Advances in Hurricane Engineering: Learning from Our Past

The directionality factor (Kd) used in the ASCE 7 wind load provisions for components and cladding is a load reduction factor intended to take into account the less than 100% probability that the design event wind direction aligns with the worst case building aerodynamics. There is concern that in hurricane prone regions the current value of Kd underestimates the real vulnerability of the building system. A two-part study is underway to evaluate the efficacy of the current incarnation of the Kd factor for hurricane-prone regions. This paper presents part one of the study, consisting of a review of the initial development of the Kd concept, as well as a comparison of how wind directionality is addressed in building standards from around the world.

Atlantic basin hurricane activity has been shown to be correlated with the Atlantic Multidecadal Oscillation (AMO). In positive phases of the AMO (warm Atlantic sea surface temperatures, SST), hurricane activity is higher than evident in the long-term average (both warm and cool cycles). In this paper we examine the correlation between the Atlantic SST anomaly and rates and intensities of US landfalling hurricanes. The ASCE 7-10 (ASCE, 2010) wind speeds were computed using tropical cyclone statistics based on the long term average. We are currently in a warm SST cycle, and landfall rates of intense hurricanes are higher in the warm SST cycle. We show the impact of these higher landfall rates on the design wind speeds given in ASCE 7-10. The effect of the SST anomaly on the design wind speeds is treated by applying weights to the hurricane probabilities assigned to each tropical cyclone used in the development of the current ASCE 7-10 design wind speed maps. The weights are based on information on landfall rates according to the storm intensity and landfall location. The net result of the process is a new set of wind speed hazard curves that are valid in a warm Atlantic SST cycle. The climate conditioned hazard curves are compared to those used to develop the ASCE 7-10 wind speed maps at ten points along the Gulf and Atlantic coasts.

Full-scale field tests for a billboard sign with dimensions of 24'-7" wide x 12'-3 1/2" tall x 5'-8 3/4" deep (7.49 m x 3.75 m x 1.75 m) is conducted at the Wind Science and Engineering (WISE) field site located at the Reese Technology Center in Lubbock, Texas. The sign is supported on one pole with a clearance from the bottom of the sign to the ground of 12'-31/2 (3.75m). Gust factors for the determination of pseudo-steady coefficients, as well as force and torsion coefficients for the sign are presented. Gust factors computed from field data show significant scatter. Force coefficients for the sign are typically less than those reported from wind tunnel studies.

The pre-standard for evaluating flood damage-resistant materials was developed to support and clarify NFIP Technical Bulletin (TB) 2 (FEMA 2008). The document addresses structural and aesthetic issues associated with flood damage to building materials but leaves a number of questions unanswered. In an effort to move forward with the pre-standard and address the unanswered questions, FEMA is conducting additional research.

When an existing building in a mapped special flood hazard area is substantially improved or repaired after sustaining substantial damage, and if the local jurisdiction participates in the National Flood Insurance Program (NFIP), the building must be brought into compliance with the NFIP requirements for new buildings. Those requirements are also found in current model building codes. In 2010, the Federal Emergency Management Agency (FEMA) published the Substantial Improvement/Substantial Damage Desk Reference to provide guidance to local officials who are responsible for enforcing local floodplain management regulations and building codes. An important part of that responsibility is making SI/SD determinations. Also in 2010, FEMA released the Substantial Damage Estimator (SDE), a software tool that helps local officials develop information to make SI/SD determinations. The SDE offers a formalized approach to collecting data in the field to develop reasonable estimates of building values and estimates of costs of repairs.

Design standards for buildings and structures in flood hazard areas were published for the time in 1998 in the first edition of the American Society of Civil Engineers (ASCE) standard Flood Resistant Design and Construction (ASCE 24). ASCE 24-98 was revised in 2005 (ASCE 24-05) and is currently being revised again. The release of the next edition is anticipated in late 2012 or early 2013. Since 2000, ASCE 24 has been referenced in model building codes, which include the 2000-2012 editions of the International Code Series (I-Codes) and the 2003-2012 editions of the National Fire Protection Association's NFPA 5000 Building Construction and Safety Code. National Flood Insurance Program (NFIP) regulations, which are largely performance standards, are the minimum requirements that must be enforced by communities that participate in the program. The I-Codes and NFPA 5000 contain provisions that are consistent with NFIP regulations, in large part by references to ASCE 24 and ASCE 7, Minimum Design Loads for Buildings and Other Structures. Communities can satisfy the NFIP requirements for participation by adopting building codes and/or floodplain management regulations. This paper presents (1) an introduction to ASCE 24; (2) a description of the relationship between the NFIP, the I-Codes and NFPA 5000, and ASCE 24; (3) milestones in the development of ASCE 24 (and ASCE 7); (4) the ASCE 24 provisions that are more specific than the NFIP minimum requirements, and those that exceed the minimum NFIP requirements; and (5) a list of the more significant changes between ASCE 24-98 and ASCE 24-05, and a list of possible revisions under consideration for in the next edition.

When Hurricane Andrew hit southeastern Florida in 1992, the model building codes in use at the time had some flood provisions, but none had complete, mandatory provisions that were consistent with the minimum requirements of the National Flood Insurance Program (NFIP). One of the objectives of the NFIP is to reduce damage to buildings caused by flooding. The Federal Emergency Management Agency (FEMA) has long recognized the advantages of enforcing floodplain management regulations through building codes. In 1994, three national model building code organizations established the International Code Council to develop a single set of comprehensive model codes, known as the International Code Series (I-Codes). FEMA, the federal agency that administers the NFIP, participated in incorporating the flood provisions into the I-Codes and continues to participate in the code development process. The flood provisions in the 2009 and 2012 I-Codes are consistent with the NFIP requirements for buildings and structures. Early editions of the Florida Building Code (FBC) that were based on the I-Codes did not contain flood provisions, but the 2010 FBC retains the flood provisions from the I-Codes and has Florida-specific amendments related to flood-resistant construction. CodeMaster, published by the Structures & Codes Institute, is a series of guides on topics related to building codes. CodeMaster: Flood Resistant Design (2011) focuses on the flood provisions in the 2009 and 2012 editions of the International Building Code and the International Residential Code and the referenced standard for flood loads in ASCE 7-10, Minimum Design Loads for Buildings and Other Structures (ASCE, 2005b; 2010) published by the American Society of Civil Engineers (ASCE) and the referenced standard for design and construction requirements in flood hazard areas, ASCE 24 (ASCE, 2005a).

The purpose of this paper is to discuss methods for determining wind loads on buildings and other structures that warrant comment, correction or improvement. The assessment is intended to serve as a resource as a new version of the American Society of Civil Engineers ASCE-7 Standard is being prepared. Issues discussed in the paper include: wind speeds in non-hurricane regions; alternative analytical methods for determining wind loads and wind effects on Main Wind Force Resisting Systems and Components/Cladding; aerodynamic pressure coefficients; pressures on rooftop equipment; component and cladding pressures on arched roofs; and the wind tunnel procedure.

