Study On Wind Effects And Wind-tunnel Test Method For Super Large Cooling Towers Group

With the development of Chinese electrical utility, the height of the coolingtower is going to surpass the level of200m, with breaking through the world record,which is far exceed the limit height of165m in our current cooling tower designspecifications. In order to ensure the security and economy of such large coolingtower, as well as to provide reference for our specification revision, the wind load,response characteristics and wind-induced interference for200m high super largecooling tower are studied by wind tunnel tests, CFD simulations and finite elementanalysis of response in the present work. The main contents and conclusions aresummarized as follows:(1) The shortcomings of the rigidity model manometric experiment for innersurface pressure were analyzed by CFD simulations. The results show that the windtunnel test is still reliable. The wind tunnel test results indicate that the wind pressureof inner surface is less affected by Reynolds number and wind direction, while theventilation ratios of stuffing layers and cross baffle significant ly impact the internalwind pressure. In general, the internal pressure along hoop and meridional arebasically the same.(2) The influence of various parameters for rough strips on pressure distributionwere analyzed. It was found that roughness coefficient k/s is an important parameterto describe the roughness size, which closely affect the pressure distribution.Meanwhile, the estimation method of pressure according to k/s and the relationshipbetween roughness Reynolds numberRe Rand minimum pressure coefficient Cpminare obtained, according to which the test wind speed and roughness could beaccurately predicted for the simulation of target curve.(3) Wind loads characteristic and their design value of both outer and innersurface were investigated. It is shown that the wind loads on outer surface of thecooling tower behaved the remarkable three-dimensional effect, and the internalpressure is also not strictly uniformly distributed along the hoop and meridional. It isproved that local deformation of cooling tower play a major role in the total response,and the wind-induced response is closely associated with the distributioncharacteristics of wind pressure, while there is no clear description on relationshipbetween wind-induced response and the value of drag coefficient. It was also found that the design value of outer and inner surface could be simplified by representativecurve and constant-0.50respectively, which are proved to be sufficient safe for thestructure.(4) Similarity relations for full aero-elastic model of cooling tower were derivedand verified by finite element method, which show that full aeroelastic model cansatisfy the similar relationship of both cooling tower shell and column, and canachieve a successive distribution for mass and stiffness. A full aero-elastic model wasdesigned and processed with sufficient accuracy, and it’s wind-induced response wasmeasured by high-precision laser displacement meter and the data was carefullyanalyzed. The results show that the wind-induced response and dynamic amplificationof cooling tower is not obvious in normal climate.(5) Wind tunnel tests were carried out to study wind-induced interference effecton two200m high adjacent super large cooling towers based on rigidity modelsynchronization manometry wind tunnel tests. The results show that wind-inducedinterference affected the distribution of wind pressure significantly, particularly forthe downstream tower in the tandem configuration. Meanwhile, it is found thatinterference factors calculated by different load parameters are different, even for thesame load parameter, different eigenvalues could result in different values. Inaddition, test result show that surface roughness has less effects on the wind loadinterference factor.(6) Wind tunnel tests were carried out to study wind-induced interference effecton two200m high adjacent super large cooling towers based on full aero-elasticmodel. The test results indicate that interference effect has less effect on dynamicresponse characteristics, while gust loading factor increases slightly. Also,interference factor result from wind loads does not represent the effect of interferenceon the response, wind-induced interference effects of cooling tower should beexamined by wind-induced response. In addition, interference factor calculated bywind-induced response are less affected by surface roughness.