NUMERICAL SIMULATION OF WIND INTERFERENCE EFFECT FOR A STADIUM AND A GYMNASIUM Gang Xu 1, Xing-qian Peng 2, Li Wu 1, Hai Zhu 1 1 Graduate student, College of Civil Engineering, Huaqiao University, Quanzhou, Fujian, China, 36221, xugang2839@126.com 2 Professor, College of Civil Engineering, Huaqiao University, Quanzhou, Fujian, China, ABSTRACT 36221, pxq@hqu.edu.cn The wind load of the long-span roof of a stadium and a gymnasium in Quanzhou was calculated by CFX1. software and the SST k ω turbulence model. The numerical simulation of wind pressure on the surface of the roof of the gymnasium was carried on. A comparison between the wind pressure from the test and the simulation was made and the result showed that the two match quite well. The results in this paper verify the feasibility of numerical simulation of large roofs in wind engineering. Then, the wind interference analysis between the stadium and gymnasium was carried on. The effect to wind load in different relative positions of the long-span roof complex and their interference mechanisms were explored. An overall evaluation of the construction planning for the stadium and the gymnasium has been delivered. The difference of the interference effect on closer and farther covering is distinguished. KEYWORDS: LARGE-SPAN STRUCTURE, NUMERICAL SIMULATION, LIFT COEFFICIENT, INTERFERENCE EFFECT 1 Introduction The wind load characteristics of large-span roof structure lies on not only its structural shape, but also its position on the built environment. Distribution of modern architecture is very intricate. Generally, there are other buildings or structures around. The aerodynamic interference of such buildings or structures can not be ignored, particularly when the two buildings are very close to each other. The interference will be quite significant [Shen (24) and Chen (25)]. Of course, such interference are closely related with the position, shape, and many other factors of the interference buildings, so giving a universal significance of the conclusions is very difficult, but some meaningful conclusions are drawn through an analysis of specific examples [Katarzyna et al. (2)]. In this paper, according to long-span projects of a stadium and a gymnasium in Quanzhou, the numerical simulation of wind loads and wind interference effects of the large sports complex were calculated by CFX1. - a computational fluid dynamics software and the SST k ω turbulence model. Firstly, the numerical simulation of wind pressure on the surface of the roof of the gymnasium was carried on; its aim is to verify the accuracy of the numerical simulation methods by the comparisons of the simulation results with the wind tunnel test results [wind-tunnel test research report (25)]. Finally, the wind interference
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan analysis between the stadium and gymnasium was carried on. The effect to wind load in different relative positions of the long-span roof complex and their interference mechanisms were explored. An overall evaluation of the construction planning for the stadium and the gymnasium has been delivered. 2 The CFD numerical simulation of the average wind pressure and data processing 2.1 Numerical simulation and geometric trellis The large cross-building complex in Quanzhou is composed by the stadium (open cantilevered structure) and the gymnasium (closed structure). The distance between the stadium and the gymnasium is about 3m. The overall layout of stadium is made of an oval shell and a semi-cylindrical shell structure. The plane is uni-axial symmetry oval, the long axis is 218m and the short axis is 19.8m. The height of the highest roof is 32.5m. The structure of the gymnasium is novel and unique. The overall layout of the gymnasium is composed by four independent cantilevered roofs. The style of the gymnasium is curved reticulated shell structure with the distance between east and west is about 92m, north and south is about 22m. The height of the roof is 45.5m with a largest span of 48m. The model position, wind direction and the layout of measuring point of numerical simulation model are shown in Figure 1. In this paper, the world's leading computational fluid dynamics software CFX1. is selected to do the calculation. The full-scale model of the stadium is established according to the actual construction plans, neglecting the torch column ancillary components, the final domain for calculating is 6m 4m 25m by calculated many times, basin set to meet the blocking probability is less than 3% of the requirements under the valley stadium model front 1/3, the entire computational domain is divided into a domain I and domain II, the region is sufficient to simulate the whole atmospheric environment of the building. The non-structural hexahedral grid is used in domain I with 48,744 unstructured hexahedral elements divided. The 1,18,692 tetrahedral units and hexahedral elements of non-slip near the wall are divided in domain II, the total of elements are 1,517,436 units. The model grid is shown in Figure2. Fig1. Model position, wind direction and the layout of measuring point Fig2. The trellis of a stadium and a gymnasium 2.2 The physical model of turbulence The Reynolds stress model is used to calculate and the near wall function is used to simulate the complex flow phenomenon near the wall. The Reynolds stress model is a closed equation which includes 11 partial differential equations that consist of the average movement of three momentum equations and a continuity equation, six Reynolds stress equations and an ω equation [B. P. Hoxey et al. (1997)]. 2.3 Boundary Conditions
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan Inlet: the wind speed [Huang (22)]: Z ( α U Z = U, which Z ) Z is the reference height andu is wind speed of reference height, Z =2m and U =14m/s in accordance with the wind tunnel tests method B-type surface roughness coefficient of.16 [Richartson et al. (1997)], the turbulence intensity in Inlet, when Z 5m, I z =.23; when 5m <z 35m, Z.21 I z =.94 ( ) [Suggestion of Japan (1996)]. 35 Outlet: the fully developed turbulence model is used; the free-slip wall is used on the side and sune surface; the non-slip rough wall is used on ground; the non-slip smooth wall is used on the wall of building [Lin (25)]. 2.4 The data processing of numerical simulation In aerodynamics, the surface pressure usually is used non-dimensional pressure coefficient C which is expressed as: p p Cp = (1) p p C p is the wind pressure coefficient of the i test hole location, is the measured pressure value on the surface, p is the average measured total pressure and p is the average static pressure at the reference point. On both sides by the wind and the cantilevered parts, internal and external (from top to bottom) simultaneous measurement of the surface ΔC pressure coefficient is expressed as: Δ C = u d p p (2) u Δ C is the wind pressure coefficient of the i test hole, p is the pressure value of the d sune surface. p is the pressure value of lower surface. t The role of the wind to covering is lift force, the lift coefficient is introduced to determine the lift force of every covering under different direction which is expressed as: r r C = C A A (3) F P i / A is overall size of covering, α is the normal direction of the surface of the measuring point. α is 15 for this project, CP is the wind pressure coefficient of the i measuring point which can be the wind pressure coefficient of the sune surface, lower surface or composition force. 3 Analysis of the numerical simulation results 3.1 The comparison of roof block wind pressure coefficient between numerical simulation and wind-tunnel test results In order to validate the reliability of the numerical simulation, at first, the numerical simulation results and the wind-tunnel test results of the stadium and the gymnasium are t
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan compared, considering the length, the paper presents the roof block wind pressure coefficient deduced from numerical simulation under the wind direction of and 9 degrees. The roof block wind pressure coefficient deduced from the test and the numerical simulation respectively under the wind direction of and 9 degree are in fig 3. Fig3.a tested at Fig 3.b simulated at Fig3.c tested at 9 Fig 3.d simulated at 9 Figure3. The wind pressure coefficient of composition force of the coving roof block tested & simulated at some wind direction As can be seen from the fig, as regards the more complex geometry and large-span cantilevered roof structure of the stadium, the numerical simulation with the experimental results is still relatively close, with an average error of about 33 percent in the overall accuracy which could satisfy the project accuracy requirement basically. The average error in and 9 degree are 33.28% and 33.65% respectively. Since the complex geometry of the stadium and gymnasium, numerical modeling of the wind load value is difficult, and the computer's computing power is limited, so the number of the grid could not be demarcated fine enough, what s more, the roof block demarcating error, boundary conditions the entrance of the wind profile, roughness, etc. can not be completely coincide with the wind tunnel test environment, so there are some differences compared to the wind tunnel test, such as the vortex and separation around the construction objects generated by fluid, leading more difficult in numerical simulation. In general, the trend between the result of the numerical simulation and the result of the wind tunnel test is generally the same. The numerical
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan simulation value is greater than the wind tunnel test on the windward edge, trail edge and the corner. The two results basically inosculate and the simulation result is feasible. 3.2 The gymnasium to the stadium aerodynamic disturbance It can be seen from the position of the stadium and the gymnasium that the gymnasium will have a greater aerodynamic disturbance to the east and west cantilevered roof. The paper mainly selects the east cantilevered roof interference effects to carry on the analysis. The results from the numerical simulation of the stadium itself and both are analyzed, the lift coefficients of the east cantilevered roof under different wind directions are worked out. The lift coefficients of wind pressure of the upper surface and the under surface are in Fig 4, the total lift coefficients of wind pressure are in Fig 5. Line-a shows no disturbed case and line-b shows disturbed case. (1) The lift coefficient is smaller than that when the gymnasium did not exist. When the gymnasium is in the lower reaches, for that the lower reaches of gymnasium slowed the flow of wind to some extent, resulting in the sub-pressure becoming smaller, but the range of variation is not wide, because the stadium do not have much effect on preventing the flow of wind owing to the higher beam with overhanging. The lift coefficient is larger when the gymnasium is at the parallel position than that when the gymnasium did not exist. It mainly due to the funneling when the wind through them, the wind speed increased, resulting in the sub-pressure of the surface becoming larger. When