Bogusz Bienkiewicz

Proper orthogonal decomposition and reconstruction of multi-channel roof pressure

The proper orthogonal decomposition (POD) and reconstruction of pressure on the roof of a low-rise building is described. The point pressures acquired simultaneously at 494 taps, uniformly distributed on the building surface, were used in the analysis. The pressure space correlation was computed and employed to calculate the eigenvalues and eigenvectors (modes) for two cases: roof pressure field (RP) and roof and walls pressure (RWP). The first two of the RP modes were found to be respectively similar in shape to the pressure mean square and the mean pressure derivative computed with respect to the wind direction. Strong interaction between the wall and roof pressures was identified from the POD analysis. The POD results for the RP case were employed to reconstruct the point and area-averaged roof pressure. Approximately 30% and 2% of the computed modes were needed to represent the largest negative peak of point and area-averaged roof pressure, respectively. Only one mode was sufficient to reconstruct the mean point and area-averaged roof pressure.

Local wind loading on the roof of a low-rise building

Abstract Fundamental studies of wind loading on the roof of a 1:25 geometrical scale model of the Texas Tech University test building were performed in a boundary-layer wind tunnel. The investigation was focused on the pressure distribution in the windward corner region of the roof. A range of wind directions and roof parapet heights was considered. The surface flow patterns were visualized and compared with the roof pressure. The largest negative mean and peak pressure coefficients were found for cornering winds, wind direction of 235°C. The pressure coefficients obtained using the dynamic pressure at the roof height were, respectively, −3.5 and −11. Very low roof parapets increased the overall maximum roof suction, while high parapets decreased it. For the wind direction of 225°, the highest negative presurre occurred in the secondary vortex region and the highest pressure gradient occurred in the reattachment region. It was found that the fluctuating roof pressure was sensitive to the intensity and the integral scale of turbulence in the oncoming flow. Comparison of the wind tunnel data with field data showed that the mean and the positive peak pressures were in good agreement. The negative peak pressures measured in the wind tunnel underestimated the field data.

Proper orthogonal decomposition study of approach wind-building pressure correlation

The wind-induced pressure acting on buildings and structures is greatly influenced by the approaching wind. A detailed analysis of correlation between wind and building pressure is difficult using classical statistical methods. Proper orthogonal decomposition (POD) overcomes this difficulty by allowing us to consider separately the principal coordinates and the associated eigenvectors, characterizing the dominant spatio-temporal structures of the analyzed field. In this paper, the fluctuating wind pressure acting on the building models is analyzed using POD and the correlations between the decomposed pressures and the approaching wind are examined.

Proper orthogonal decomposition of roof pressure

Abstract This paper illustrates application of the proper orthogonal decomposition in an investigation of the aerodynamic loading on a roof of a low-rise building. Pneumatically averaged pressure in a corner region of the roof, measured for concerning wind, was used in the analysis. The pressure space covariance was employed to establish the principal modes and coordinates of the roof aerodynamic loading. These quantities were employed next to perform the modal reconstruction of the roof pressure. The convergence of the roof pressure reconstruction suggests that only a few of the principal modes and coordinates are required to represent the area-averaged roof aerodynamic loading.

Wind-tunnel study of wind loading on loose-laid roofing systems

Abstract The paper presents the results of an experimental study of the wind loading on loose-laid paver system on the flat roof of a low-rise building. Attention is focused on the underneath pressure distribution, the distribution of correlation of external and underneath pressures, and the statistical characteristics of the underneath pressure. The effects of a void space under pavers on the pressure distribution and the pressure correlation, and the effects of roof parapet height on the pressure correlation, are also discussed.

Proper orthogonal decomposition of building wind pressure specified at non-uniformly distributed pressure taps

Abstract The paper discusses the proper orthogonal decomposition of wind pressure specified at non-uniformly spaced locations. The developed methodology is illustrated by considering pressure on the roof of a low-rise building. The results are presented for both non-uniformly and uniformly distributed pressure taps. The case of the non-uniformly distributed taps was obtained by discarding some of the uniformly distributed taps. In the analysis, the pressure space covariance was computed and employed to solve an eigenvalue problem. The eigenvalues and eigenvectors for the non-uniformly distributed pressure taps were computed using a symmetric matrix resulting from transformation of the non-symmetric, integrated covariance matrix. The results were compared with those for the uniformly distributed pressure taps. The best agreement was obtained when the rectangular rule of integration was applied to the non-uniformly distributed pressure taps.

Wind loading and resistance of loose-laid roof paver systems

Abstract This paper discusses wind loading on and wind resistance of generic roofing systems consisting of loose-laid pavers on a flat roof of a low-rise building. It is found that the space between pavers improves the wind resistance of the system. It reduces the paver net wind uplift and leads to an increase in the failure wind speed. The space beneath the pavers weakens the system performance. The between paper-to-underneath paver space ratio controls pressure underneath the pavers. The higher the ratio and the larger the pavers, the higher the wind resistance of the system. Square pavers are found to be more wind resistant than rectangular pavers.

Mean and Peak Wind Loads on Heliostats

Mean and peak wind loads on flat rectangular or circular heliostats were measured on models in a boundary layer wind tunnel which included an atmospheric surface layer simulation. Horizontal and vertical forces, moments about horizontal axes at the ground level and at the centerline of the heliostat, and the moment about the vertical axis through the heliostat center were measured. Results showed that loads are higher than predicted from results obtained in a uniform, low-turbulence flow due to the presence of turbulence. Reduced wind loads were demonstrated for heliostats within a field of heliostats and upper bound curves were developed to provide preliminary design coefficients.

Application of autoregressive modeling in proper orthogonal decomposition of building wind pressure

Application of autoregressive (AR) modeling in study of the results of proper orthogonal decomposition (POD) of wind-induced pressure on a low-rise building is described in the paper. The POD principal coordinates (PC) were obtained for point pressure measured simultaneously at 495 taps on the surface of a model of a low-rise building. The AR models were next fitted for the first four PC, and they were compared with the AR models developed for point pressure at representative taps on the building windward wall and roof. It was found that the order of the AR required for the principal coordinates was relatively low, and it did not exceed the order of the AR established for the considered point pressure.

Wavelet study of approach-wind velocity and building pressure

Application of wavelet transform (WT) to study approach wind speed and building roof pressure is described in the paper. The WT is briefly introduced and employed in two illustrative examples, and in analysis of wind-tunnel data. The local wavelet spectra and other WT-based quantities are presented. Significant variation in the local energy, not accounted for in Fourier spectra analysis, is noted. Further study of these variations and their effects on wind-induced loading, as well as investigations of physical processes associated with patterns exhibited by the WT results, are desired.