Hurricanes Katrina and Ike, and to a lesser extent Rita, Ivan and others that have struck the Gulf and Atlantic coasts over the last several years, have caused extensive damage to property. Much of the damage was to the foundation systems or was propagated to the building through poorly performing foundations. From damage investigations after the storms, it almost appears that there is insufficient information for the practicing engineers about how foundations should be designed for these highly vulnerable locations and buildings or the information that is available is being ignored. While the storm surge elevations were extensive for both Hurricanes Katrina and Ike, the performance of building foundations suggests that there is either little design being performed for these foundations or the expected loads on the foundations are not clearly understood by designers. Hurricane Ivan caused damage to shallow building foundations where survival was probably assured had the foundations been deep-driven piles. Many of the building and foundation losses were caused by extensive erosion and scour at the foundation and it is clear this important coastal ‘hazard' is not being sufficiently accounted for in the design of the foundation. This paper will explore not only some of the foundation damage mechanisms that have occurred along the U.S. coastlines but suggest design approaches that engineers should consider using to help insure greater survival rates of the foundation. The current foundation design guidance that is available for the practitioner from expert coastal construction sources is not sufficient in covering the many important aspects of foundation design that must be considered to resist failing under the extreme loads of coastal storm surge events.

Hurricane Ike was the third costliest hurricane to hit the United States, causing almost $30 billion in damage after making landfall on the Bolivar Peninsula, Texas, in September, 2008. Following the disaster, a case study of the peninsula was conducted to evaluate hurricane-generated wave and water level effects on wood framed, pile-elevated coastal residences. This study provided information about the structural characteristics most influential to the survival of these homes and may be used to validate future wave force prediction and fragility models for elevated coastal residences. A total of 1922 homes on the Bolivar Peninsula were sampled, and hindcasts of significant wave heights and water levels during the storm were run to relate environmental factors and structural attributes to a home's probability of survival. Freeboard, calculated as the difference between the elevation of the lowest horizontal structural member and the combined wave crest and storm surge elevation, was shown to play a critical role in determining survival: survey data indicated that a positive freeboard resulted in a near 100% survival rate for homes subjected to significant wave heights above 3.5 ft (1.07 m). Survival was also strongly influenced by maximum significant wave height itself; below significant wave heights of approximately 3.5 ft, the freeboard requirement for survival dropped sharply. Uncertainty of numerical hindcasts of significant wave heights was considered when evaluating survival as a function of wave height; however, these estimations were considered more accurate than assuming depth-limited breaking waves at each house location. Homes experienced much larger survival rates in areas subject to small significant wave height conditions than in locations affected by larger waves. This analysis was the first step in accurately determining expected damage resulting from specific wave and water level loadings. A more generalized analysis will mitigate possible effects of construction techniques, survey bias, debris impact, and environment-specific wave height estimations on quantitative survival criteria.

The third edition of Federal Emergency Management Agency (FEMA) P-55, Coastal Construction Manual: Principles and Practices of Planning, Siting, Designing, Constructing, and Maintaining Residential Buildings in Coastal Areas (CCM), was published in 2000. The third edition is based in part on lessons learned from Hurricane Andrew (1992) and other hurricanes in the 1990s. The manual was updated and reorganized for the fourth edition, published in 2011. The fourth edition is based in large part on lessons learned from Hurricanes Charley (2004), Ivan (2004), Katrina (2005), and Ike (2008). This paper summarizes the wind-related updates in the fourth edition of the CCM, including best practices for the enhanced performance of the main wind force resisting system and building envelope in residential buildings in coastal areas.

The fourth edition of FEMA's Coastal Construction Manual, published in 2011, provides guidance and current best practices on the design and construction of disaster-resistant residential structures in coastal areas. The fourth edition is an update of the third edition, published in 2000, and is based largely on lessons learned from Hurricanes Charley (2004), Ivan (2004), Katrina (2005) and Ike (2008). The information related to flood design that has been revised or added includes: Descriptions of recent coastal flood events; Updated information on flood hazard identification; Expanded discussion of sea and lake level rise; Updated references to codes and standards; New guidance on calculating wave slam on elevated floor systems; Improved calculations for flood-borne debris impact loads; Improved guidance on foundation scour; Updated and expanded guidance on flood load combinations

An important question in the wake of Hurricane Katrina (and other storms) was whether building damage was due to "wind" or "water." The question was particularly technically challenging in those cases when nothing but the building's foundation remained after the storm. Professional opinions and conclusions concerning the cause of damage to many specific, individual residences and commercial buildings, extending from the west side of Mississippi to western Alabama, were developed using a detailed, site-specific, forensic analysis process. The analysis included state-of-the-art estimates of the time-history of storm surge, currents, wave heights, and winds at each specific building location during Hurricane Katrina. These, in combination with the known building characteristics and FEMA/ASCE guidance as well other published results, usually led to a fairly clear picture of the timeline of loads on the building and the likely cause of damage. The proper interpretation of hindcast site-specific model results provides a detailed timeline of all the elements of a storm's impacts on a specific building during passage of a storm and can answer the question of wind or water.

ASCE is in the process of preparing a publication that will assist engineers is distinguishing damage during hurricanes that is caused by wind versus flood. The publication is tentatively titled "Guidelines for Hurricane Engineering Damage Assessments." The work is a joint effort between ASCE.s Technical Council on Forensic Engineering (TCFE) and ASCE.s Technical Council on Wind Engineering (TCWE). Chapter authors are all recognized experts in their fields This paper describes the need for a publication that will serve as a resource, and describes the purpose and organization of the publication.

In addition to its massive physical impacts, Hurricane Katrina (August 23-30, 2012) engendered many disputes between policy holders and insurance companies over the agent and timing of damages to covered structures because most of the impacted properties had separate policies - each with exclusions - covering wind and flood damage. Thousands of lawsuits involving millions of dollars hinged on what became known as the "wind vs. flood" question. We resolved this question in an objective, unambiguous way through hindcasting of the storm conditions at high resolution and calculation of the structure's response to the combined wind and hydrodynamic loads. Results for a representative property are presented to illustrate the procedures and results. Discussion covers subsequent evolution of the legal landscape with respect to the exclusion clause and dual causation in insurance policies. Recent court rulings appear to place the burden on the insurer to prove not only that excluded damage was, in fact, the result of the excluded cause but also that the damage from an excluded cause preceded damage from a covered cause.

Engineering and architectural professionals play an important role during the forensic assessment of hurricane damage. These professionals are often asked by owners and insurance companies to assess the cause(s) and extents of building system failures following a hurricane event. The opinions provided during these assessments often are translated into an estimate of damages or an insurance claim/settlement. It is, therefore, critical that these professionals remain impartial and are guided by sound engineering principles during the investigation and assessment efforts. Too often design professionals or self-proclaimed experts with limited resources and limited forensic experience are asked to perform these assessments. This can and has resulted in expert opinions that neglect engineering principles and incorporate junk science. Furthermore, these expert opinions are sometimes based solely on interpretation of codes and standards intended for new construction, instead of considering the actual in-situ performance of the existing architectural and structural elements. In some instances, such experts also utilize junk science to substantiate their opinions, becoming an advocate for their client(s) instead of remaining an impartial and independent expert. This paper reviews building assessments completed by unqualified professionals and identifies instances where established engineering principles were simply ignored.

This paper evaluates the cost-effectiveness of wind retrofit measures. As a result of incidents such as Hurricane Andrew, the model building codes have opening protection requirements for new construction, and some states offer insurance credits for wind retrofit measures. The principal intent is to demonstrate that at certain locations and under specific conditions, wind retrofit projects are cost-effective. The evaluation focuses on the wind retrofit measures outlined in the Federal Emergency Management Agency's (FEMA) Publication P-804, Wind Retrofit Guide for Residential Buildings, dated December 2010. The purposes of the evaluation are to: Support the effectiveness of opening protection requirements that have been adopted as a result of incidents such as Hurricane Andrew; Encourage property owners to consider wind retrofit projects; Potentially facilitate expediting the grant review process to accomplish more wind retrofit projects under FEMA's Hazard Mitigation Assistance Program or other grant programs; Assist the insurance industry in implementing credit for specific wind retrofit measures

The Samuel L. Byrne House, built in Cambridge, Maryland, in 1876, has withstood a century and a half of direct exposure to high winds, torrential storms, and harsh sunlight. Left largely unaltered over the years, the building is undergoing extensive renovations to mitigate future weather-driven deterioration. This disaster-resilient system will take advantage of the timber frame's unique bracing and structural links to distribute uplift, overturning, and shear forces across a designated load path and into the building's new concrete and brick foundation walls. Many supplementary disaster resilient and sustainable building strategies, including the use of stainless steel hurricane straps, wooden dowels, perimeter roof flashing, custom ridge vents, missile-resistant solar panels, storm windows, operable cedar shutters, a modified cold roof design, and a system to divert storm water from the foundation are being incorporated into the renovation and retrofit. The building envelope is also undergoing critical repairs and alterations to improve energy performance and building sustainability; insulation, air barriers, and reinforcing studs and rafters will be installed throughout the exterior walls and roofs. Operating within the parameters of the Maryland Historical Trust and the Cambridge Historic Preservation Commission, the renovation explores the contextual approaches to disaster resilience and sustainability that have afforded the structure its remarkable longevity.

Lessons learned from building performance in Hurricane Andrew led to immediate improvements in the South Florida Building Code in 1994 and the eventual creation and adoption of the statewide Florida Building Code (FBC) in 2002. Alongside these building code improvements, numerous public and private programs to encourage wind resistant construction were also created. Concurrent with the adoption of the FBC, Florida Statute 627.0629 required all property insurers in Florida to recognize the wind resistive benefits of the FBC in their residential rate filings by February 2003. In effect, this statute requires insurers to provide discounts for homes that are built in accordance with the FBC, as well as to homes that were built prior to the FBC with similar wind-resistive features. The creation and implementation of this insurance incentive program has faced many challenges including: determination of the appropriate discounts by feature or combination of features, inspection protocols for documenting wind resistive features, dealing with intentional and unintentional errors introduced in the inspection process, and ensuring the actuarial soundness of the program. This paper discusses the evolution and current state of the mandated insurance discount program in Florida, its basis in wind engineering, and its implementation via wind mitigation inspections.

In 2010, the Insurance Institute for Business & Home Safety (IBHS) released guidance in the FORTIFIED Home Hurricane Program for retrofitting existing residential structures to resist high winds and wind-driven water intrusion associated with tropical storms and hurricanes. This program utilizes engineering principles and draws on more than 20 years of experience in hurricane damage investigations to develop retrofits designed to reduce losses by improving the performance of vulnerable systems, not just selected components. The program seeks to reduce damage, property losses and displacement of homeowners by making improvements to the roof, reducing water intrusion through attic ventilation systems, strengthening of gable end construction, protecting openings, and strengthening of critical elements of the continuous load path from the roof to the foundation. A key aspect of the program is that it uses an incremental approach to retrofitting, which allows homeowners to strengthen their home in steps where the most common failure points are addressed first. Each incremental step thereafter builds upon the strengthening already completed. The expected performance of the home improves with each step and begins to approach, and for some elements to exceed, that of new homes built to the latest building codes and standards. The program requires an initial inspection and assessment of the home to identify what retrofits, if any, are required to achieve each FORTIFIED Hurricane Designation level: Bronze, Silver or Gold. Once required retrofits have been completed, the program includes verification protocols to ensure retrofits meet established criteria. This paper describes the FORTIFIED Home Hurricane program, the basis for the levels developed, and the expected benefits in reducing hurricane loss vulnerability.

Although building codes have progressed toward requiring design and construction practices that increase building resistance to high winds, many existing homes in hurricane-prone regions were designed and constructed to much less restrictive codes. In 2005, Hurricane Katrina damaged or destroyed 234,230 single-family homes, including manufactured homes, in Mississippi alone. In 2010, the Federal Emergency Management Agency (FEMA) published FEMA P-804, Wind Retrofit Guide for Residential Buildings, to provide much-needed guidance on retrofits that would help protect existing residential structures from hurricane damage. The purpose of a residential wind retrofit project is to reduce the vulnerability of a home to damage from wind and wind-driven rain intrusion during a high-wind event. The three areas of a home that are typically the most vulnerable to failure from high winds are roof and wall coverings, openings, and load path connections. FEMA P-804 provides guidance on three sets (packages) of retrofits that address these vulnerabilities. Retrofit projects may be eligible for federal financial assistance through FEMA's Hazard Mitigation Grant Program and the Pre-Disaster Mitigation Program. The application for funding must include a Benefit-Cost Analysis (BCA). FEMA's BCA software tool can be used to conduct the analysis.

Post-hurricane damage assessments have documented the failure of wall components and connections including the loss of various types of siding materials. In addition, recent U.S. model building/energy code changes are expected to lead to increased use of continuous insulation, particularly foam sheathing attached to the exterior surface of light-frame wall framing, to achieve advanced energy code compliance. A particular need for multi-layer wall systems is the understanding of wind loads on the various layers so that designers and product manufacturers can ensure acceptable building envelope performance of energy efficient wall systems in high-wind events, such as hurricanes. To address the knowledge gaps and practical concerns related to multi-layer wall systems with air-permeable exterior cladding, the Foam Sheathing Committee (FSC) of the American Chemistry Council (ACC), the Vinyl Siding Institute (VSI), the National Association Home Builders Research Center (NAHB RC), State Farm Mutual Automobile Insurance Company, Insurance Institute for Business & Home Safety (IBHS), and others have initiated research into wind loads on layers of multi-layer wall systems. Results presented in this paper include determination of wind pressure loading on layers of exterior wall assemblies that include vinyl siding, continuous insulation (rigid foam sheathing) on the exterior side of the wall assembly or Oriented Strand Board (OSB) exterior sheathing, insulation in the wall cavity and gypsum board interior sheathing. This paper presents results obtained from full-scale wind tunnel tests and compares them with results obtained using dynamic pressure chamber tests. The full-scale wind tunnel tests indicate that porous exterior siding experiences higher wind loads and that the interior gypsum wall board experiences lower loads than previously measured in dynamic pressure chamber tests. Results of the different test methods are compared and implications are discussed.

Recent debates at code hearings and in standards development committees have raised questions regarding the capacity of wall systems with exterior rigid foam sheathing attached directly to wall studs to resist out-of-plane wind pressures. This research assesses the out-of-plane performance of exterior wall systems with rigid foam sheathing under dynamic negative pressure wind loading, with specific focuses on measuring the capacity of various wall systems with rigid foam sheathing and evaluating the pressure equalization effects across the layers of the wall system. Several wall systems with exterior foam sheathing and various levels of interior and exterior finishing were evaluated.

The study of wind-driven rain (WDR) has shown a significant progress in past few decades. The applications of semi-numerical and computational fluid dynamics (CFD) methods have shown major advances to reasonably estimate the amount WDR impinging on building facade. The agreement witnessed between numerical results and actual measurements on existing buildings reinforces the above fact. However, standardized testing methodology for WDR which can produce reliable and repeatable test results is still in its development stage. Buildings' component-wise testing methods for WDR effects prescribed in standards and building codes are limited to a simplified application of water with a uniform and cyclic static pressure. The study presented herein focuses on the representation of WDR and the different parameters involved in simulation of WDR in full and large-scale testing facilities. As to the holistic testing approach, many parameters are involved in determining the amount of rain water deposition on building envelope. These parameters consist of wind field characteristics (wind speed profile and turbulence intensity), rain rate, raindrop size spectrum and its integral parameters (drop number concentration and liquid water content per unit volume of air, mean-weight and volume median diameters etc.) and rain duration. The terminal raindrops velocity also has a direct effect on the calculations of WDR rate and in the determination of the level of impact caused by raindrops on building façade. The paper also discusses the processes involved in hurricane level WDR simulation using a 2-fan WDR generator at Florida International University (FIU). The objective is to develop flow management techniques using a 2-fan prototype system that can be applied to simulate the target parameters at the large-scale 12-fan Wall of Wind hurricane wind and rain simulator. WDR is generated using different type of nozzles arranged in a grid pattern with a controlled discharge rate. Rainfall data (drop size distribution and rain rate) collected during tropical cyclones have been used as target for simulating realistic WDR at the testing facility and preliminary results are presented and discussed.

A study at the Insurance Institute for Business & Home Safety (IBHS) Research Center was conducted to evaluate the potential for water entry into the attic through typical vents and through a sealed and unsealed roof deck under wind-driven rain exposures. A duplex building was constructed where sheathing joints on one-half of the roof deck were sealed and the other half was not sealed prior to installing an asphalt shingle roof covering. Water entry through gable end and soffited eave vents were also evaluated. Drainage panels were installed between the lower chords of the roof trusses. This drainage system allowed for the collected water to be segregated by zones. A target rain deposition rate of 8 inches per hour was used for all tests. Water entry through gable end vents and soffited eaves was evaluated using up to three wind exposure regimes. Water entry tests for the sealed and unsealed roof decks were conducted after the shingles were removed. Water entry through the un-taped roof deck joints exceeded that through the taped deck and the vents. This study demonstrated the value of sealing the roof deck and provided information on the relative importance of water entry through vents compared to the roof.

Statistical analyses of residential insurance claims data collected in the aftermath of Hurricanes Charley (2004), Frances (2004), Jeanne (2004) and Wilma (2005) in Florida and Hurricane Ike (2008) in Texas indicate that a strong relationship exists between the wind-related damage sustained by a single-family dwelling and its floor area. In particular, holding all other factors that can affect the vulnerability of a building to wind-related damage constant, as the floor area of a single-family dwelling increases, the wind damage ratio decreases. This behavior is believed to be largely due to how the size of the highest suction zones created by the wind relative to the total surface area of the building change as the building's footprint grows. It is also believed that this size effect is amplified by the generally better material and construction quality found in homes with larger floor areas. This paper presents the hurricane claims data that indicates the wind damage ratio of a single-family dwelling is inversely related to its floor area and discusses the factors that might influence this relationship.

This paper evaluates the use of closed-cell Spray-Applied Polyurethane Foam (ccSPF) as a structural retrofit of existing wood-framed roofs. ccSPF is used as a thermal insulator in the building envelope and it forms a tenacious bond to most construction materials. As a result ccSPF is used to strengthen the wind resistance of wood roof structures. ccSPF is currently being installed in houses in Florida although the long-term performance of the material in presence of water is unknown. This study seeks to evaluate the effect of water leaks on the moisture buildup and wind uplift capacity of ccSPF-retrofitted wood roof panels. Nine full-scale roof attics were constructed and exposed to long-term environmental conditions and simulated rain in two phases, (five in Phase I and four roofs in Phase II). Numerous leak gaps were cut into the roofing systems simulating effects of long-term water leaks. It was found that moisture buildup occurs in roof sheathing and wood framing members, but it the increased wind uplift capacity of ccSPF-retrofitted roofs was not significantly affected. The high tensile strength of the water-saturated ccSPF to wood bond was confirmed in long-term tests using small-scale samples. While there was a reduction in tensile failure stresses after time due to exposure to moisture, the minimum values were more than xx psi, well above the wind uplift suctions producible by hurricane wind speeds. In the second phase, techniques to mitigate the moisture buildup were investigated. Preliminary results indicate that incorporating underside air vents within a ccSPF-retrofit may reduce the drying time of the roof sheathing. Further, installing a self-adhered waterproofing membrane underlayment on the roof deck substantially reduced the absorption of moisture into the wood roof system. A discussion of future testing is provided.

Part I compares three standards for glazing design:1) 2005 Minimum Design Loads for Buildings and Other Structures, ASCE 7-05; 2) 2010 Minimum Design Loads for Buildings and Other Structures, ASCE 7-10; and 3) 2008 International Code Council / National Storm Shelter Association Standard for the Design and Construction of Storm Shelters, ICC 500-08. Part II examines glazing performance under ASCE 7 and ICC 500-08 requirements. Part III is about glazing protections versus hurricanes and tornadoes. Conclusions of this paper are 1) fenestration products, tested with allowable strength design are similar for ASCE 7-10 or ASCE 7-05, namely Cat 3 hurricane protection, 2) fenestration products, tested with allowable stress designs, may fail at ultimate wind speeds, 3) ASCE 7, both 2005 and 2010, do not protect against tornadoes, 4) glazing for essential facilities, should upgrade essential facility products to meet shelter standards and 5) glazing products can be designed to meet shelter standards.

This paper summarizes the methods and results of three experimental investigations regarding the vulnerability of residential windows and window protection systems to windborne debris impact. The first experimental investigation quantified the momentum threshold required to damage unprotected residential window glazing due to the impact of asphalt roof shingles and wooden dowels (vegetation). The second experimental investigation documented the performance of metal shutters due to the impact of whole concrete roof tiles and 2 x 4 lumber. The third experimental investigation addressed the frangibility of roof tile systems under impact by roof tiles, and the puncture vulnerability of metal hurricane shutters under impact by roof tile fragments. These studies confirm that the use of window protection to mitigate damage from windborne debris is effective. However, properly installed and code approved protection systems are vulnerable to failure in strong hurricane conditions.

This paper will present effective window, door and flashing design guidelines, installation case studies and corresponding results from field testing of several installed windows and doors. Openings create the discontinuities in the building facade that often allow water and air infiltration into the building. Over the last 20 years, window and door design has garnered much attention, which has led to more stringent Building Codes. The revised Building Codes have, purportedly, led to improved design and manufacturing of windows and doors. However, recent field testing of installed windows and doors indicates that, on the majority of projects, inadequate quality control leads to installed products that leak unacceptable levels of air, water or both. Quality control issues at the manufacturing plant can be simple mistakes, such as factory workers installing the wrong material around the perimeter of a sash or frame, which can significantly reduce the performance level of a highly-rated product. Improperly drawn or misinterpreted Shop Drawings often result in the omission of field-applied sealants, again significantly reducing the performance of a highly-rated product. Installation errors reduce the effectiveness of sealants which also significantly reduces the performance of highly-rated window and door products. Fortunately, field testing helps identify the modifications necessary to bring the windows and doors in compliance with Contract and Building Code requirements. Field testing also verifies that modifications have effectively restored the windows and doors to their intended performance levels. Field testing has proven to be an invaluable tool to ensure installed window and door products are functioning as intended, so we recommend field-testing windows, doors and associated flashings for both renovation and new construction projects.

Façades must screen the unwanted from entering the building. This includes wind, rain, windborne debris, solar radiation, noise, and even intruders. Energy efficient glass has become a basic requirement for building facades across the world; but, in areas subject to windborne debris generation during storms, impact resistance may be mandated by code, or at least through design specifications. Glass types, configurations, post breakage performance, and design configurations as they relate to windborne debris resistance and common hazardous glass locations as outlined in regulations are discussed. Configurations to meet desired performance attributes such as safety, hurricane, security, and sound abatement with consideration for energy efficiency are described and commonly accepted products are compared. Post breakage behavior and glass shard containment are also examined.

This paper presents an advanced engineering evaluation using nonlinear analysis of hyperelastic material that provides significant improvement to structural silicone glazing (SSG) design in high performance curtain wall systems. Very high cladding wind pressures required in hurricane zones often result in bulky SSG profile dimensions. Architectural desire for aesthetically slender curtain wall framing sight-lines in combination with a desire to reduce aluminum usage led to optimization of silicone material geometry for better stress distribution. To accomplish accurate simulation of predicted behavior under structural load, robust stress-strain curves of the silicone material are essential. Physical property testing was provided by the silicone manufacturer via specialized laboratory protocol. A series of rigorous curve fit techniques were then made to closely model test data in the finite element computer analysis that accounts for nonlinear strain of hyperelastic silicone. Comparison of this advanced design technique to traditional SSG design hi-lights differences in stress distribution contours in the silicone material. Simplified structural engineering per the traditional SSG design method does not provide accurate forecasting of material and stress optimization as shown in the advanced design. Full scale specimens subject to structural load testing were performed to verify design capacity, not only for high wind pressure values, but also for debris impact per ASTM E1886 and ASTM E1996. Also, construction of the test specimens allowed development of SSG installation techniques necessitated by the unique geometry of the silicone profile. Finally, correlation of physical test results with theoretical simulations is made so evaluation of design confidence is possible. This design technique will introduce significant engineering advancement to the curtain wall industry.

There is no question that a lot has been learned about hazard mitigation design of the building envelope in the last twenty years. The real question is, based on what we have learned and implemented, "Are our Building Envelopes Ready for a Powerful Storm?" The scope of this paper was to answer the above question. This was done by:defining what constitutes a safe and secure building envelope capable of satisfactorily sustaining a powerful storm; identifying the critical areas in which advancement would have been necessary in order to have achieved that goal; posing a series of additional questions in each category designed to identify the necessary key advances; explaining the proper methods and addressing the importance and significance of each category; and, finally, answering each of the additional questions. The categories identified as being critical to having a safe and secure building envelope were: research, design and intricacies of design, testing, product approvals, shop drawings, design review, installation and inspection. It was determined that a safe and secure building envelope does not depend on any one of these critical areas, but rather all of these critical areas. These processes act as a chain with the strength of the overall chain only as strong as its weakest link. The strong links, in which significant advances have been made, were found to be the areas of research, design, testing and product approvals. The weak links, in which very little or insufficient advances have been made, were found to be intricacies of design, shop drawings, design review, installation and inspection. After having studied the answers carefully, in my opinion, our building envelopes are not ready for a powerful storm.

Recently, there has been considerable interest in the assessment of surface-damaged curtain wall glass, especially by the insurance industry in the aftermath of high wind events such as hurricanes. Whether the affected structures are 30 years old or recently completed, at stake can be the potential replacement of hundreds or thousands of insulating glass units with related costs in the millions. Using six case histories, some involving natural disasters and some involving other damage during construction operations, this paper presents information related to the field and laboratory investigation of surface-damaged glass, as well as determination of cause and restoration techniques for certain types of damage. An understanding of the limitations of these restoration techniques is also essential for providing a proper prognosis for the damaged glass. This paper presents valuable information for architects, engineers, contractors, insurance industry representatives, and building owners and managers interested in assessing glass surface damage, determining causation, and evaluating appropriate remedial measures.

Glass damage and failures of cladding components during extreme wind events can be attributed to a variety of different mechanisms, including extreme wind pressures, missile impacts from debris, and structural performance of glass within its support system. Wind speeds in excess of design values, either during the storm or through the channelization of the winds due to the close proximity of other structures, can lead to surface pressures similarly beyond design. However, extensive cladding damage has been observed in regions where wind speeds experienced were known to be below the design values. This research provides insight into the performance of the JP Morgan Chase Tower and the adjacent parking/office structure, in the downtown Houston area in the aftermath of Hurricane Ike. It is shown through both computational fluid dynamics (CFD) and wind tunnel analyses that slight variations in wind direction led to increased development of vortical flow between the high-rise and adjacent structure, as well as increases in the observed pressure loads on the exterior surface. The combined effects of elevated pressures and the existence of vortical flows possibly initiated damage and then propagated further damage from airborne debris. Similar mechanisms of debris damage were witnessed during Hurricane Alicia (1983), which also impacted Houston.

Design for wind loading is of particular importance in Florida due to the influence of hurricanes. The design wind speeds when applied to analytical code provisions result in significant wind pressures compared to other locales. This places a premium on the need for refinement of design wind loads through wind tunnel testing. Accurate measures of pressure and force coefficients, in the presence of the surrounding built environment, account for much of the benefit. Occupant comfort is also an issue, although the treatment of hurricanes can be a source of confusion. Use of lower return periods in the evaluation of comfort should address this. The Florida Building Code contains specific provisions for High Velocity Hurricane Zones that exclude the consideration of Exposure B, which would otherwise be used in many instances. Wind tunnel studies are designed to satisfy the intent of the code with respect to safety while still reflecting realistic conditions for serviceability.

The Shanghai Center is slated to be the tallest building in China, topping out at a height of 632 m. Due to the tower being situated in a typhoon zone, as well as its height and unusual shape, RWDI undertook a comprehensive wind engineering program. The program included an initial, exploratory set of tests which were done to assist the project team in optimizing the tower's shape and orientation with respect to the prevailing winds. Once the shape had been finalized, detailed tests were conducted which included a pressure study, an aeroelastic model study, and a high-speed Reynold's number test because of the tower's rounded corners. Based on the wind engineering studies, Thornton-Tomasetti was then able to develop an efficient design for the tower that consisted of super-columns with outriggers that work in conjunction with the inner building core, comprising an effective system for resisting wind loads.

In the design of tall buildings, wind effects can significantly affect a number of aspects of design; but, there is often little consideration of the type of wind climate in which the building is located. Wind climates governing the design of tall buildings are geographically dependent. In regions where hurricanes dominate the extreme wind climate, it needs to be recognised that these are unrelated to the parent wind climate. At latitudes beyond those where hurricanes occur, synoptic storms can dominate both the everyday and the extreme winds. It is for buildings in these areas, that many of the current criteria have been developed. Some of these assumptions, and the effects of their application to tall buildings in hurricane regions, are discussed in this paper. Issues relating to more appropriate design criteria for application in hurricane regions are also presented.

Solar arrays installed on roofs of low-rise commercial buildings are especially popular since it makes good use of previously unexploited real estate. These systems are usually restrained to resist wind-induced lift-off and sliding using ballast, penetrations, adhesives, or combinations thereof. In the case of ballasted systems, reducing the added weight on the roof is often the primary objective of designers, while maintaining a certain level of reliability. This can be accomplished through aerodynamic design and other methods, but sometimes more simply through the utilization of load sharing between adjacent panels. This is accomplished through the vertical bending stiffness of the racking system. This paper presents a discussion of the varied wind tunnel testing and analysis methods currently employed around the world, with a focus on the correlation effects associated with load sharing. A finite element model, in conjunction with wind tunnel data on a roof-top array, is used to demonstrate the impact of array stiffness on wind loads in a time- and spatially-varying sense. Practical guidance to determine ballast requirements is also provided.

Wind loads on large, low-profile, roof-mounted solar arrays placed on low-rise building roofs were examined using scale models in a boundary layer wind tunnel. The effects of building size and array tilt angle were examined. It was found that wind loads on the array increase with building size. Normalizing the array effective wind area by the building wall size leads to design curves that collapse onto a single curve for each array geometry. For tilt angles below 10°, there is a linear increase in the pressure coefficients as tilt angle is increased. For arrays with tilt angle of 10o and above, the wind loads do not depend on tilt angle and are relatively constant.

Reliable wind loading information on solar panels is essential for the large-scale deployment of solar technologies in the United States. However, only a limited amount of published wind tunnel and computational fluid dynamics (CFD) research is available to address aerodynamic loading information for photovoltaic (PV) panels and solar water heaters. Because of this knowledge gap, no provisions exist in the current building codes and standards for the design of PV mounting systems in areas that are vulnerable to extreme winds. This study presents aerodynamic data based on experimental testing of a residential roof-mounted PV panel. The goal of the study was to develop a test-based dataset of aerodynamic loading effects for the test cases considered. Full-scale testing of a single PV panel mounted on residential building models was conducted with the 6-fan Wall of Wind (WoW) hurricane simulator. During the full-scale tests, multi-axial load cells measured the aerodynamic reaction forces at four locations where the PV racking system was attached to the roof. Net aero-dynamic force coefficients were obtained directly from the reaction forces measured. Model scale testing of similar configurations was conducted in a boundary layer wind tunnel (BLWT) at a scale of 1:12. During the wind tunnel tests, the PV panel model was equipped with 28 pressure taps to measure the overall pressure distribution on the panel. Net aerodynamic force coefficients were determined from the simultaneous wind tunnel pressure time histories measured from upper and lower solar panel surfaces using the pressure integration method. A systematic comparison between the aerodynamic force coefficients derived from full scale and wind tunnel testing is in progress. Noteworthy fluctuations of the glass panel were observed during the full-scale tests for cases when the PV panel was mounted parallel to a flat roof. This phenomenon is seen in the uplift force time histories, indicating PV panels may experience across-wind loading effects under certain mounting configurations. Overall, good agreement was observed between the large- and small-scale measurements for critical PV inclination angles.

This paper describes how modelers of the Florida Public Hurricane Loss Model (FPHLM) account for and implement various roofing characteristics that reflect the historical building practices and code enforcement in Florida. The life cycle of roofing structural elements, particularly the roof cover, are considered by using an algorithm to estimate whether or not a structure has been retrofitted based on the original year built. The paper will discuss the role of building practices and life cycles of roofing components and their use in vulnerability models as well as discuss the influence of roof retrofits on the vulnerabilities of typical personal residential structures.

A wind tunnel investigation of lift-off of concrete roof pavers from a high-rise building using a 1:10 scale model is described. Corner vortices were the key reason for the observed movement of the pavers. It was also found that if Reynolds number effects were ignored the model tests overestimated the lift off velocity that occurred at full scale during a wind storm. However, making a Reynolds number correction to the model results based on published results on behavior of roof suctions under corner vortices brought them into good agreement with full scale. The effectiveness of strapping systems on the two edge rows of pavers and on all pavers was investigated. As well, vertically restraining edge pavers was tested. The full strapping system was found to be significantly more effective than straps on two edge rows only, and further substantial improvement was obtained when edge restraints were added.

This paper presents a large-scale Wall of Wind (WoW) investigation of external and underneath wind pressures on roof tiles installed on a low-rise building. The dimension of the low-rise building tested was 2.74 m (9 ft) long, 2.13 m (7 ft) wide, and 2.13 m (7 ft) high (wall height). The building was tested with interchangeable gable roofs with three different slopes (2:12; 5:12 and 7:12). The field tiles of these gable roofs were tested under three different tile profiles. namely: high (HP), medium (MP), and low profiles (LP) in accordance with Florida Building Code of Practice; but, only the MP profile is discussed in this paper. For the ridge part of the roof, two different types of tiles were considered, namely: rounded and three-sided. The results showed that the ridge tiles were exposed to high pressure coefficients as compared to the field tiles, including those located at the corners. Considerably higher pressure developed on the gable-end ridge tiles compared to those at the roof center. Positive mean underneath pressure on the field tiles develops on the windward side, thus increasing the net pressures on the windward surface of the roof. On the leeward side, however, reductions in net pressure to a non-significant level were observed due to the opposite direction of the internal and external pressures.

This paper presents the initial results and future experiments of an ongoing multi-year research initiative to investigate the performance of asphalt roof shingles exposed to hurricane conditions. Two unique experimental facilities were used to (a) calibrate current wind uplift test requirements for asphalt shingles with expected performance in actual hurricanes, (b) investigate the effects of aging on the wind resistance of shingles, and (c) evaluate the influence of edge attachments and fastener schedules on the wind resistance of shingles. This research was complemented by wind tunnel studies to quantify turbulent wind flow over the surface of a residential roof. These goals represent a necessary baseline to address the wind performance of asphalt shingle roofing systems. The outcomes of the research will provide a quantification of the performance of new and aged shingle roof systems, refine the specific causes of the inadequate performance observed in field studies, and create a roadmap to mitigate damage associated with shingle loss in high wind events. The current test method for wind uplift performance of shingles, ASTM D7158, will be calibrated against the controlled realistic wind loading, bridging the gaps between standards and expectations. This new information will be used to validate or refine the current test methodology and provide the latest engineering knowledge to the roofing industry for the development of reliable, wind-resistant shingle roofing systems for the long-term.

The rising price of fossil fuels, government incentives and growing public awareness for the need to implement sustainable energy supplies has resulted in a large increase in solar panel installations across the country. For many sites the most economical solar panel installation uses existing, southerly facing rooftops. Adding solar panels to an existing roof typically means increased loads that must be borne by the building's structural elements. A successful solar panel installation on an existing rooftop includes documenting the as built condition of the structure and then calculating and applying loads from new solar panels. The structural engineer is called on to design economical structural reinforcement to bring existing buildings into compliance with current building code requirements for wind, snow and seismic loads. The most economical reinforcement design varies and depends on whether the existing roof is built from wood framing, wood trusses, steel framing or pre-engineered structural members.

In the years prior to the full development and implementation of comprehensive building codes, it was common practice for rooftop equipment to be installed without being firmly attached to the building structure. The thinking was that the weight of the roof top unit would be sufficient to keep the roof curb and rooftop unit on top of the building. Recent research into the effects of major natural disasters has shown that in order to resist significant seismic and wind loads, rooftop units must be firmly restrained from movement off of their curbs, and the roof curbs must be positively attached to the building structure. These requirements are changing the way that roof curbs are designed, built and installed.

During a hurricane the objects on the roof of the building can be subjected to wind pressures that exceed the walls of the structure, depending on the shape of the roof. In fact, design of attachments of roof top equipment had been recognized as insufficient in the ASCE7-05 Standard, as an amplification factor of 1.9 was added in Section 6.5.15.1 to the calculation of forces to due wind pressures on roof-top equipment. The new ASCE7-10 Standard has changed the approach to calculating wind forces using "Ultimate" wind speed maps and different force coefficients. In this paper we have performed ASD calculations using the ASCE7-05 Standard and the "Ultimate" wind speeds corresponding to the "strength" method in the ASCE7-10 Standard for some typical Roof Top Unit (RTU) curbs. We compare the results of the subsequent design of the RTU curbs using both Standards and list conclusions. Also included in the paper is a step-by-step procedure to be used when evaluating and/or designing RTU curbs and their attachments to the roof system for hurricane wind loads. Typically, RTU curbs are fabricated from light-gage metal and, therefore, the AISI Code is most often applicable. Diagrams are included showing the various parts of the RTU curb and the corresponding load paths that are evaluated.

Damage surveys performed during recent hurricane seasons have identified roof-top equipment as a primary point of failure on low-rise building envelopes. Rooftop equipment failures caused by high winds have resulted in millions of dollars worth of damage to buildings, primarily from water infiltration into large openings exposed by detached rooftop equipment, wind-borne debris damages caused by blown-off rooftop equipment, and delays in post-storm building re-occupancy due to subsequent mold and mildew infestations attributed to water infiltration. These observations indicate that existing rooftop equipment wind load design methodologies and anchorage techniques may not be adequate to withstand severe wind loads. The objective of this study was to obtain realistic wind loading data for rooftop equipment by conducting full-scale and wind tunnel experiments on typically sized roof-mounted air conditioning (a/c) units. Wind loads on full-scale rooftop a/c units were measured using the 6-fan Wall of Wind (WoW) hurricane simulator at Florida International University (FIU). The full-scale study considered three a/c condenser units placed on the roof of a low-rise building model having a flat slope. The a/c units were instrumented with force and pressure sensors to capture the aerodynamic loading effects. A similar investigation was carried out in a boundary layer wind tunnel (BLWT) at a scale of 1:10. The wind tunnel a/c unit models were instrumented with 32 pressure taps each, and the pressure integration method was used to calculate the net aerodynamic loading effects. Preliminary results indicate that the force and moment coefficients obtained from the full scale and model scale studies are in reasonable agreement when rooftop a/c units are away from the edges of the roof.

A numerical simulation system incorporating an atmosphere-ocean-wave coupled model and a typhoon potential vorticity bogussing scheme was developed to calculate strong winds, heavy rainfall, storm surges and high waves caused by maximum potential typhoons in a future climate. The risks of coupled hazards in Ise Bay, Japan, under a future climate were evaluated using this simulation system. Results demonstrate the possible occurrence of a 6.9-m storm tide at Nagoya Port, which remarkably exceeds the largest storm tide ever recorded in Japan, which is 3.5 m. Moreover, times exceeding design values for the storm tide, significant wave height, and wind speed were shown simultaneously to evaluate the risk of coupled hazards. Results show that Tsu Port is an extremely dangerous area for coupled hazards and that the time exceeding design values was 74 min.

The effects of global climate change on hurricane activities have been the subject of constant debate in recent years. An analysis of the hurricane database (HURDAT) maintained by the National Hurricane Center reveals a change in the annual hurricane (or storm) genesis frequency in the Atlantic basin in recent years. Some studies have projected increases in both the frequency and intensity of landfalling hurricanes along the U.S. coastline in the near future. This paper presents the results of a preliminary study of the influences of possible change in the hurricane genesis frequency on the hurricane landfall rates and wind speed distributions along the U.S. coastline. A state-of-the-art stochastic hurricane simulation model was utilized to simulate and project the future hurricane activities. Instead of generating the number of storms per year based on a constant mean frequency, computed using historical data, two new hurricane genesis models were derived to project the future storm genesis frequency. The stochastic hurricane simulation model incorporating the new genesis models was utilized to generate multiple realizations of hurricane seasons for the next 50 years (2012 to 2161). The results of this preliminary study show appreciable increases in both the landfall rates and wind speeds under two different hurricane climate change scenarios. In particular, the results indicate that more storms are likely to strike the Florida peninsula and coastal regions of Georgia, South Carolina and North Carolina.