the gymnasium is in the upper reaches, because the airflow is blocked by the gymnasium, the speed of the airflow will be reduced to a certain extent, while the airflow in the building is separated and reattached, the flow turbulence of the roof the lower reaches become higher, the separation of the flow in the front roof reduced, resulting in negative pressure becoming smaller, so the lift coefficient is smaller than that when the gymnasium did not exist. (2) To the lift coefficient of the wind pressure under the surface, when the gymnasium is in the lower reaches, the lift coefficient become larger, That explains that the lower reaches of gymnasium slowed the flow of wind to some extent,resulting in the drive effect obvious, but the range of variation is not wide, because the east cantilever roof is located in the wake zone of the west cantilever roof, the upper and lower surface pressure are smaller; when the gymnasium is in the parallel position, as the Gap tube effect, the wind speed increased, the rad outflow of the wind reduced the negative pressure under the surface, resulting in the lift coefficient becoming small; when the gymnasium is the upper reaches, because the east coving is located in the wake zone of the stadium, the flow separated in the front of the gymnasium and attached to the east cantilever roof, resulting in negative pressure under the surface becoming smaller, so the lift coefficient becomes smaller. (3) To the wind force, when the gymnasium is in the lower reaches, since the airflow is blocked by the interference of the gymnasium, the total lift force coefficient of gymnasium is little smaller than that when it does not have gymnasium. But cantilever roof is higher, the flow blocked that is not obvious, so the lift coefficient on the surface and under the surface decreased slightly; when the gymnasium is in the parallel position, the total lift force coefficient of gymnasium is larger than that when it does not have gymnasium. The total lift force of the east cantilever roof increased because the gap tube effect when the gymnasium is in the upper reaches, the total lift coefficient is smaller than that when the gymnasium did not exist. Due to the existence of the upper reaches of the gymnasium and the downstream flow, the lift coefficient of the east become smaller, in the 9 wind direction, the total lift force of the east cantilever roof is the largest.
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan.2.1 CF 3 6 9 12 15 18 21 24 27 3 33 36 -.1 -.2 -.3 -.4 Wind direction -.5 a up b up a under b under CF -.5 3 6 9 12 15 18 21 24 27 3 33 36 -.1 -.15 -.2 -.25 -.3 -.35 -.4 -.45 Wind direction a total force b total force Fig 4.The lift coefficients of wind pressure Fig 5.The total lift coefficients of wind pressure of the upper surface and the under surface of the upper surface and the under surface 3.2 The stadium to the gymnasium aerodynamic disturbance In order to reflect the quantitative interference effect of the stadium to the gymnasium, interference factor IF is employed: C Pi IF = (4) CPi is the average wind pressure interfered and C PA is the average wind pressure non-interfered. The gymnasium is composed by the Training Hall and the Competition Hall, the middle set up a 9mm wide joints, the roof block wind pressure coefficient of Training Hall is 1-15 and the roof block wind pressure coefficient of Competition Hall is 16-62, the numerical simulation of the gymnasium and the stadium is analyzed respectively,because the gymnasium and the stadium are central axis symmetrical structures, here only to analyze the ~ 9 and 27 ~ 36, the roof block interference factor of gymnasium is calculated under different wind direction, the roof block interference factor distribution maps under different wind direction is shown in Figure 6. (1) At the wind direction, the interference factor of roof block 1-15of the training hall is larger, the maximum is 2, the biggest negative is -4, the geopolitical interference factor is smaller, most interference factors of the competition hall Roof is about 1.2. Because the air flow accelerated in a passage between the stadium and the gymnasium, resulting in Gap tube effect, the roof block 1-15 interference factor of the training hall is larger. The interference effect is obvious near the stadium, the wind pressure of block 1 and 2 turned into negative and interference factor is negative too. As the roof of Competition Hall roof is far from the stadium, the interference factors of roof block 16-62 are smaller. (2) At the 45 wind direction, the interference factors of roof block 1-15 of the training hall is smaller, the minimum is.48, the geopolitical interference factor is greater, the roof block interference factor of the competition hall is smaller, most of the roof block16-62 interference factors of the competition hall s Roof is about 1. because of the existence of the stadium and the gymnasium is higher than the stadium, the flow blocked by the training hall and wind speed decreased, the wind pressure of the training hall s roof reduces, so the interference factor is smaller. (3) At the 9 wind direction, the wind pressure trends of the gymnasium is similar to the 45 wind direction case, the roof block 1-15 interference factor of training hall is smaller, the minimum is.2, and interference effect is more obvious, the roof block interference factor of Competition Hall is smaller and most of roof block interference factor is about 1. (4) At the 27 wind direction, the roof block 1-15 interference factor of the training hall is smaller, the minimum is.18, interference effects are obvious, the geopolitical C PA
