. [Google Scholar]
  47. Spalart, P.R.; Deck, S.; Shur, M.L.; Squires, K.D.; Strelets, M.K.; Travin, A. A new version of detached-eddy simulation, resistant to ambiguous grid densities. Theor. Comput. Fluid Dyn. 2006, 20, 181–195. [Google Scholar] [CrossRef]
  48. He, K.; Minelli, G.; Wang, J.; Gao, G.; Krajnović, S. Assessment of LES, IDDES and RANS approaches for prediction of wakes behind notchback road vehicles. J. Wind. Eng. Ind. Aerodyn. 2021, 217, 104737. [Google Scholar] [CrossRef]
  49. Liu, J.; Niu, J. CFD simulation of the wind environment around an isolated high-rise building: An evaluation of SRANS, LES and DES models. Build. Environ. 2016, 96, 91–106. [Google Scholar] [CrossRef]
  50. Lateb, M.; Masson, C.; Stathopoulos, T.; Bédard, C. Simulation of near-field dispersion of pollutants using detached-eddy simulation. Comput. Fluids 2014, 100, 308–320. [Google Scholar] [CrossRef]
  51. Liu, H.; He, X.; Zheng, X. Amplitude modulation in particle-laden atmospheric surface layers. J. Fluid Mech. 2023, 957, A14. [Google Scholar] [CrossRef]
  52. Blott, S.J.; Pye, K. Particle size distribution analysis of sand-sized particles by laser diffraction: An experimental investigation of instrument sensitivity and the effects of particle shape. Sedimentology 2006, 53, 671–685. [Google Scholar] [CrossRef]
  53. Jeong, S.H. Simulation of large wind pressures by gusts on a bluff structure. Wind. Struct. 2004, 7, 333–344. [Google Scholar] [CrossRef]
  54. Davenport, A.G. Note on the Distribution of the Largest Value of a Random Function with Applications to Gust Loading. Proc. Inst. Civ. Eng. 1964, 28, 187–196. [Google Scholar] [CrossRef]
  55. Cheng, X.X.; Ke, S.T.; Li, P.F.; Ge, Y.J.; Zhao, L. External extreme wind pressure distribution for the structural design of cooling towers. Eng. Struct. 2019, 181, 336–353. [Google Scholar] [CrossRef]
  56. Yang, X.; Yang, Y.; Li, M.; Wang, P. Effects of free-stream turbulence on non-Gaussian characteristics of fluctuating wind pressures on a 5:1 rectangular cylinder-ScienceDirect. J. Wind. Eng. Ind. Aerodyn. 2021, 217, 104759. [Google Scholar] [CrossRef]
  57. Kumar, K.S.; Stathopoulos, T. Power spectra of wind pressures on low building roofs. J. Wind. Eng. Ind. Aerodyn. 1998, 74, 665–674. [Google Scholar] [CrossRef]
  58. Kumar, K.S.; Stathopoulos, T. Wind loads on low building roofs: A stochastic perspective. J. Struct. Eng. 2000, 126, 944–956. [Google Scholar] [CrossRef]
Figure 1. Types of wind-induced damage to building envelopes: (a) roof damage, (b) wall cladding damage, (c) wall insulation damage.
Figure 1. Types of wind-induced damage to building envelopes: (a) roof damage, (b) wall cladding damage, (c) wall insulation damage.
Buildings 15 02779 g001
Figure 2. Computational domain and boundary conditions.
Figure 2. Computational domain and boundary conditions.
Buildings 15 02779 g002
Figure 3. Mesh distribution for 0° wind direction model: (a) computational domain mesh, (b) refined near-wall mesh, (c) boundary layer mesh.
Figure 3. Mesh distribution for 0° wind direction model: (a) computational domain mesh, (b) refined near-wall mesh, (c) boundary layer mesh.
Buildings 15 02779 g003
Figure 4. Particle size distribution of sand at different heights [51].
Figure 4. Particle size distribution of sand at different heights [51].
Buildings 15 02779 g004
Figure 5. Measurement points layout on (a) roof, (b) windward (leeward) surfaces, (c) two sides, and (d) cross-sections definitions.
Figure 5. Measurement points layout on (a) roof, (b) windward (leeward) surfaces, (c) two sides, and (d) cross-sections definitions.
Buildings 15 02779 g005
