Click for next page ( 56


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 55
3 Wind-Engineering Research Needs INTRODUCTION Wind-related problems cover a wide range of topics including damage to both low-rise structures (housing and small commercial and industrial buildings) and mid- and high-rise buildings (residential and office), bridges, towers, stacks, and power lines. In addition, a number of non-w~nd-Ioad- related issues deserve research attention. Reentrainment of exhaust gases from laboratory building fume hood exhausts into building ventilation systems and local dispersion of toxic gas releases or of toxic fiances from accidental spills represent health and safety issues to which wind engineering can make major contributions to eliminate or reduce the hazard. These are areas that do not have adequate design methodologies at present. Other issues that affect the economic success of a project include pedestrian acceptability of wind and blowing dust or snow near building entrances and in plaza areas. Wind-engineering research can also result in the reduction of soil erosion caused by wind and the proper modeling of iceberg movement and of] spills, both of which are important issues related to offshore oil operations. Moreover, wind energy potentially represents approximately 10 percent of electric system installations in the United States. The United States currently sustains several billion dollars per year in property and economic Tosses due to windstorms, along with significant loss of life. In September 1989, Hurricane Hugo caused $4 billion to $5 billion in insured Tosses, yielding a total loss value of $7 billion to $S billion to the U.S. mainland alone. Yet even with this large actual Toss record and tremendous potential loss waiting to occur, the United States spends no more than about $4 million per year on wind mitigation, most of which is allocated to storm warning capability. Less than $] million is spent each year in wind-eng~neenng mitigation research (National Research Council, 1989~. Indeed, even the annual loss figures presented above suffer from a high degree of uncertainty due to lack of funding for research into windstorm effects The United States has not commissioned a study of wind damage loss assessments since the mid-1970s. In spite of the lack of funding, engineers have explored the reasons for the heavy financial Tosses in the United States. A task committee of the American Society of Civil Engineers (ASCE) published a series of nine, coordinated papers in 1989, which detailed many of the reasons for large losses and outlined a number of solutions (American Society of Civil Engineers, 1989~. The insurance industry also published a document in 1989 demonstrating that strengthening housing units during construction to resist _ so

OCR for page 55
56 Bred and the Built Environment hurricane winds would result in a minimal additional cost in the range of I.5 to 4.0 percent (All-Industry Research Advisory Council, 1989~. Although much is known about how to mitigate wind damage, much is yet to be learned. Sustained research is needed In the areas of defining wind loads, determining more economical ways to resist these wind loads In both new and existing structures, and finding ways to implement solutions In the construction environment, where training of construction workers is limited and inspection is a cost that local communities would prefer not to incur In this chapter, various areas of wind eng~neenng are identified and the issues related to research are presented. RESEARCH MEI]IODOLOGY Various methodologies have been developed to advance our knowledge base in wind-engineenug research and practices. These methodologies include physical modeling using atmosphenc, boundary-layer wind tunnels; numerical modeling taking advantage of powerful computing capability; and full-scale field measurements to vend predictions made by physical and/or numerical modeling. In addition, innovative experimental approaches, such as tornado simulation, are required to address particular problems (Lund and Snow, 1991~. Recent developments in the area of probabilistic methods and statistical inference provide powerful tools to facilitate implementation of uncertainties arising from the complex nature of wind problems. Post-wind- storm disaster investigations also provide the best and most direct way to assess the current state of w~nd-hazard mitigation practices and can contnbute significantly to our efforts to reduce the threat of hazards. Physical Modeling That portion of the atmosphere within the first 1000 to 2000 ft of the earth's surface comprises the atmosphenc boundary layer (ABL,J. Physical modeling in wind tunnels of atmospheric winds within the ABE has matured over the past 30 years. This modeling capability permits engineers to address the effects of winds on the built environment on a routine, en~neering- design basis. Windflow about buildings and over complex terrain, wind loads on structures, dispersion of pollutants, wind ejects on pedestrians, snow or sand deposition and drift, and heat transfer from structures are some of the problems that can be addressed through wind tunnel modeling. BoundaIy-layer wind tunnels are designed specifically to mode} the variation in mean wind speed with height above ground and the vertical distribution of turbulence, or gustiness, in the ABL. Figure 3-1 shows how the wind speed varies with height and how that variation changes with the character of roughness on the ground surface. Boundary-layer wind tunnels differ from aeronautical wind tunnels in that they contain long test sections to allow the development of appropriate velocity and turbulence profiles.

OCR for page 55
Wnd-Engyneenng Research Needs 57 M.u, Spoed Promo Anita R~ of Gust fluciua6~ ~& ~ as) ~1 1500 - ~D 1000 - ID Or 500 2000 . 1 500 ~ntVh~d10O1 1000 / // /7~, 500 Gradient ~ JOO' _ , ' /~/ 2000 1 500 100~) 500 o m Wed JOq 8g1 771 //~/ FIGURE 3-1 Mean wind velocity over level terrains of differing roughness. Source: Davenport, ~ G. 1967. Gust Loading Factors. Journal of Structural Division. Reproduced by permission of American Society of Civil Engineers. Figure 3-2 shows the first major boundary-layer wind tunnel built at Colorado State University in 1961. Scale models of terrain, buildings, bridges, power plants, or other features of interest can be installed in the wind tunnel for study. Figure 3-3 shows a Apical wind tunnel study for wind loads on a structure, in which a mode] under study, along with a mode} of the surrounding city, has been installed in a boundary-layer wind tunnel. The thermal structure of the atmosphere is important in many applications, such as dispersion of pollutants. Boundary-layer wind tunnels can be constructed to include the temperature variation of the atmosphere in the mode} wind simulation. Stably stratified winds (when the ground is colder than the air above, such as night conditions or winds over water or snow) or unstably stratified winds (when the ground is warmer than the air above, such as winds on a hot, Sumner day) can be modeled by appropriate cooling and heating of the model ground surface or air. Modeling technology is sufficiently developed that several consulting engineering firms are now routinely testing for wind loads on buildings, dispersion of pollutants, pedestrian comfort in wind, and snow deposition and drifting. These studies are inexpensive for major development projects, often accounting for less than 0.1 percent of the total project cost. In some cases, cost savings of 5 to 10 percent or more may be obtained as a result of these tests. A wide variety of wind tunnel tests are available on a commercial or research basis to define wind effects on structures. A plastic model of a

