Dissertation
Dissertation > Industrial Technology > Building Science > Building structure > Structure theory to calculate > Structural loads and structural capacity > Structural load analysis

Study of Wind Effects and Equivalent Static Wind Loads on Closing-ground Building with Roof-opening

Author LiShouKe
Tutor ChenZhengQing
School Hunan University
Course Civil Engineering
Keywords Wind Tunnel Test Roof Opening Wind-induced Internal Pressure Extreme Wind Pressure Wind-induced Response Equivalent Static Wind Loads
CLC TU312.1
Type PhD thesis
Year 2013
Downloads 0
Quotes 0
Download Dissertation

Closing-ground buildings usually present some openings on the center or cornerof roof because of the need of function and wind-induced damage, such as theretractable roof of stadium, skylights of houses or industrial buildings, and damage onthe roof of buildings. If a closing-ground building has a dominant opening on the roof,the wind loads would be significantly different with that of closed building. Theprevious studies about wind loads of dominant opening building mainly focused onwall-opening buildings, such as windward or leeward opening buildings. This thesismainly focuses on envelope wind loads, internal pressures, extreme wind loads,wind-induced responses, and equivalent static wind loads on the roof-openingclosing-ground buildings by means of wind tunnel tests and theoretical analysis. Themajor contents and results are listed as follows:1.7TTU rigid models are manufactured with a scale of1:50, which are3centerroof-opening models,3corner roof-opening models as well as one no roof-openingmodel. Wind pressures of these models are tested in HD-2wind tunnel. First, thereliabilities of HD-2wind tunnel and pressure testing system are confirmed bycomparing the results of no-opening cases among the HD-2wind tunnel tests, fullscale tests, and UWO wind tunnel tests. Second, on the basis of three kinds of centerroof-opening cases and three kinds of corner roof-openings cases, the characteristicsof wind pressure coefficients on roof-opening envelopes are studied in detail. The testresults are compared with several provisions, it seems that the suction forces on theroof obtained from provisions are overestimated and the press forces areunderestimated.2. The wind-induced internal pressures in roof-opening closing-ground buildingsare specially studied. First, the coherences of internal pressure coefficients are studied,which ensures the accuracy of describing the internal pressures by using onetime-history. The effects of wind directions, opening ratios, leakage ratios, turbulencetendencies and volumes of structures on internal pressure coefficients are studied indetail, and it is shown that the internal pressure coefficients of roof-opening buildingsare negative, which can not be well evaluated by the provisions, as well as the resultsobtained from wall-opening building tests. So an adaptive internal pressurecoefficient is proposed for the designing of roof opening closing-ground building. Second, the fluctuating mechanisms of internal pressures are specially studied. It isshown that the single dominant opening Helmholtz equation can be used to simulateinternal pressure. Based on the wind tunnel tests and simulating results, the empiricalformulae about the ratio between fluctuating internal pressure and external pressure,the ratio between extreme internal pressure and external pressure are fitted with thenon-dimension opening parameters.3. The characteristics of extreme pressure coefficients for cladding componentsare specially studied. First, the distribution of skewnesses and kurtosises as well asthe characteristics of probability density function about upside, downside and netpressure coefficients are studied. It is shown that the probability density function issignificantly different from Gaussian distribution, and the modified Hermitepolynomial model is the best probability density function which is used to fit theprobability density function for roof-opening closing-ground buildings’ wind pressurecoefficients at most cases. Based on the best fitted probability density function, thispaper proposes a modified Hermite peak factor procedure, which can take accountinto the band width effects and non-Gaussian characterstic of wind pressure. Second,this paper proposes a non-Gaussian simulation peak factor procedure for overconingthe exhausting defaults of classic peak factor procedure. By means of comparing theresults obtained from these two new procedure and several traditional peak factormethods, it is shown that the peak factors of the opening roof’s wind pressure areunderestimated by Davenport peak factor method and Sadek-Simiu peak factormethod in some cases, but Modified Hermite peak factor procedure can be used to getthe peak factors safely, and Non-Gaussian simulation peak factor procedure is the bestway to get the peak factors. The results of Non-Gaussian simulation peak factorprocedure are closest to the observed peak factors.4. The wind-induced responses and equivalent static wind loads (ESWL) arespecially studied. For taking account into the relevance of background ESWL andresonant ESWL, this paper proposes a full three-component procedure to getwind-induced responses and equivalent static wind loads by stochastic dynamicmechanism. The overall fluctaing wind-induced responses and equivalent static windloads are divided into three parts which are background component and resonantcomponent as well as background-resonant coupling component by the new procedure.By testing examples of opening roofs, it is shown that the full three-componentprocedure is an accurate method same as CQC method in theory. In normal wind anglecases the wind-induced responses and equivalent static wind loads are underestimated by the traditional three-component procedure which neglect the background-resonantcoupling component, and in oblique wind angle the results are overestimated. Lastly,for considering the equivalent of multi-object, this paper proposes a multi-objectequivalent static wind loads procedure which makes the full three-component singleESWL procedure as the basic vector and use constrained least square method to getthe results. It is shown that the new multi-object ESWL procedure can overcome thefault which is usually presenting exceptional large local ESWL by the traditionalmulti-object ESWL method, and the results of the most important object can beaccurately equivalent.

Related Dissertations
More Dissertations