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan interference factor is smaller, most of the interference factors of roof block 1-62 of the competition hall is.8 ~ 1. of which the roof block 22-32 is smaller, most interference factors are about.6. Because of the stadium in the upper reaches of the gymnasium and its block the airflow, resulting in a significant shielding effect. The flow speed will be reduced to a certain extent, the wind pressure of the roof reduce, so the interference factor is smaller. Since the competition hall in the wake flow district, the shielding effect is not very obvious, the wind pressure of competition hall is affected a little and the interference factor is bigger. (5) The shape and position of the interfere affected the average wind pressure of the tested buildings, the stadium and gymnasium is a central axis symmetrical streamlined structure and the covering of the gymnasium is higher then the stadium which is benefit to the wind load of the stadium and gymnasium. When the stadium is in the lower reaches of the gymnasium, the flow blocked by the stadium, wind speed slows down, the wind pressure of gymnasium s roof reduces. When the stadium is in the upper reaches of the gymnasium, the gymnasium is blocked by the stadium, the wind pressure of gymnasium s roof reduces. Consequently, the architecture planning of the stadium and gymnasium is quite reasonable. 3 IF 2 1 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61-1 -2-3 1.25 1.75.5.25 IF -4-5 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 Fig.a. roof block 1-62 Fig.b. 45 roof block 1-62 IF 1.2 1.8.6.4.2 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 IF 1.9.8.7.6.5.4.3.2.1 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 Fig.c. 9 roof block 1-62 Fig.d. 27 roof block 1-62 Fig 6.The roof block interference factor distribution of gymnasium at some wind direction 4 Conclusions In this paper the wind loads of the long-span roof of the stadium and gymnasium in Quanzhou were calculated and the wind interference analysis between the stadium and gymnasium was carried. Some conclusions are as follows: (1) The wind loads was studied comparatively by wind tunnel test and CFD simulation and the law of wind pressure distribution is similar. The results in this paper verify the feasibility of numerical simulation of large-span roof buildings. (2) The stadium that in the upper reaches reduces the lift force of the approach cantilevered roof of the gymnasium, increases the lift force when in parallel position, reduces the lift force when in the lower reaches. (3) The wind pressure of the stadium's roof is affected by the gymnasium. When the stadium is in the upper reaches of the gymnasium, the flow blocked by the stadium, the wind pressure of gymnasium s roof reduces, shielding effect is not very obvious. When the
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 29, Taipei, Taiwan stadium is in the lower of the gymnasium, the flow is reduced by the stadium, the wind pressure of gymnasium in the training hall roof reduces. When the stadium is in parallel position, the wind speed increases between the stadium and the gymnasium, gap tube effect is obvious, the wind pressure of gymnasium s roof increases. (4) The shape and position of the complex affects the average wind pressure of the tested buildings, the stadium and gymnasium is a central axis symmetrical streamlined structure and the covering of the gymnasium is higher than the stadium which is benefit to the wind load of the stadium and gymnasium. Consequently, the architecture planning of the stadium and gymnasium is quite reasonable. Acknowledgements This research is supported by the following projects: the National Natural Science Foundation of China under Grant No. 5784, the Fujian s Science and Technology signal special project under Grant No. 25YZ116, the Xiamen s Science and Technology university innovation project under Grant No. 352Z28339, the Quanzhou s Science and Technology planned project under Grant No. 27G7. References Chen X. C. (25), the theoretical study and application of wind-induced response and equivalent wind load of large-span roof structure, Zhejiang University doctoral dissertation. Huang B. C. (22), Wind Resistance Analysis theory and Application for Structure, 2 rd Edition, Tongji University Press, Shanghai, 31-32. Khanduri, ACStathopoulos and T.Bedard C. (1998), Wind-induced interference effects on Buildings-a review of the state-of-the-art, Engineering Structure, 2 (7), 617-63. Lin B. (25), The wind tunnel test and CFD numerical simulation of the wind load of Daqing Petroleum Institute gymnasium roof, Xi'an: the structure of the twelfth National Conference on Wind Engineering, 724-73. Katarzyna Klemm,Wojciech Marks, Agnieszka J. Klemm. (2), Multicriteria optimization of the building arrangement with application of numerical simulation. Building and Environment, 35(1): 537-544. Richartson G.M., Hoxey R.P., Robertson A.P., Short L.J. (1997), The silsoe structures Building Com- parisons of pressures measured at full scale and in two tunnels, Journal of wind engineering and Industrial Aerodynamics, 72: 187~197. Shen G. H. (24), large-span roof structure of wind resistance studies, Zhejiang University doctoral dissertation. Suggestion of Japan. (1996), AIJ Recommendations for Loads on Buildings. wind-tunnel test research report. (25), The Civil Engineering State Key Laboratory of Disaster Prevention of Tongji University, The study of wind tunnel tests, wind-induced response and equivalent static wind load of Quanzhou Straits Sports Center Stadium.