OCR for page 55
Bill - F AS of 58 .G , ~ o C' . ~ ~ . _ , ._ ~ (D CL C' ~ TIC _ -_ ~ _ ,, ~ J l ~ 1 ret ~ ~ Cal 1 ~ | ~ I l a' (D _ ~ _ . I) ~ ~ CO CO ._ _ ._ W X C ~ X IS elf - V, 0 C O =_ ~ 3 ~= c o o %. ~ ~ ~ ~ C ~ ~ _ ._ ~ W~ ~D'- o e_ lL C ~_ C0 O C~ C ~ ~e _ _ o V o le~ o Ct - o e_ C~ _ V, C~ e_ ct e_ e_ - Ct C) e_ o o o ~Y ~ V

OCR for page 55
~~-- ~ Amp :;:::::::::::: ::~: :~: - .- Isis :` ~ -: : ~ -:s - ~:~111~1~ :~s -it - ;~:~ ^: ~ At: OGRE 3-S Model under smog in ~ bound layer wind Amp. Source: Cell Peterka Peterson Inc. Encore under saw can be ins1~mented a1 hundreds of location 10 measure local Ouch pressures far use in design of 1be sl~cmre~s daddy. For End loads scam on ~ lamer panes groups of pressure sensors can be combined 10 he ares-=eraged pressure forces. ~1ema~e~, ~ model ~ he s1mcm~e can be moused on ~ balance 10 medium he i~l=1=eou~ Acme End loads Spay 10 1be entire s1mcmre. ~ silly bounce can also merge the combined, ~em~ plied wind load and 1ntem~ inertia remove lose of He sincere. For over ~pUc~io~, tracer amen such ~ 1~-conce~rsdon ~~oc~o~ or c^on diadem can be released Tom points of poDudon release ~ 1be wind heel model 10 measure 1beir mean Ad peak concerto ~ dad pOiD1S of interest gab a gas cbroma10gr~b. In L w~ far =~' 1bo hachon of ~ 10~C gas released ~ a fame good -=s1 on ~ build roof ~1 ~ reentrained into ~ building intake can be measured Ad modified Ada 1be design pie of ~ budding. Figure S-4 she dab Mom s roof veal caused by And 0~ around 1be building. If ibis plume contained Sac gases, reen1rsi~eD1 iD10 Dewy buDding-~r bakes could pose bomb conceal. Correcdon of ibis he of problem her he budding is placed info seance can be veal heave Ad peyote m be selected 10 herons leek of 10~C ma1edal~ he same measurement Coach can be used iD emerging response plying far sccidenlal spins or g~ releases or 10 design deuces 10 ~~( Con1HiD, 0[ dell those releases.

OCR for page 55
60 Hued and the Built Environment FIGURE 34 Building exhaust downwash model under test in wind tunnel. Source: Cermak Peterka Peterson, Inc. Wind tunnel tests can also be used to measure wind fluctuations in pedestrian areas near structures so that pedestrian wind exposur~and consequent comfort levels can be gauged. In addition, winds in approach areas to helicopter pads can be measured to identity high turbulence or wind shear that could endanger helicopter operations on windy days. Wind-speed measurements in wind tunnels can also be used to locate anemometers to prevent shielding from upwind obstacles or to find correction factors for data from shielded anemometers. Transport of particulate material is a highly complex process that, in many cases, can be modeled satisfactorily in a wind tunnel. Snow deposition and drift in the presence of wind or snow drifting are important for designing building entrances, highway geometries, parking lots, or pedestrian walkways to avoid drift buildup, and accompanying maintenance costs. Depths of snow drifts on roofs can be estimated to aid in design of roof structures for snow load. For small projects such as single- or multiple-family dwellings, small commercial structures, or industrial buildings, wind turmel tests are often too costly relative to the overall project cost. For these structures, which make up the overwhelming bulk of construction in the United States, reliance must be placed on building codes to define wind loads. Because of the high cost of full-scale tests, wind tunnel data, obtained on a research basis, have been used to develop the w~nd-loading provisions of these codes.