Figure 6. Determination of statistical duration for wind pressure results in (a) clean wind and (b) sand-laden wind.
Figure 6. Determination of statistical duration for wind pressure results in (a) clean wind and (b) sand-laden wind.
Buildings 15 02779 g006
Figure 7. (a) Mean and (b) fluctuating wind pressure coefficients of longitudinal measurement points on symmetry axis (A–B–C–D) of TTU.
Figure 7. (a) Mean and (b) fluctuating wind pressure coefficients of longitudinal measurement points on symmetry axis (A–B–C–D) of TTU.
Buildings 15 02779 g007
Figure 8. Wind field characteristics: (a) Mean wind speed Uz/U0 and root mean square of fluctuating wind speed σ u , where U0 represents reference wind speed at a height of 2L; (b) power spectrum of fluctuating wind speed. Wind field was established based on field data from Huang et al. [51], with fitted wind profile expressed as U z / U 10 = ( z / z 10 ) 0.3 .
Figure 8. Wind field characteristics: (a) Mean wind speed Uz/U0 and root mean square of fluctuating wind speed σ u , where U0 represents reference wind speed at a height of 2L; (b) power spectrum of fluctuating wind speed. Wind field was established based on field data from Huang et al. [51], with fitted wind profile expressed as U z / U 10 = ( z / z 10 ) 0.3 .
Buildings 15 02779 g008
Figure 9. Instantaneous vortex structures of (a) clean wind field and (b) sand-laden wind field (without model).
Figure 9. Instantaneous vortex structures of (a) clean wind field and (b) sand-laden wind field (without model).
Buildings 15 02779 g009
Figure 10. Average wind pressure coefficients on sections M (1–2–3–4) (subscript 1) and N (5–6–7–8) (subscript 2) under wind direction angles of (a) 0° and 45°, (b) 15° and 60°, and (c) 30°, 75°, and 90° in clean wind and sand-laden wind. The designations (a1,a2,b1,b2,c1,c2) correspond to the respective angles and sections.
Figure 10. Average wind pressure coefficients on sections M (1–2–3–4) (subscript 1) and N (5–6–7–8) (subscript 2) under wind direction angles of (a) 0° and 45°, (b) 15° and 60°, and (c) 30°, 75°, and 90° in clean wind and sand-laden wind. The designations (a1,a2,b1,b2,c1,c2) correspond to the respective angles and sections.
Buildings 15 02779 g010
Figure 11. Fluctuating wind pressure coefficients on sections M (1–2–3–4) (subscript 1) and N (5–6–7–8) (subscript 2) under wind direction angles of (a) 0° and 45°, (b) 15° and 60°, and (c) 30°, 75°, and 90° in clean wind and sand-laden wind. The designations (a1,a2,b1,b2,c1,c2) correspond to the respective angles and sections.
Figure 11. Fluctuating wind pressure coefficients on sections M (1–2–3–4) (subscript 1) and N (5–6–7–8) (subscript 2) under wind direction angles of (a) 0° and 45°, (b) 15° and 60°, and (c) 30°, 75°, and 90° in clean wind and sand-laden wind. The designations (a1,a2,b1,b2,c1,c2) correspond to the respective angles and sections.
Buildings 15 02779 g011
Figure 12. Sand concentration contour plots in sand-laden wind for model region at wind direction of (a) 0° and (b) 45°.
Figure 12. Sand concentration contour plots in sand-laden wind for model region at wind direction of (a) 0° and (b) 45°.
Buildings 15 02779 g012
Figure 13. Instantaneous wind field velocity streamlines and particle velocity vectors for model at a 0° wind direction on (a) facade and (b) horizontal plane at a height of 0.6L.
Figure 13. Instantaneous wind field velocity streamlines and particle velocity vectors for model at a 0° wind direction on (a) facade and (b) horizontal plane at a height of 0.6L.
Buildings 15 02779 g013
Figure 14. (a) Statistics of skewness and kurtosis for 0° wind direction model; (b) 0°–I skewness; (c) 0°–S skewness; (d) 0°–I kurtosis; (e) 0°–S kurtosis.
