Wind Tunnel Model Study Research Articles

Wind loads on double-skillion roof houses

ABSTRACT Recent damage surveys have shown that double-skillion roofhouses, characterised by two monoslope roofslopes with a vertical ‘rise’ connecting the upper and lower roof slopes, are vulnerable to cladding and structural failures in windstorms. Wind loading Standards such as AS/NZS 1170.2 do not currently provide pressure coefficients for deriving design wind loads specifically for these types of double-skillion roofs. This paper presents a 1/50 scale wind tunnel model study on a typical double-skillion roof house. The study found that the upper roof slope experiences large suction pressures especially near the upwind corner for oblique approach winds. In addition, the lower roof slope and rise are subjected to large positive pressures. The structure (i.e. rafters) near the end-walls also experiences large hold-down loads. Cladding loads and rafter hold-down loads are significantly larger than values obtained from applying data currently available in AS/NZS 1170.2

Wind load fluctuations on roof batten to rafter/truss connections

Batten to rafter connections in light framed housing can be vulnerable to progressive or cascading failures where a localised failure can cause the loss of a large section of the roof envelope. The synchrony of loads at neighbouring connections may affect the initiation of such failures. A 1/50 length scale wind tunnel model study was performed on a gable roof house to record spatial and temporal pressure fluctuations on the roof surface and data were studied to determine the flow separation mechanisms causing different loading patterns on batten to rafter connections. The cross correlation between load time histories was used to give a measure of synchrony between loads experienced at neighbouring connections and indicate the direction that fluctuations move across the roof. Orthogonal wind directions result in 2-dimensional flow separation that produce more synchronous loads at batten to rafter connections than cornering wind directions, where conical vortices produce high uplift forces on connections. The patterns of loading and their correlations give a means of identifying which parts of the roof and which approach wind directions may result in the initiation of a progressive failure of batten to rafter connections.

Wind loads on contemporary Australian housing

Design pressures given in wind loading standards are based on wind tunnel studies conducted during the 1970s to 90s on rectangular hip and gable roofs. However, most contemporary houses have complex hip-roof geometries with a range of plan footprints. Wind tunnel model studies were carried out on representative one- and two-storey houses to determine cladding design and truss hold-down loads. These loads were compared to design loads determined from wind loading standards AS/NZS 1170.2 and AS 4055. AS 4055 gave conservative design loads in most situations, and AS/NZS 1170.2 underestimated the loads near the windward edges and ridge on the roof. AS/NZS 1170.2 does give satisfactory design loads for wall cladding and truss hold down.

Alan G. Davenport's mark on wind engineering

This paper overviews the contributions to wind engineering made by the late Professor Alan G. Davenport. These include an engineering model of strong wind in the atmospheric boundary layer and its simulation in wind tunnels. His mean and turbulent wind profiles and his spectrum of longitudinal turbulence are known world-wide. He developed the theory of wind buffeting of line-like structures and introduced the gust factor method, which determines the magnitude of the peak wind-induced response, including the effects of wind-induced resonance. Also he has made key contributions to the development of statistical methods which are needed in the prediction of wind loads and wind-induced effects. These include the method of up-crossings, which predicts the probability of exceedance of particular extreme values. Foremost, he will be remembered for the wind tunnel model studies of many of the world's milestone buildings and structures.Professor Davenport's approach to wind loading was to combine its key components in a chain of thought, analogous to a physical chain, whose strength is determined by its weakest link. In recognition of this and his many other achievements, the International Association of Wind Engineering has officially named his approach to wind loading as the “Alan G. Davenport Wind Loading Chain”.

Vulnerability of roofing components to wind loads

The vulnerability of roofing components of contemporary houses built in cyclonic regions of Australia is assessed for increasing wind speeds. The wind loads and the component strengths are treated as random variables with their probability distributions derived from available data, testing, structural analysis and experience. Design details including types of structural components of houses are obtained from surveying houses and analyzing engineering drawings. Wind load statistics on different areas of the roof are obtained by wind tunnel model studies and compared with Australian/New Zealand Standard, AS/ NZS 1170.2. Reliability methods are used for calculating the vulnerability of roofing components independently over the roof. Cladding and batten fixings near the windward gable edge are found to experience larger negative pressures than prescribed in AS/NZS 1170.2, and are most vulnerable to failure.

局所相似モデルの概念と風洞実験による検証 : 通風時の換気量予測法に関する研究(第1報)

A new model has been proposed for evaluating the discharge coefficient and flow angle at an inflow opening for cross-ventilation based on the previous analysis of the flow structure by the Large Eddy Simulation. This model is based on the idea that the cross-ventilation flow structure in the vicinity of an inflow opening creates dynamic similarity under the condition that the ratio of cross-ventilation driving pressure to dynamic pressure of cross flow at the opening is consistent. Wind tunnel model study was carried out to reproduce cross-ventilation by attaching suction fan at the leeward side of the building model. Discharge coefficients at stagnant surrounding condition with suction flow correspond well with those exposed to external wind with extremely large suction flow. Total pressure at the opening is preserved even when flow is separated at the upstream corner. It was confirmed that the proposed model can be applied to almost any directions and opening positions.

Design Wind Loads on Gable-Ended Low-Rise Buildings with Moderate and Steep Roof Slopes

Wind tunnel model studies were carried out to determine the external wind pressure distributions on large, low-rise buildings with roof pitches of 15° and 35°. The studies showed a significant increase in the magnitude of the mean and peak negative pressure coefficients with an increase in length to span aspect ratio, on the downwind roof and leeward wall, near the windward gable-end of buildings with a steep 35° pitch roof, for oblique approach winds. These oblique approach winds also generate large design wind load effects on the frames near the gable-end. There is little variation in mean and peak pressure coefficients with aspect ratio, across the frame of the building with a moderate 15° pitch roof. AS/NZS 1170.2, which specifies increased suction pressures with aspect ratio on the downwind roof slope and leeward wall of steep pitch roofbuildings compared to the previous edition of the Standard, provides satisfactory design wind load effects on frames near the gable-end. However, some design load effects on frames near the gable-end of the 15° pitch roof building are underestimated by AS/NZS 1170.2.

Fluctuating wind loads across gable-end buildings with planar and curved roofs

Wind tunnel model studies were carried out to determine the wind load distribution on tributary areas near the gable-end of large, low-rise buildings with high pitch planar and curved roof shapes. Background pressure fluctuations on each tributary area are described by a series of uncorrelated modes given by the eigenvectors of the force covariance matrix. Analysis of eigenvalues shows that the dominant first mode contributes around 40% to the fluctuating pressures, and the eigenvector mode-shape generally follows the mean pressure distribution. The first mode contributes significantly to the fluctuating load effect, when its influence line is similar to the mode-shape. For such cases, the effective static pressure distribution closely follows the mean pressure distribution on the tributary area, and the quasi-static method would provide a good estimate of peak load effects.

Effect of building length on wind loads on low-rise buildings with a steep roof pitch

A wind tunnel model study was carried out on long, low-rise buildings with a steep roof pitch to determine the effect of the length-to-span aspect ratio on the external wind pressure distributions. The study showed a significant increase in the magnitude of the negative pressure coefficients on the leeward roof and wall, with an increase in aspect ratio, for oblique approach winds. These large suction pressures also generate large design wind load effects on the frames near the gable-end. The 1989 edition of the Australian standard for wind loads, AS 1170.2-1989 was found to underestimate the wind loads on steep pitch gable-roof buildings of aspect ratio greater than 3, on areas near the windward gable-end, and hence the critical bending moments in the supporting structural frames. The current Australian/New Zealand wind load standard, AS/NZS 1170.2-2002 specifies increased negative pressure coefficients on the leeward half of high pitch roof buildings, and critical bending moments in the supporting frames calculated from these distributions agree quite well with values obtained from the wind tunnel study. However, other major standards severely underestimate the critical bending moments, and the effective pressure coefficients producing those bending moments, especially on the leeward roof slope.

Wind-induced responses of tall buildings experiencing complex motion

A wind tunnel model study on the effects of eccentricity between centre of mass and centre of stiffness of a tall rectangular building on the wind-induced response characteristics and wind excitation mechanisms was carried out. The results indicated significant effects of eccentricity on the along-wind, cross-wind, and twisting moment responses for both cases of the incident wind normal to the narrow and wide faces of the building model. The periodic twisting moments, created either by the flow separation at the upstream corners and the reattachment process of the separated shear layers, or by the eccentricity between the resultant lift force and centre of stiffness, were found to be a dominant wind excitation mechanism causing a significant increase in the twisting moment response of the building model. In addition, the twisting moment responses could be greatly amplified by the effects of resonance between the frequency of the periodic twisting moment and the torsional natural frequency of the building model. For the building model with the centre of stiffness located downstream from the centre of mass, increases in the twisting moment response and the resultant acceleration response at the corner of the building model were more prominent.