OCR for page 55
~nd-Engyneenng Research Needls 61 Unfortunately, funds have been insufficient to develop, through research and testing, the level of understanding of wind loads that would provide maximum economy of construction. Currently, no private organizations possess the resources or mission to fund the appropriate level of research needed for code development. Neither has the National Science Foundation nor any other federal agency found sufficient funds to support such an effort. Thus, research funds are vitally needed for better definition of wind loads in a codified form. Numerical Modeling Numerical modeling of turbulent flows around complex boundaries is difficult and demands the use of very large computers. Consequently, the calculation of wind flow around structures, of the dispersion of pollutants over buildings or complex terrain, and of wind loads on structures using numerical solutions of the governing differential equations is still in its infancy. These calculations have the potential to predict wind flow, pollution dispersion, and wind loads with great accuracy. To date, the mean concentrations of pollutants far downwind of buildings and over some complex terrain features and the mean pressures on simple cubical buildings have been predicted with some skill. Fluctuating concentrations or pressures are still not within the range of calculation capability, nor can mean values be calculated for other than very simple geometnes. The availability of inexpensive but powerful computers will make possible significant progress in numerical calculation capability in the near future. Still needed are improved numerical algorithms, protocols for easy implementation of parallel processing, and improved graphical display software. For many years, empirical mathematical models have been used for the prediction of pollutant dispersion by winds. However, in many instances, these empirical models are subject to large errors. Numerical solutions to the governing equations could significantly increase the accurac y of such predictions. Two major areas in which research is needed are fast algorithms arid turbulence modeling for winds in which temperature gradients are important. The development of numerical modeling warrants special attention. Initially, the requirements wall be for groups with special expertise working in cooperation with others interested in computational fluid dynamics and with access to powerful computers. Further development should then lead to the broader usage of these techniques. Field Measurements Field measurements of wind and its effects are necessary to validate predictions made by physical or numerical modeling. Complexities in field

OCR for page 55
62 Wnd and die Built Environment measurements arise from the surrounding topography, the wind climate, the unpredictable nature of wind, the complicated structural system of buildings, and the unavailability of off-the-shelf instrumentation. In addition, perhaps the most basic, intrinsic difficulty of field measurements is their nonrepeatability. The natural wind has too many large variations in intensity, direction, and turbulence levels, as well as in meteorological aspects, to elect it to repeat itself precisely. These complexities make field measurements expensive and time consuming. New technologies in instrumentation and data acquisition systems provide opportunities in field measurements that were not available previously. Solid-state electronics, remote sensing devices, and computerized data acquisition systems permit reliable and detailed measurements. Data on the response of high-r~se buildings, window glass, and other structures or elements can be obtained using these technologies. To reduce the cost of field measurements as well as to gain meaningful results, field measurements should be targeted for specific objectives. A few, carefully plaT,ned, long-term experiments should be conducted in the field to provide baseline data. These experiments should be supplemented by well- targeted, short-term experiments using mobile equipment, perhaps in the region of hurricane landfalls. The site for an experiment should be topographically clean with frequent high winds. Complex topography should be avoided. A fairly extensive array of meteorological instruments must be deployed to assess gust sizes and directions, air densities, and storm characteristics. It is necessar, to have at least one tall meteorological tower and a few short ones at each site. The structures for which responses are measured should be structurally simple: transmission lines, long-span bridges, or free-standing towers are good candidates for measuring responses. International, cooperative projects with other windstorm-prone countries can help to reduce manpower costs and speed up data collection. Field measurements have the potential of providing significant benefits In the future. It is recommended that a few, long-term field experiments be established to provide baseline data. Along with these expenments, physical modeling in the wind tunnel and numencal modeling should be planned to develop improved modeling technologies. Innovative Experimental Approaches Although boundaIy-layer wind tunnels provide a major too} for exploring wind loads on structures, they cannot provide all necessary experimental capability. These facilities do not, for example, directly simulate tornadoes, thunderstorm outflows, or hurricane eyewall winds. The extent to which these phenomena are different in wind-Ioading characteristics from the straight- line winds associated with extratropical windstorms or frontal passages, which boundary-layer wind tunnels simulate well, is unknown. Investigation into unique physical facilities that could mode! these phenomena should be

OCR for page 55
~nd-Eng~neenng Research Needs 63 undertaken. The ability of numerical models to perform this type of modeling should also be investigated. Some laboratory simulators have been used to mode! tornadoes on small scales. These facilities provide great insight into the structure of tornadoes, but they are not funded at a level sufficient to exploit their full potential. A combination of laboratory measurements, numerical simulations, and fuB-scale testing is the most likely route to fuller understanding of tornado flows. Larger laboratory facilities than currently exist would benefit this investigation. Improvements in Doppler radar might provide field measurements of tornado velocity fields for comparison to laboratory and numerical simulations. Additional improvements in radars or other instruments capable of measuring tornado velocity fields are also needed. The resistance of small structures, such as wood frame buildings, small office or business buildings, and masonry buildings to wind loads is currently difficult to predict. Model versions of these structures cannot duplicate their complicated failure modes. Physical modeling facilities are needed in which simulated wind loads can be applied to full scale models of such smaller structures. Otherwise, a very large wind tunnel can impose the wind loads by winds occurring in nature. Research is needed to determine the most effective method for testing these structures and to design and build such facilities. A numencal component should accompany this research effort so that the results from numerical studies can be used to reduce the extent of the needed physical studies. Facilities to mode} wind-wave action on structures are also needed to better analyze and quanta y the complex, fluid-structure interactions expenenced by offshore platforms. Structural Safety and Reliability Traditionally, structural safety in the design process is ensured by including appropriate safety factors to account for shortcomings stemming from a lack of full knowledge. from insufficient data. or from inherent . O , , _ _ ~ .~ . . ~ ~ t ~ ~ ~ ~ ~ ~ ~ ~ _ vanan~ty In tne pro Stem s parameters. Untortunately, the factor ot safety concept does not provide a quantitative measure of stn~ctural safety or reliability. Probabilistic assessment of structural safety is receiving increased attention and acceptance with the emergence of probability-based design formats such as the load and resistance factor design. Recent developments in the area of probabilistic methods and statistical Inference offer a convenient mathematical framework to cope with uncertamnes arising trom a variety of sources. Research is needed to quantii y the uncertainties associated with venous problem parameters, to examine propagation of these uncertainties, and to assess the influence of uncertainties on the design process. ~

OCR for page 55
64 Wind And the Buill Environment Damage Investigations Post-wind-sto~m disaster investigations offer the best opportunity to assess how successfully current knowledge and technology are being applied to reduce the Impacts of extreme wind hazards. The examination of darr aged structures and lifeline systems allows wind engineers to identity future needs in wind-engineenng research and practices, and to develop better ways to effectively implement these technologies once they are developed. SEVERE WIND FORCES J Windflow over bluff bodies, such as buildings or bridges, is different in many respects from flow over streamlined bodies such as wings or airfoils. Such bodies are immersed in the atmospheric, boundary-layer flow in which the mean velocity vanes rapidly with height above ground and the turbulence in the approaching wind is much higher than that which typically impinges on an airfoil in flight. The turbulent nature of the wind has been illustrated in Figure 2-16, which shows that wind is composed of a mean with turbulent fluctuations (gusts). The gustiness in the wind and its variation with height above ground (Figure 3-~) are the main features that distinguish it from aeronautical flows characterized by smooth flow. Bluff body aerodynamics control wind loads on such diverse structures as buildings, bridges, tall solid towers, trussed towers, stacks, cooling towers, and cables. The basic w~nd-Ioad mechanisms are buffeting caused by turbulence (gustiness) in the approaching wind, wake excitation caused by turbulence generated in the region immediately downwind of the structure, wake excitation caused by vortex shedding from the structure (periodic shedding from alternate sides of the body of packets of fluid that have rolled up into a rotating mass), aeroelastic effects in which the wind loads are altered by the motion of the body, and galloping excitation caused when aerodynamic damping (resistance to motion in a fluid caused by fluid viscosity) and mechanical damping (resistance to motion caused by internal friction within the structure itself are overcome by aeroelastic effects. Local cladding loads on the surface of buildings are a significant design issue because water leakage through unintentional wall and roof cracks is a major cause of building damage. The largest fluctuating pressures usually act outward from the building surface and can cause cladding to fall during a storm. Although w~nd-Ioad codes specify local wind pressures, wind tunnel tests often find significant variations from that loading. One of the largest local pressures on a building is frequently found near a corner of the roof. The flow mechanism responsible for this phenomenon is called roof vortex, shown in Figure 3-5. Additional research into local pressure fluctuations has the potential to improve empirical Dredictinns of Thin nr~cc,~r~c for ,,~- in building codes. Wind loads on structures fluctuate randomly in time in response to random changes in wind speed and direction and to random pressure .__ of -v A- _- r~01~O lV1 Urn

OCR for page 55
~nd-Engyneenng Research Needs 65 I. ~ ~ ~~ ~ . ~ ~ .... .... FIGURE 3-5 Roof vortex in wind tunnel mode! leading to high roof wind loads. Source: Fluid Mechanics and Wind Engineering Program at Colorado State University. fluctuations generated by the wind flow about the structure. Wind-Ioading mechanisms have not been satisfactorily descnbed mathematically because of the extreme complexity of the turbulent flows responsible for w~nd-Ioading variability. As a result, wind loads are now largely expressed empirically. New developments in chaos theory provide possible avenues for understanding the nature of turbulence and wind loading in the presence of turbulence. An analytical component of wind-Ioad research emphasizing chaos theory might provide advances in understanding these complex w~nd-Ioad mechanisms. Low Buildings A few research projects have been funded by the metal buildings industry to study design loads appropriate for small buildings. As a result, a revised w~nd-Ioad code has been produced for use in the design of

OCR for page 55
68 Wnd arid die Buill Environment Damping ant! Stn~cture Control To decrease the effects of fatigue, structure defections can be reduced by increasing the structure stiffness or by increasing its damping ability. Prediction of damping in many structures, including buildings, is relatively primitive. The uncertainty in wind loading on a building can easily be 50 to 100 percent due to uncertainty about damping behavior. Increasing the damping In structures could significantly decrease the cost of construction of buildings, bridges, towers, and other flexible structures. Recent research and practice indicate that design for control of damping in structures is within reach if adequate resources can be applied to research into damping methods. Current damping practice includes the use of tuned mass dampers (massive weights near a building top that are attached to the building frame through spnngs), nscoelastic dampers (thousands of small devices placed throughout a building to dissipate kinetic energy in the structured, or aerodynamic fanugs (changes in the structure shape to reduce the wind loads causing the motion technique used on bridges but not practical for buildings). These approaches to limiting motion are relatively expensive. Active control of structure motion is a promising research area for reducing the cost of many structures. It involves the sensing of structure motion with a control system that activates motion reduction dences (such as a tuned mass damper). Once activated, the system tends to reduce the structure motion. Active control devices are common on aircraft and have the potential to significantly reduce the cost of engineered structures susceptible to w~nd-induced motion. Considerable research is required before these devices could be considered to control wind-induced motion of buildings, bridges, or towers. Similar active systems, properly designed and installed, can also pronde potential benefits to the structure's resistance to earthquake loads. Sloshing of fluid in a tank is another method for dissipating energy in a structure. Limited applications have been used in the United States to restrict motion in water towers. Recent research demonstrates potential application for damping of buildings and bridges by using fluid sloshing. Research in the United States is needed to develop fluid-lioshing damping technology to a practical design level. Bridges Conventional bridges are not very sensitive to the dynamic effects of wind because of their relatively high stiffness. On the contrary, suspended- span bridges, which include suspension and cable-stayed bridges, are very sensitive to wind effects. In addition to buffeting effects of wind, they are susceptible to aeroelastic effects, which to a great extent caused the Tacoma Narrows disaster. These suspended-span bridges are often even more sensitive to wind during various construction phases than they are after completion.

OCR for page 55
Wnd-Eng~neenng Research Needs 69 The aerodynamic stability of a bndge is governed by the bridge geometry, its spectrum of natural frequencies, and its damping. Information on the aerodynamic behavior of bndges is determined through physical modeling in wind tunnels. A wide range of mode} tests are available, such as section models, taut strip models, and full-bridge models. Section model tests, which are currently the primary investigative tool, help to determine aerodynamic characteristics of the bridge section that are then utilized In an analytical mode! to determine the bridge's overall dynamic behavior. However, an analysis based only on section mode! test results conducted in smooth flow often fails to describe the three-dimensional bridge behavior in natural wind conditions. Improvements have been suggested in this regard, such as appropriate modeling of turbulence in the approach flow. Full-br~dge models, if they are both structurally and aerodynamically accurate, can provide information on the overall dynamic behavior of the bridge and offer the convenience of modeling the surrounding terrain to accurately sunulate the approach flow conditions. Motion reduction dewces can help to improve the aeroelastic stability of a bridge. These can be considered during the design phase or can be incorporated once the bridge is built. The behavior of freestanding bridge towers during construction deserves special attention and, often, motion reduction devices are needed to control their motion. Motion Perception Many structures move sufficiently in the wind that occupants can sense the motion and may object to its magnitude. These structures include office buildings, residential buildings, offshore platforms, airport control towers, bridges, and other flexible structures. Very little research has been performed on the levels of motion that are acceptable for various uses. The design of many buildings is governed by the acceptability of motion to its occupants; the current level is based on only a few, uncontrolled studies. However, the constriction costs of some structures might be significantly reduced if these levels were relaxed on the basis of more solid research into acceptable levels of motion. Offshore Winds and Their Effects Wind-related issues concerning offshore drilling activities may be divided into two general areas: the design and analysis of offshore structures, and offshore exploration and operation. Wind speeds at various return intervals are the essential input for the design of conventional fixed structures. For these relatively stiff structures, ondy the steady wind effects are of interest and they typically contribute less than 10 percent of the total environmental Toads. However, for exploration deep water, where conventional platforms may not be appropriate due to

OCR for page 55
70 Wind and the Built Environment their sensitivity to the dynamic wave load effects, attention has been focused on the development of innovative structural systems to explore the frontiers of offshore reserves. The most promising of these systems are the so-called compliant structures, such as the tension leg platform. Recent, preliminary studies have suggested that, under certain conditions, wind loads are comparable to wave loading at low frequencies, which emphasizes the need to improve our understanding of the wind-loading mechanisms and to quantiiN,r their effects. A limited amount of data from wind measurements taken over the ocean exhibit considerable variability. The problem is essentially due to the difficulty of taking measurements, and it is compounded by the variable nature of the sea surface, which continuously translates and deforms. If the wind flow field and its characteristics are not much influenced by the exact form of the surface, but rather by the energy loss and rate of momentum transfer due to surface friction, then the relationship established for the wind characteristics over land may be applicable over the sea surface. The steady wind loads are expressed in terms of wind velocity and aerodynamic force coefficients. The overall-platform aerodynamic force coefficient is determined by synthesizing force coefficients of the several components and substructures of the platform, based on a projected area approach utilizing code recommended values. Generally, these values are conservative due to complex structural configurations and the influence of interference and shielding. There is a need for better quantification of the interference and shielding effects to develop a procedure for more accurate assessment of steady, aerodynamic load effects on platforms. Wind intensity significantly affects offshore exploration activities, especially the operation of floating rigs and drilling ships. Windstorm information is vital for the planning of drilling operations. The operation of platform cranes, the transfer of personnel, and offshore helicopter flight operation are also affected by wind conditions. Wind forecasts and real-time data are crucial for the smooth operation and safety of these activities. Fatigue Problems Fatigue occurs when a structure fails after a large number of cycles of oscillation or vibration at a stress level well below that which we cause failure after only a small number of cycles. Fatigue is a major problem for bridges and for many other structures or portions of structures. The fluctuating nature of wind causes cyclic loadings on roof and wall panels of buildings. In a siow-mov~ng hurricane, such cyclic loadings can cause fatigue failures of these panels, resulting in their removal and consequent wind and water damage to the building interior. For example, widespread damage to residential units in Darwin, Australia, during Cyclone Tracy in December 1974 was attributed to fatigue failure in roof panels. In addition, damage in the Caribbean during Hurricanes Gilbert (September 1988) and

OCR for page 55
Wnd-Eng~neenng Research Needs 71 Hugo (September 1989) suggests that metal roof panels in these areas failed due to the fatigue of materials. In the design and field operation of a wind turbine, fatigue failure of the rotor has been the dominant concern. The rotor can be subjected to as many as ~ x 108 c~vcles of ultrahigh stress level during its life span. Fatigue failure of many engineering materials has been studied through C-YCiiC load testing ~ which the load changes from positive to negative within each cycle, or from zero to positive within each cycle. Wind loading, however, involves cyclic loading in the presence of a significant mean load that can change sign from stolen to storm. It is not clear that the fatigue load models developed without the presence of a mean load can satisfactonlY Predict fatigue loading due to wind. Further research in load history and cyclic loading should be pursued to improve understanding of fatigue problem in windstorms. Future use of higher-strength materials, entailing lighter components, is likely to increase fatigue-related problems, thus making such research even more critical. CODES AND STANDARDS Building codes in each locality control design and construction of buildings and structures in that locality. Most communities in the United States adopt, in large part, one of three mode} building codes, namely, the National Building Code of the Building Officials and Code Administrators International, the Standard Building Code, or the Uniform Building Code. Wind-Ioad provisions in these mode} building codes are patterned after the ASCE Standard on Minimum Design L,oads for Buildings and Other Structures, ANSI/ASCE 7-~. ANSI/ASCE 7-~S is the only consensus w~nd-Ioad standard currently available in the United States. All three mode] codes utilize the basic w~nd- speed map of the ANSI/ASCE 7-~. However, the si~rnIar~cy in w~nd-Ioad provisions between model building codes and the ANSI/ASCE 7-~S stops with the w~nd-speed map. The factors that influence the magnitude of And loads on a building, addition to wind speed, are the terrain surrounding the building, the shape of the building, arid the desired safety of the building frame and components. The mode} building codes use some of these factors from the ANSI/ASCE 7-~8, modify some factors based on experience, or ignore some of the factors as a part of tradition. In addition, some factors in mode] building codes are adopted from industry manuals. The use, adoption, and modification of w~nd-Ioad factors by the mode} building codes result from an attempt to simplify the w~nd-Ioad provisions, but also represent the lobbying efforts of industries and special interest groups. Even with these modifications, final wind Toads for most buildings are fairly consistent in all mode} building codes, though anomalies exist and, in some cases, the wind loads between mode! building codes differ by 50 percent or more. All three mode] building codes provide the use of ANSI/ASCE 7-~8

OCR for page 55
72 Wnd and the Built Environment as art alternative to be applied at the discretion of the designer. The nation would ~ ^^~^ ~~ ~ 1 . ~ 1 . ~ . . . . . a;; Immensely or one w~na-ioac provisions in the mode! building cones were the same as those found in the ANSI/ASCE 7-~S standard. The current version of the ANSI/ASCE standard, which was crafted by a volunteer group, represents an outstanding effort but does not represent the best that could be produced by a funded development effort. Needed improvements to the ANSI/ASCE 7-~8 include better wind-speed definition; improved wind directionality; improved gust factor models; inclusion of torsional wind loads (twisting about the vertical axis of buildings); improved local pressure prediction and overall frame loads through a fancily of loading coefficients for various building shapes; improved alongwind dynamic loading prediction model; inclusion of a workable across-wind mode] of wind loading; w~nd-Ioad prediction for structures during construction; fatigue-Ioading prediction; inclusion of a risk-based design procedure; and improved standard construction details. Improvements in code provisions could save billions of dollars each year in reduced construction costs and windstorm damage. This benefit can be realized only through an extensive research program directed specifically at the various needed improvements. RETROFIT REROOFING with , . . ~ . . Retrofit, as used in this report. refers to the rovf~rina of an "Y;C:~;~= -~' ~ _ ~ ~ . ~ , . _,< _ _ _~ ~ v ~ ~^_ _V ~A1Aj5 AL All ~1;~ ~111~ 1 ~1 llCW row malenals anchor structure to Drovide increa.cer1 on_~^ ~~ ~ ~ ~ ~ ~~ ~~!LllsIlLIlO~s us Mu Prove uralnage. In lYdd, approximately 250,000,000 sq ft of metal retrofit roofing systems were installed in the United States. This market is growing at the rate of about 15 to 20 percent per year. In its simplest form' retrofit reroofing is the adding of another roof over an existing roof, in the same shape and form as the existing one. In most instances, however, retrofit reroofing systems consist of building a structure with a new roof elevated above an existing roof. Retrofit substructures are designed for both function and economy. The normal function is to provide slope and direction to the new roof for proper drainage. The economical challenge is to design lightweight members while minimizing the number of pieces and the impact of additional wind loads on the existing structure. In the northern part of the United States, the primary loading for reroofing systems is snow load. In the southern part, the lightweight reroofing systems are extremely sensitive to wind. The new geometric configuration of the roofing system may affect the wind load that is imposed on the system. For instance, if the onginal system was flat and the new system had a ridge line, additional loads would be imposed in the area of the ridge. The existing structure must accommodate these increased wind loads. The accurate assessment of wind loads is a critical factor to be considered in most reroofing systems. At present, however, the installation of new reroofs receives little engineering attention.

OCR for page 55
Mnd-Engyneenng Research Needs 73 AI)DITIONAL RESEARCH TOPICS In addition to the research needs described above relating to the impact of severe wind forces on the built environment, a number of other research topics deserve the attention of the w~nd-engineering community. Some of these have been noticed only recently by the public because of the demonstrated adverse impacts on the human environment of occurrences such as of} spills and soil erosion. Some of them, such as wind flow issues in the urban environment, pollutant dispersion, and ventilation, are highlighted because of the demand for more comfortable and healthier living. The development of new technology and the use of new materials also offer the opportunity to reexamine the use of wind as an alternative ener~v source. the research needs of which are described briefly in this section. Bind Flow in the Urban Environment O. , Acceptability for human comfort of any development includes both the aesthetic values of the space and the physical environment designed for project occupants. A plaza intended for relaxation watt not be used if the space is noisy, dirty, chilly, windy, or dangerous in some way. One of the major factors influencing the intended use of pedestrian areas is the physical comfort of the area, including wind forces on, and thermal comfort of, the individual. Wind in a pedestrian area can cause the space to be underutilized, especially if the temperature is low. A pedestrian's thermal balance is influenced by air temperature, humidity, wind speed, the presence of sunlight or shade, and the amount of clothing worn. To date, little research has been performed to guide the development of better models for pedestrian acceptability in the presence of wind, temperature, humidity, etc. Several cities, including Boston, San Francisco, and Pittsburgh, require that wind tunnel tests be performed for all projects in which the wind speeds are likely to cause pedestnan comfort problems. However, even where cities have w~nd-speed requirements, no attempts have been made to account for thermal comfort. Most cities have no requirements at all for wind comfort and leave decisions on pedestrian acceptability to developers. Since developers have an uneven interest in the ultimate quality of the developed space, quite different acceptability criteria are frequently applied to similar projects in the same city. Implementation of accurate and economical prediction methods for pedestrian environments could lead to improved productivity of new and existing projects. Research is thus needed to provide realistic guidelines for human comfort. Pollutant Dispersion Wind eng~neenug has made major contributions to the understanding and treatment of air pollution problems. However, many challenges remain.