Figure 14. (a) Statistics of skewness and kurtosis for 0° wind direction model; (b) 0°–I skewness; (c) 0°–S skewness; (d) 0°–I kurtosis; (e) 0°–S kurtosis.
Buildings 15 02779 g014
Figure 15. (a) Statistics of skewness and kurtosis for 45° wind direction model; (b) 45°–I skewness; (c) 45°–S skewness; (d) 45°–I kurtosis; (e) 45°–S kurtosis.
Figure 15. (a) Statistics of skewness and kurtosis for 45° wind direction model; (b) 45°–I skewness; (c) 45°–S skewness; (d) 45°–I kurtosis; (e) 45°–S kurtosis.
Buildings 15 02779 g015aBuildings 15 02779 g015b
Figure 16. Area and rate of increase for non-Gaussian regions on building surfaces under varying wind directions. TNG (%) is defined as ratio of sum of non-Gaussian measurement points’ areas on each surface to model’s total surface areas. Rate of increase indicates total area of non-Gaussian areas in sand-laden wind relative to that in clean wind. Orange represents the variation in TNG (bar chart), while green indicates the variation in rate of increase (dot-line graph).
Figure 16. Area and rate of increase for non-Gaussian regions on building surfaces under varying wind directions. TNG (%) is defined as ratio of sum of non-Gaussian measurement points’ areas on each surface to model’s total surface areas. Rate of increase indicates total area of non-Gaussian areas in sand-laden wind relative to that in clean wind. Orange represents the variation in TNG (bar chart), while green indicates the variation in rate of increase (dot-line graph).
Buildings 15 02779 g016
Table 1. Number of measuring points exhibiting non-Gaussian characteristics and corresponding area of non-Gaussian regions on building surfaces at various wind directions.
Table 1. Number of measuring points exhibiting non-Gaussian characteristics and corresponding area of non-Gaussian regions on building surfaces at various wind directions.
StateRoof (A)Windward (B)Right (C)Lee (D)Left (E)
NNGNGA
(%)
NNGNGA
(%)
NNGNGA
(%)
NNGNGA
(%)
NNGNGA
(%)
0°–I7851.872.77368.45229.97060.5
0°–S10869.13933.17763.69069.17767.1
15°–I2512.200.03933.610.3177.3
15°–S2715.320.63030.430.92517.5
30°–I5741.92221.45162.23316.66256.8
30°–S6741.75853.13447.83316.58683.3
45°–I3221.776.4104.5133.8175.0
45°–S5232.23130.731.7123.94934.0
60°–I3120.100.010.44123.231.0
60°–S6941.93230.81813.45028.2199.5
75°–I4228.83127.710.47956.131.8
75°–S4627.34034.43231.99261.994.0
90°–I7437.67034.562.99549.62411.7
90°–S6430.97739.32424.08141.7239.2
Note: NNG represents number of measurement points exhibiting non-Gaussian characteristics. NGA (%) refers to ratio of area of non-Gaussian measurement points on each surface to corresponding surface area.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Hu, D.; Zhang, T.; Jin, Q. Investigation of Wind Pressure Dynamics on Low-Rise Buildings in Sand-Laden Wind Environments. Buildings 2025, 15, 2779. https://doi.org/10.3390/buildings15152779

AMA Style

Hu D, Zhang T, Jin Q. Investigation of Wind Pressure Dynamics on Low-Rise Buildings in Sand-Laden Wind Environments. Buildings. 2025; 15(15):2779. https://doi.org/10.3390/buildings15152779

Chicago/Turabian Style

Hu, Di, Teng Zhang, and Qiang Jin. 2025. "Investigation of Wind Pressure Dynamics on Low-Rise Buildings in Sand-Laden Wind Environments" Buildings 15, no. 15: 2779. https://doi.org/10.3390/buildings15152779

APA Style

Hu, D., Zhang, T., & Jin, Q. (2025). Investigation of Wind Pressure Dynamics on Low-Rise Buildings in Sand-Laden Wind Environments. Buildings, 15(15), 2779. https://doi.org/10.3390/buildings15152779

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop