Search

The failures of solar array systems occur occasionally under strong or extreme winds. The wind load is one of the main loadings on such systems. To investigate the characteristics of wind loads of solar arrays mounted on flat roofs with deflectors and those on gable roofs, pressure measurement tests on rigid models were conducted. The wind pressures on the upper and lower surfaces of solar panels and also the net pressures for these two types of roofs were examined. The variations of the largest negative peak net pressures with the tributary areas, considering all wind directions, were obtained and compared with the provisions on the wind design of flat-roof-mounted solar arrays by the Structural Engineers Association of California (SEAOC) as well as the recommended net pressure-effective area curves proposed by previous studies. The results indicate that the largest negative peak net pressures of solar arrays mounted on flat roofs are larger, and decrease more rapidly with the tributary area than those on gable roofs. The results of flat roof ones agree well with the SEAOC provisions. The design wind loads proposed by previous studies are conservative when the tributary area is relatively large. [ABSTRACT FROM AUTHOR]

The characteristics of probability distribution formulas determined by classic extreme value model and non- Gaussian peak factor method were investigated respectively based on the non-Gaussian pressure results of roofs and walls on saddle-roofed structures. Results show that for the components and cladding, the probability distribution of peak factor can be estimated by Gumbel distribution (type I) appropriately, which are significantly consistent with that obtained by the generalized extreme distribution. Moreover, especially for the common fractiles of 57% and 78% in practice, previously mentioned comparison errors are minimal. The analytical expression of non-Gaussian peak factor is formulated by coefficients of skewness, kurtosis and Gaussian peak factor from time history samples, including features of parent probability distribution. Compared with classical extreme distribution method, this non-Gaussian peak factor model is capable of dealing with any strong and weak Gaussian cases, due to utilizing more comprehensive sample information. Fractile of non-Gaussian peak factor method is moderately larger than that computed by classical extreme theory with the same occurrence probability of extreme value, which consequently can evaluate large fractile probability distribution suitably. [ABSTRACT FROM AUTHOR]

The aerodynamic effect uncertainties for two kinds of typical large span roofs, a spherical shell and a cylindrical shell were analysed systematically based on repeated wind tunnel tests and the probability plot correlation coefficient method. Probability distribution characteristics of mean wind pressure and fluctuating wind pressure were investigated. The study indicates that normal distribution is the ideal model to describe the probability distribution of mean wind pressure coefficients. The mean value of variation of mean pressure coefficients for the spherical shell and the cylindrical shell are 0. 11 and 0. 13, respectively. Lognormal distribution is the ideal model to describe the probability distribution of fluctuating wind pressure coefficients. For the spherical shell, the mean value of fluctuating pressure coefficients at the leading edge and downwind part are relatively high, and the maximum value occurs at the leeward edge. The maximum value of standard deviation occurs at a 45° angle to the incoming flow. For the cylindrical shell, both the mean value and standard deviation of fluctuating pressure coefficients reach the maximum at the joint area of the rooftop and the side walls. Monte Carlo method is carried out to investigate the influence of sample size on the sampling error of mean wind pressure coefficients. Adjusting factor related to the variation of mean pressure coefficients and sample size is proposed, which will provide references for current wind tunnel tests. [ABSTRACT FROM AUTHOR]

Synchronized pressure tests on a rigid model of long-span enclosed cylindrical shell roof was conducted in a boundary layer wind tunnel, and measurements were taken for 36 wind directions. The non-Gaussian distribution characteristics and formation mechanism of fluctuating wind load were studied based on the wind load data obtained from the wind tunnel test. Firstly, the comparative analysis between probability density distribution of fluctuating wind load and standard Gaussian probability density distribution was done. The results show that the non-Gaussian distribution characteristics are relatively outstanding at the locations which are greatly influenced by the signature turbulence. Secondly, the formation mechanism of non-Gaussian distribution was discussed by analyzing the correlation of fluctuating wind loads between different taps. At the locations which are greatly influenced by the signature turbulence, the vortexes do not satisfy with the independent identically distribution, and are greatly correlated ； as a result, the characteristics of non-Gaussian distribution are relatively strong. Finally, making use of the value of skewness and kurtosis which are corresponding to the 80% cumulative probability, the Gaussian and non-Gaussian zones of long- span enclosed cylindrical shell roof were divided. The results show that the wake a re a, comer area and the areas which are impacted by inflow belong to the non-Gaussian zone, and airflow-steady areas belong to the Gaussian zone. [ABSTRACT FROM AUTHOR]

Based on wind tunnel data with long sampling time, a simplified formula for the calculation of peak wind pressures was proposed on the basis of Hermite polynomial. After obtaining the related parameters of Hermite polynomial based on approximate analytical formulae proposed by Winterstein, the unknown coefficients in the simplified formula were obtained by fitting the wind tunnel test data. By regarding maximum wind pressures and minimum counterparts of the data as reference values, the accuracy of the fitted formula was assessed. During the evaluation of the maximum wind pressures, the accuracy of the simplified formula and existing formulae by applying Hermite polynomial was compared. In the current formula, the parameters were obtained by the approximate analytical formulae proposed by Yang et al. The results show that under different sampling frequencies, durations and wind tunnel test data, the simplified formula gives an accurate estimation of the maximum values, whereas the prediction for the minimum values is unsatisfactory. Nevertheless, the discrepancy for the latter case is insignificant as the absolute values of the minimum wind pressure are small. [ABSTRACT FROM AUTHOR]

A detailed comparison on external pressure coefficient of the enclosure structure on flat roof has been made among codes and standards of various countries and regions including China (GB 50009--2012), Germany (DIN 1055-4), USA (ASCE/SEI 7-10), Japan (AIJ-2004), Europe (BS EN 1991-1-4:2005), Canada (NBC-2005), Australia/New Zealand (AS/NZS 1170.2:2011) and Britain (BS 6399-2:1997). In order to make a good comparison, wind load values of different codes and standards have been converted into the product form of the external pressure coefficient and the basic wind pressure. At the same time, external pressure coefficient has been converted into a value which is corresponding to a basic wind speed of 10 min duration and 50 years return period. The comparison and summary of factors which are related to external pressure coefficient also have been made. These factors are wind direction, building height, building size ratio, area reduction and external pressure coefficient partition. In the end, the applications of different standards have been discussed by flat roof wind tunnel tests, and some suggestions for the value selection of the enclosure structure external pressure coefficient of flat roof have been put forward. The most unfavorable wind pressure should be determined considering all 360° for the design of the enclosure structure on flat roof. The edge and corner areas should be partitioned for the wind pressure calculation. Various area reduction formulas should be used for various locations of a roof. Various external wind pressures should be used to buildings with various height or width-to-height ratios. [ABSTRACT FROM AUTHOR]

The shape factors of wind loads for the complex-shaped and important buildings are not specified in GB 50009-2012 4Load code for the design of building structures * . The wind loads of these buildings are frequently determined by wind tunnel tests. A method of determining the design wind loads for cladding and components based on the wind tunnel tests was presented in this paper. In this method, the peak pressure coefficient of wind load is expressed as the summation of shape factor and the peak fluctuating pressure coefficient, instead of the product of the gust factor and the shape factor in GB 50009-2012. The method can not only be applied to the windward cladding and components, but also to the leeward cladding and components which are located in the flow separation zones. After the non-Gaussian characteristics of wind pressure histories are considered, the peak fluctuating pressure coefficient can be simply calculated using the reduced formulae of non-Gaussian peak factor. As an example, the design wind loads on cladding and components of a large-span flat roof were investigated using both the presented method and the method in GB 50009-2012. The recommended peak pressure coefficients on cladding and components of flat roof were presented for the wind-resistant design. It is clear that the problem of design wind loads on cladding and components is solved effectively by the presented method and that the load panels specified in AIJ-RLB-2004 code are reasonable. It is illustrated that the design wind loads on cladding and components located in the flow separation zones may be underestimated by the national code if the shape factors are obtained by wind tunnel test. [ABSTRACT FROM AUTHOR]

With wind tunnel tests, the effects of terrain and building length-width ratio on wind loads of buildings with a flat roof were investigated，in terms of the mean，the RMS and the peak pressure coefficients，and the mean lift coefficients on the roof. The results demonstrate that the terrain has little effects on the mean coefficients on the roof, but has significant effects on the RMS and peak pressure coefficients. The RMS pressure coefficients on the windward separation area of the typical roof section in terrain C is about 1. 5 times those in terrain A, and the distance from the windward edge to the flow reattachment is 0. 7 H, 0. 3H and 0.3H(H is height of the model)，respectively, in terrain A, B and C when the wind direction is orthogonal to the short edge of the building. Increasing the building length along the wind direction has little effects on the pressure coefficients along the typical roof sections, makes more area of the roof be covered by the flow reattachment, and reduces the mean lift force coefficient on the roof. Both the terrain and the building length-width ratio have significant effects on the most unfavorable peak pressure coefficients for all wind directions at the roof corner area, but have little effects on those at the middle and center roof areas. For the roof corner area, the most unfavorable peak pressure coefficients in terrain C is larger than those in terrain A and B, and those for the roof with a length-width ratio of 2. 5 is 2 0% larger than those with a length-width ratio of 1. 5 and 2. 0. [ABSTRACT FROM AUTHOR]

Fluctuating wind pressures on a group of low-rise buildings with flat roofs were measured by wind tunnel test, and the influences of building row, column and building area density on the interference effects of peak wind pressure on the roof zone of the buildings located at the comer, edge and center of the building group were investigated by a single factor rotation method. The results demonstrate that shielding effects on peak wind loads of most roof zones are noticeable. The shielding effects are particularly significant for the entire roof zones of the center buildings and for the corner and central roof zones of the corner and edge buildings, in addition, amplification effects occur at the edge middle roof zones, where the largest increase reaches 14% . Increasing the building row or column has more significant effects on corner and edge building, where the shielding effects of the corner and center roof zones are further amplified. For the building group with three rows and three columns, the interference factors of all the roof zones of the center building, and those of the comer roof zones (near the building group center) of the comer and edge buildings, are more sensitive to the building area density, and the values decrease linearly (proportional coefficient is close to - 1 . 0) with an increase in the building area density. In contrast, the interference factors of the corner roof zones away from the building group center for the corner and edge buildings are less sensitive to the building area density. [ABSTRACT FROM AUTHOR]

High-frequency-pressure-integration tests were carried out in a wind tunnel to investigate wind load characteristics on smooth cantilevered roof and wavy-shaped cantilevered roof, both of which have the same size. The effects of wavy-shaped roof surface have been evaluated in wind tunnel test. Wind pressure distribution shows that, wavy shape does not change the main mechanism of wind load, but intensifies the separation on the front edge of the roof. Bending moment coefficient is a controlling factor for the wind-resistance design of the main roof structure, and can be derived from integration. Relative to the smooth surface cantilevered roof, the mean bending moment coefficient of the wavy-shaped cantilevered roof is increased by 23% and the RMS (root mean square) of bending moment coefficient of the wavy-shaped cantilevered roof is increased by 20%, under the most unfavorable wind direction angle. But wavy shape decreases the energy concentration of the fluctuating wind load when reduced frequency is between 0.1 and 0.2. Wavy shape has strong influence on the extreme wind pressure, since wind uplift on wavy-shaped cantilevered roof is 13% larger than that on smooth surface cantilevered roof, which is the controlling factor of the wind-resistant design of cladding structures. The influence of wavy shapes should be considered for the wind-resistant design of the wavy-shaped cantilevered roof to ensure safety. [ABSTRACT FROM AUTHOR]

Wind tunnel tests were conducted on three arched roofs with different rise-to-span ratios (1/4, 1/6 and 1/8). The effects of the parameters of the wind loading and structures on the wind-induced response and equivalent static wind loads (ESWL) were investigated for arched trusses. The experimental results show that when the wind direction is 0°, the maximum suction occurs on the roof apex under high rise-to-span ratios, and at the leading roof edge under low rise-to-span ratios. For the engineering convenience, one simplified mode was chosen to express the ESWL for multiple nodal displacements and member stresses. This method was employed to investigate the effects of the rise-to-span ratio, span length, roof mass, wind velocity and wind direction on ESWLs of arched trusses in the middle and edge areas. Suggested values of ESWLs are proposed for engineering structural design. Structural responses under these ESWLs agree well with actual peak responses. [ABSTRACT FROM AUTHOR]