OCR for page 55
74 Hued arid the Built Environment Atmospheric dispersion problems have historically been addressed with emp~ncal equations based on a Gaussian dispersion model. These equations work reasonably wed In situations that do not involve terrain or large, nearby structures. On instances where complicated geometries come into play, the models may be in error by factors of 10, 100, or more. Dispersion problems can be divided into two categories: extraordinary events that occur rarely, and regular or daily events. An example of the former would be industrial spills occurring infrequently but with immediate threat to human life. Reingestion of fume hood or laboratory exhausts that could induce long-term health effects through In,w-]eve~1 ren~t~1 -~rnr~cilr~ would be an examD]e of the. latter ~ < _ _ ~ _ _ _~ vim ~ ~ ~ Emergengy response to toxic spills is now often based on little or no advance planning. In industrial settings, likely spill locations and magnitudes can often be anticipated and planned for. Wind tunnel studies can be performed ahead of time and the results stored on a Commuter fr)r olli~lr reference. The computer can rid _ then be connected to meteorological instrumentation to develop a real-time prediction of toxic cloud extent. Although some simple systems of this type do exist, additional research is needed to optimize the design. Laboratory or fume hood exhausts from buildings are frequently reingested into the air intake system, thereby exposing occupants to dangerous levels of chemicals on a regular basis. Wind tunnel modeling of these cases can readily identifier solutions. However, research is needed in both numerical and wind tunnel modeling to develop methods of prediction that are more economical than current wind tunnel testing methods. New pollution sources are required to show that national, ambient air quality standards are being met using modeling procedures approved by the Environmental Protection Agengy. New air quality regulations will require a demonstration that health and safety thresholds are not exceeded or that the cancer risk is insignificant. However, present Gaussian-Wpe or other empincal models are inadequate to describe pollution dispersion in a number of situations, and wind tunnel or numerical modeling is required to adequately address these situations, which include dispersion in winds about buildings, complex terrain, nonuniform roughness (urban or industrial settings), blowing dust or particulates, area or volume sources, mobile sources, mountain valley wind systems (thermally driven flows), and land or sea breezes. Although wind tunnel modeling of pollution dispersal is relatively well developed and numerical modeling is developing, there are a number of research areas in which work wall enhance our ability to quantitatively evaluate pollution levels. These include Reynolds number effects, dense gas effects, plume buoyancy effects, dispersion in stable or unstable atmospheric stratification, and hybrid modeling in which physical and numerical modeling are combined.

OCR for page 55
~nd-Engineen~g Research Needs 75 Ventilation Increasing energy efficiency has become a major goal in the United States. Concern about the increase of carbon dioxide production and acid rain effects from the burning of fossil fuels has led to the consensus that the nation's per capita energy usage must drop. Wind engineering is one of the bases of ventilation technology, since Grind is the driving force for infiltration of air through building skins and also causes direct heat transfer from the building exterior surface. Improved models of inf;Itration and heat transfer can thus improve energy efficiency in buildings. Use of wind speeds to control natural ventilation could have significant energy benefits. Band Energy Nationwide, more than 10,000 wind turbines generating about 700 MW of electrical capacity were in operation by the end of 1984. According to the Office of Technology Assessment (1985), wind energy as an alternative energy source could provide 2l,000-MW capacity, representing about 10 percent of electric system installations. The trend in research and development has been toward larger rotors to achieve economies of scale. A wind turbine with a capacity of 4 MW requires a rotor diameter approaching 400 ft. In wind energy systems of all size ranges, the rotor is the part most vulnerable to structural failure. The rotor is stressed by a variety of forces: gravity-induced stress reversals, centrifugal forces, w~nd-induced thrusts, and wind turbulence. Of these forces, the last two are directly related to the properties of wind. As mentioned above, fluctuating wind gusts can cause fatigue problems. Research in time and spatial variations of wind gusts and their dynamic effect on turbine rotors should be pursued vigorously to understand cyclic loading on the rotor structure and to mitigate fatigue failures. Soil Erosion Soil erosion by wind is a global problem that induces both on-site and off-site damage. The on-site damage includes sandblasting of plants, exposure of plant roots, loss of plant nutrients, and Toss of agricultural productivity. The off-site damage can be in varied forms: air and water pollution, sand deposition on highways, dust damage to households, automobile damage, and landscape damage. Off-site wind erosion costs in the western United States are estimated at between $3.8 billion and $12 billion per year (Piper, 1988~. These costs are much higher than the on-site costs. For example, in New Mexico, on-site costs of wind erosion are estimated to be $10 million annually (Dens and Condra, 1989), whereas off-site costs are about $466 Bullion per year (Huszar and Piper, 1986~.