Wind loading of parallel free-standing walls on bridges, cliffs, embankments and ridges

This paper describes a study of boundary-layer wind loads on parallel walls, used as noise barriers on freeways and motorways. Approximately 20 representative situations were studied in a wind-tunnel model study. The paper gives details of the experimental techniques used, comparison with previous wind-tunnel tests (flat ground only) and the main results of the study.

Wind tunnel model studies to predict the action of wind on the projected 558 m Jakarta Tower

A study of wind effects was carried out at the Boundary Layer Wind Tunnel Laboratory (BLWTL) for the projected 558-m high free-standing telecommunication and observation tower for Jakarta, Indonesia. The objectives were to assist the designers with various aspects of wind action, including the overall structural loads and responses of the Tower shaft and the antenna superstructure, the local wind pressures on components of the exterior envelope, and winds in pedestrian areas. The designers of the Tower are the East China Architectural Design Institute (ECADI) and PT Menara Jakarta, Indonesia. Unfortunately, the project is halted due to the financial uncertainties in Indonesia. At the time of the stoppage, pile driving had been completed and slip forming of the concrete shaft of the Tower had begun. When completed, the Tower will exceed the height of the CN-Tower in Toronto, Canada by some 5 m.

Surface shear stress measurement system for boundary layerflow over a salt playa

A novel method for measuring shear stress via a floating surfaceelement beneath a turbulent boundary layer has been tested and implemented.The final design is used for measuring the surface shear stress in theboundary layer that forms over the salt playa of western Utah. Wind tunnelmodel studies were conducted to establish the characteristics of two competingsub-designs, and to determine their sensitivities and potential error sources.The wind tunnel based experiments included independent comparisons with theresults of wall velocity gradient based measurements using hot-wire data.Overall, the results show that a floating element on a pool of liquid, herewater, provides a more accurately measured surface shear stress than thatdetermined utilizing an air bearing. In addition, the water pool technique issimpler to implement and produces a higher degree of repeatability. The windtunnel results provided the design criteria needed to build a larger-scaledevice used to measure the surface shear stresses over the Great Salt LakeDesert in western Utah. This final design has been implemented and fieldresults are presented and compared with sonic anemometer based estimates ofthe wall shear stress.

Wind load effects and equivalent pressures on low-rise house roofs

Wind loads on the roof of a typical low-rise house are determined by carrying out a wind tunnel model study. The region near the windward gable end are subjected to the largest wind loads. The large suction pressures near the ridge for oblique approach winds exceed data given in AS1170.2. The mean and fluctuating wind pressures and load effects on the roof tributary supported by the second truss from the windward gable end (i.e. truss subjected to the largest wind loads) are analysed in detail. Smaller load effects are generally obtained by applying the covariance integration method with wind tunnel data, compared with values derived from using AS1170.2 data. The equivalent static pressure distributions obtained from the load–response-correlation method are generally smaller than the peak pressures derived from AS1170.2, except near the ridge for oblique approach winds.

Mean wind loading at the leading ends of free standing walls

Full-scale mean wind loads on panels within free standing walls are compared with previous and new wind tunnel model studies. The high loads adjacent to free ends are evident in all studies. Various methods to reduce this end-panel loading were tried in the wind tunnel to examine their effectiveness. It was observed that short return walls and wall porosity did not significantly reduce the loading as previously suggested.

Deposition of large particles from warehouse fire plumes—a small-scale wind tunnel model study

