Study Of Wind Tunnel Experimentation And Wind Effects On Roof Structures With Wall Openings

Long-span roof buildings often present some openings on the curtain walls because of functional requirements and wind-induced damage, such as the doors, windows of stadium, airport terminal, houses or industrial buildings. If a long-span roof building has a dominant opening on the curtain walls, the wind loads would be significantly different from that of an enclosed building. Existing studies on the wind loads of buildings with dominant openings mainly focused on cases that only the curtain walls has a dominant opening. However, it is necessary to consider the circulation of air in-and-out of the openings existing on different sides of curtain walls, which leads to more complicated measurements and investigations. In order to resolve this problem, this thesis presents detailed research on the external wind loads, internal pressure, extreme wind loads and wind-induced responses of buildings with wall-openings by combining wind tunnel tests and theoretical analysis. The major contents and results are listed as follows.(1) The wind-induced internal pressures of multiple wall-opening buildings are specially studied. Firstly, the correlation of internal pressure coefficients is discussed, which ensures the accuracy of using a uniform value to describe all the internal pressures. The effects of wind directions, internal volume, opening area of wall, opening locations and the ratio of rise to span on the internal pressure coefficients are also studied in detail. Secondly, by analyzing the connection between the spectra of internal pressure and fluctuating pressures inside buildings, it is concluded that an internal pressure spectrum includes not only the energy component of atmospheric turbulence, but also the turbulence caused by the opening features. The rms fluctuating internal pressure coefficients tend to increase with the enhancement of the resonance and energy at low frequencies and especially, for structures with multiple openings, wind pressure correlations at orifices also affect fluctuating internal pressure considerably. Then, comparing the local shape coefficient inside buildings with that stipulated in the Chinese code, and it seems that the values on the underside of a roof obtained from the code are largely underestimated for most wind directions for single wall opening and the values for multiple wall openings are well evaluated by the present code in spite of a small number of exceptions. Lastly, the flow-exponent is studied. No matter whether there is an opening in the side wall, the flow exponent is always consistent with the previous study results for the cases where the ratio of total areas of windward openings to that of leeward openings is equal to or more than3and an adaptive value ranges of flow exponent is proposed for the case where the total area of leeward openings is greater than that of windward openings.(2) The characteristics of net wind pressure coefficient and extreme pressure coefficient for cladding components are specially discussed. Firstly, the distributions of net mean and fluctuating pressure coefficient are studied. It is shown that the distribution trends of mean and fluctuating pressure coefficient are similar and for the roof with bigger rise-span ratio(S/B≥0.10), the distribution of pressure coefficient does not present conical vortex characteristics when the wind is normal to the corner of the roof. Secondly, the net lift coefficient and net shape coefficient of the roof with multiple facade openings are clarified. It is shown that the net lift coefficient of the roof is negative and changes with the rise-span ratio and openings combination. The test results of net shape coefficient are compared with the Chinese load code, which shows that the values obtained from the code are underestimated for most cases. Lastly, the theory of GPD (General Pareto Distribution) is used to estimate the net extreme wind pressure on the claddings. The distribution of extreme wind pressure changes significantly with the difference of rise-span ratio and openings combination. Then, based on the estimated extreme wind pressure, peak factors can be evaluated. The results show that the peak factors vary in different regions of the roof and the peak factors of many regions are beyond usual empirical value ranges. So, adopting uniform peak factor for the roof with multiple wall-openings is not recommended.(3) The wind-induced responses, wind-induced dynamic factors and wind load coefficients are particularly studied. To ensure the accuracy of calculation for the wind-induced responses, it is necessary to take account of real components and imaginary components of the wind load spectrum simultaneously. The effects of vibration mode numbers, mode coupling and background-resonant coupling on the overall wind-induced responses are discussed. The mode contribution factor and the cumulative mode contribution factor are proposed to identify the modes which contribute greatly to the wind-induced vibration response and judge the reasonability of computing modes. The results of computing and analyzing the wind-induced dynamic factors and wind load coefficients indicate that the adoption of a uniform value for the roof structure is unwise, and we should consider both of them at the same time to identify the worst wind direction.