OCR for page 55
76 Wnd and He Built Environment Reducing dam age due to wind erosion wall require improvements in the technology of wind erosion prediction as well as erosion control. Near-ground wind characteristics, dust transport models, the erodibility of soil, and instrumentation of a soil sampler are all fruitful] topics for research in mitigating wind erosion. Modeling of Iceberg Movement and Oil Spills Wind plays an important role in iceberg drift. The potential threat of an iceberg Impact with an offshore instalIation is of great concern in some northern, onshore of] fields. The overall ice management task involves projecting iceberg position in view of the environmental conditions determined from weather forecasts. Winds and surface currents are also important driving forces for the movement of of! spills. Forecasts of actual trajectories of of] slicks provide essential input to spill mon~tonng and control activities. Research to refine models for both iceberg and of! slick movement would thus directly aid in addressing the environmental and safety risks associated with the production and transport of oil. RECOMMENDATIONS A number of observations, conclusions, and recommendations can be drawn from the discussion presented in this chapter: I. A strong research effort is needed to better define wind loads in a codified fob for small projects, such as housing for single or multiple families, small commercial structures, or industrial buildings, which comprise the overwhelming bulk of construction in the United States. For these types of structures, wind tunnel tests are often too costly to consider; thus reliance must be placed on building codes to define wind loads. Because of the high cost of full-scale tests, wind tunnel data obtained on a research basis have been used to develop the w~nd-loading provisions of current codes. Unfortunately, funds have been insufficient to attain the level of understanding of wind loads that would provide maximum economy of construction. No private organizations have the resources or mission to fund the level of research needed for better code development. Neither the National Science Foundation nor any other federal agency has found sufficient funds for this task. 2. The availability of inexpensive but powerful computers will make possible significant progress in numerical calculation capabilities in the near future. Still needed are improved numerical algorithms, development of turbulence modeling (especially considering the temperature gradients and protocols for easy implementation of parallel processing), and improved graphical display software.

OCR for page 55
Wnd-Er~gyneenng Research Needs 77 3. It is recognized that full-scale field experiments are costly. Nevertheless, field measurements can provide significant, future benefits. It is recommended that a few, long-term field experiments be conducted to yield baseline data, provided that a National Wind Science and Engineering Program can be established to secure sustained funding for the eypenments. In tandem with these experiments, physical modeling in the wind tunnel and numerical modeling should be planned to develop improved modeling technologies. 4. A combination of laboratory measurements, numerical simulation, and full-scale testing represents the quickest path to a more complete understanding of tornado flows. Larger laboratory facilities than currently exist would benefit this investigation. Additional improvements In radar or other instn~ments capable of measuring tornado velocity fields are also needed. 5. Closely spaced w~nd-obse~v~ng stations along the coastline are needed to obtain better definition of hurricane winds. One option is to develop a deployable w~nd-recording station that can be placed in the path of an approaching hurricane. 6. Additional research into the phenomenon of w~ndflow around buildings should be conducted to improve the empincal prediction of cladding pressures. 7. New developments in chaos theory provide possible avenues for understanding the nature of turbulence and wind loading in the presence of turbulence. An analytical component of w~nd-Ioad research emphasizing the chaos theory might provide advances in understanding the complex wind- load mechanisms. 8. A major research effort is required to improve knowledge of dynamic loading effects on flexible structures, so that relatively simple procedures can be employed by a designer to account for dynamic load effects. 9. Active control devices may be effective in controlling w~nd-induced motions of structures. However, considerable research on these devices is required to ensure that they perform as designed after construction. 10. Sloshing of fluid in a tank could be used for damping the motion ~ . ~ . ~ . . . . . to be demonstrated of buildings and bridges, but its effective use wait have through further research. Il. Additional research is needed to better define the levels of motion that are acceptable for various uses in flexible structures. Findings from this research might substantially reduce the construction costs of some structures. 12. Aerodynamic force coefficients used for offshore platform design are conservative in general due to complex, structural configurations and the influence of interference and shielding. Better quantification of the interference and shielding effects to develop a procedure for more accurate assessment of steady, aerodynamic load effects on these platforms is strongly needed. The motion of compliant offshore structures subject to strong winds needs to be investigated by both computational and experimental methods to better understand and quantify the dynamic effect of wind. , ., ~ . .

OCR for page 55
78 Wind Id He Buill Environment 13. Further research in load history and Cyclic loading should be pursued to improve understanding of the fatigue problem in windstorms. Use of higher-strength materials, entailing lighter components, is likely to increase fatigue-related problems, thus making such research even more critical. 14. Most cities have no requirements for wind comfort and leave decisions on pedestrian acceptability to developers. As a result, quite different acceptability criteria are frequently applied to similar projects in the same city. Research in this area is needed to provide realistic and uniform guidelines for human comfort. 15. Research is needed to develop better numerical models and more economical wind tunnel testing methods for prediction of fume hood exhausts to ensure that the public is not endangered by the reingestion of exhausts into air intake systems. 16. Fluctuating wind gusts can cause fatigue problems on rotors of wind turbines. Research in time and spatial variations of wind gusts and their dynamic effect on turbine rotors should be pursued vigorously to understand cyclic loading on the rotor structure and to mitigate fatigue failures. 17. To improve the technology of wind erosion prediction and control, research is needed on near-ground wind characteristics, instrumentation of a soil sampler, dust transport models, and erodibiTicy of soil.