The report describes measurements of the deposition of large particles from a small scale wind tunnel model of a chemical warehouse fire plume. A common feature of such fires is the discharge of relatively large particles with falling speeds of the order of m s −1, partly generated by mechanical damage, which can fall out of the fire plume in a different pattern to that of the gases and fine particles. These large particles may also contain toxic components, so it is desirable to know their fallout pattern. The deposition of large particles from fire plumes has therefore been modelled directly as an adjunct to earlier small scale wind tunnel experiment on gaseous plume dispersion [1]using the same experimental conditions, so that the relative behaviour of gas plume and heavy particle dispersion could be compared. As far as we are aware, this is the first experiment of this type to be carried out. There are constraints on the range of deposition conditions that can be modelled, due to scaling problems. However, it proved possible to develop a viable technique and some useful data was obtained over a range of particle and plume conditions of interest. Nonetheless, the work should be regarded mainly as a proving trial of the technique and an indicator of the significant parameters as the measurements made were too few in number (essentially of three particle deposition conditions) to provide a broad-based indication of large particle deposition. Experiments were carried out for two gas plumes, buoyant and non-buoyant, both with and without a building shell around the source. Large particle concentrations near the surface proved markedly different to those for a gaseous plume, showing a much more rapid reduction in concentration with increasing distance. Both particle falling angle and particle inertia affected the particle plume dispersion, the highest concentrations at the ground occurring for the case with small falling angle and inertia, the lowest concentrations for the case with large falling angle and moderate inertia. Concentrations for the remaining condition, with large falling angle and high inertia, fell between these two, probably because the high inertia constrained the rate of lateral dispersion of the particles. Compared with the effects of particle falling angle and inertia, the effects of a building shell around the source were relatively limited. This was greatest for the non-buoyant plume where the building shell tended to act as a trap for large particles. The buoyant plume tended to carry particles beyond the building shell, so that in this case its effects were quite limited.

Wind-Tunnel Studies of Buildings and Structures

The ASCE Standard 7 and many other codes of practice now allow the use of wind tunnel model studies for developing wind loading information to supplement and, in some situations, replace code wind loads. As a result, it is important for the engineering community to be aware of the circumstances in which wind-tunnel model tests may be advantageous or even necessary and to become familiar with the underlying physical principles and procedures, which must be followed to ensure that meaningful information is obtained. The ASCE Manual 67 was published in 1987 to assist engineers, architects, and other users of wind-tunnel services. The wind-engineering field has been evolving rapidly, and members of the Aerospace Division have been busy in developing an update. This new manual is currently in the ASCE Blue Ribbon Review process and is scheduled to be published in 1996. The new edition of the Manual of Practice for Wind-Tunnel Studies of Buildings and Structures has two distinct parts. Part 1 is an expanded version of ASCE Manual 67, which has been updated to reflect changes in technology and which has additional information for model studies of atmospheric dispersion around buildings, quality control, and procedures for combining wind-tunnel findings with climatic wind data in order to provide statistical predictions of full-scale wind loads and wind-induced effects. Part 2 is a detailed commentary that provides explanations, supporting information on methodology, and examples of typical tests; it also includes an extensive bibliography. The material presented provides a preview of this new manual. Due to space limitations, only part 1 and the bibliography are included at this time. The full manual will be available in 1996 and interested readers can get further information from the ASCE.

Generic models for pedestrian-level winds in built-up regions

Wind conditions over streets in built-up cities were examined in a boundary-layer wind-tunnel model study. Presented in this paper are results of wind speeds affected by a number of parameters such as the spatial density of street blocks, the building height over surroundings, the relative location of buildings and the direction of approaching wind. Based on literature information and current findings, some generic models have been established with the empirical relations between wind conditions and building configurations. These models can be applied for the preliminary estimation of pedestrian-level wind environmental conditions in built-up regions.

The similarity requirements for wind tunnel model studies of gas diffusion

The similarity requirements for wind tunnel model studies of buoyant and heavy gases diffusion near buildings are presented and discussed. Exact similarity of buoyant and heavy gases behavior requires explicit scaling of the released gas density and equality of the model and full scale Froude numbers. The latter requirement leads to very low wind speeds for wind tunnel tests when small scale models are used. This condition makes the physical modeling of flow and diffusion fields complicated and thus reliable wind tunnel operation becomes very difficult. Approximate modelling methods, which relax the requirement of gas density scaling in order to operate at higher wind tunnel speeds, are also examined.

Dispersion in an open-cut coal mine in stably stratified flow

Discharges from combustion within a coal pit which occur during night-time inversion conditions may result in stagnant accumulation of smoke and dangerous gases which could inhibit mining operations. A wind-tunnel model study was performed to identify the range of flow and mixing conditions which could exist when stably stratified atmospheric surface flows pass over a large open pit. Flow penetration into the pit depended upon approach-flow stability (Froude number) and the strength of the thermal inversion within the coal pit. Measurements of wind speed and temperature were made upwind, within and downwind of the pit. Concentration measurements were made within the pit, of surface sources released along pit walls. Pollutant levels were found to be strong functions of the approach-flow pit Froude number, source location, and release time.