A probabilistic framework is presented for evaluation of hurricane wave and surge risk with particular emphasis on real-time, automated estimation for hurricanes approaching landfall. The basis of this framework is the characterization of hurricane scenarios through five model parameters: the landfall location and heading, the central pressure, the forward speed, and the radius of maximum winds. Description of the uncertainty in these parameters, through appropriate probability models, leads then to quantification of hurricane risk through a probabilistic integral. Its assessment relies on the development a surrogate model for the rapid evaluation of hurricane waves, water levels, and runup. This surrogate model is established using a response surface methodology fed by information from hundreds of pre-computed, high-resolution model runs. The surrogate model is able to evaluate the maximum wave height, water level, and runup for any storm at a cost that is more than seven orders of magnitude less than the high fidelity models and thus meet time constraints imposed by emergency managers. An illustrative example is presented that considers applications of these tools for hurricane risk estimation for Oahu. The development of standalone executables to facilitate this risk estimation is also discussed.

In spite of the rapid development of computer power, the real-time application of coupled storm surge simulation with fine resolution is often constrained by needed computational time and resources. In this study, an alternative, efficient and robust technique has been developed to estimate high resolution local storm surges in coastal areas. Recently, this technique has been incorporated into a Disaster Response Intelligent System (DRIS) to support decision-making prior to, during, and following a major disaster. This incorporation makes it possible for decision-makers to respond quickly in dynamically changing circumstances during natural disasters.

The Federal Emergency Management Agency (FEMA) commissioned the Georgia/Northeast Florida Storm Surge Study (GANEFLSSS) as part of the recent effort to update coastal Flood Insurance Rate Maps. The study will simulate hurricane-induced coastal flooding with the tightly-coupled SWAN+ADCIRC model. The SWAN+ADCIRC model inputs meteorological and tidal forcing to calculate depth-averaged currents, water levels, and waves. This paper details model setup and preliminary model validation of the GANEFLSSS. Considerations in model setup included: Establishing optimal resolution based on data availability, computational resources, potential wave propagation, and proximity to coast. Identifying and capturing hydraulically important features that could conduct or impede storm surge flow. Selecting appropriate spatially-variable model parameters, such as Manning's n, wind shielding parameters, and eddy viscosity. The validation effort includes both astronomical tide and five historical storm simulations. The production run storm suite will include several hundred synthetic storms. At writing of this report, validation is currently underway with production runs to follow. Production runs on the 3 million node model mesh will take place on a 528-core high performance computing cluster (HPCC).

This inability to effectively leverage the intellectual and computational resources of the research community devoted to mitigating hurricane impacts means that the latest advances hazard characterization and modeling are often not fully harnessed to support risk assessment and decision making to mitigate the impacts of such events. In response, this paper discusses the formation of a virtual organization supported by scalable cyberinfrastructure to deliver an integrated platform for hazard characterization and system vulnerability assessment. The paper focuses specifically on the development of this portal and its GIS-supported visualization tools, with an overview of its layering approach to impact assessment and eventual acceptance of field data/observations. By developing such digital workspaces that synergize existing datasets, models, simulation tools and risk assessment frameworks of the authors and their collaborators, this project offers the opportunity for integrated assessment of hurricane impacts on civil infrastructure to support not only research and education, but also decision making by stakeholders in coastal regions.

This paper will provide an overview of FEMA's Mitigation Assessment Team (MAT) Program, as well as observations and recommendations coming out of the MAT post-disaster investigations. Since the early 1990s, FEMA has deployed assessment teams in response to Hurricanes Andrew, Iniki, Opal, Fran, Georges, Charley, Ivan, Katrina, and Ike. The MAT has also deployed teams following floods, tornadoes, and terrorist attacks. The most recent MAT deployments were to the Southeast U.S. and Missouri following the tornadoes of Spring 2011. MAT investigators are deployed quickly following disasters, and teams are made up of FEMA representatives, state and local officials, and public and private sector experts in engineering, architecture, construction, building codes, and natural hazards research, as well as industry representatives. Observations from MAT findings help establish improved disaster-resistant construction codes and standards, designs, methods, and materials used for both new construction and post-disaster repair and recovery. The observations and recommendations of MATs are presented in reports published by FEMA. MATs deployed after hurricanes have focused on the performance of residential, commercial, and critical facility construction affected by high-velocity flood flow, wave action, the impact of waterborne debris, erosion and localized scour, and high winds. This paper will focus on the observations of the interaction of flood and wind during hurricanes, as well as the load combinations of flooding and high wind.

Hurricane Irene made landfall near Cape Lookout, NC, on 27 August 2011 as a Category 1 storm and then tracked across North Carolina and re-entered the Atlantic Ocean before making a second landfall near Little Egg Inlet, NJ as a weaker storm. Irene then tracked up the east coast as a tropical storm into New York City and New England. Although not a strong storm in terms of wind speeds or storm surge, Irene impacted over 60 million people and some of the most populated areas in the U.S. This paper presents observations made by reconnaissance teams from North Carolina through Connecticut. The focus is on wind and storm surge damage with the majority of the wind damage occurring in North Carolina and storm surge damage observed mostly in North Carolina and Connecticut. Observations are included for different classes of structures including residential and commercial buildings, hotels, and industrial facilities. Construction practices that exhibited high failure rates are identified and discussed. The possible ramifications of a stronger future storm following a similar track are briefly presented.

The results of damage assessments following recent hurricane events suggest that predictions of wave transformation across barrier islands during overtopping events are unreliable and, even worse, inaccurate. In many cases, inaccurate model predictions of wave heights in high hazard areas are translating into improper standards for "lowest structural member" elevations on residential and commercial structures. The substantive result of which is increased cost of construction, whether initial or replacement costs. The purpose of this research is to improve guidance on building elevations in high hazard flood plain areas where wave action is the primary structural damage mechanism. A number of novel, autonomous wave gauges will be temporarily mounted to fixed structures on Dauphin Island, Alabama, in advance of an overtopping event during future hurricane seasons. Measured wave heights and water levels will provide an opportunity to characterize wave transformation across the barrier island. When combined with surveyed building elevations and a detailed inventory of structural and foundation characteristics, the measured waves and water levels will provide an opportunity to identify critical building elevations that delineate survival and destruction, as well as determine the adequacy of structural connections and foundations.

Two primary needs that underpin societal responses to flooding are timely and reliable forecasts of near-term flood conditions and accurate understanding of long-term flood potential. Computer models have long been used to forecast flood conditions for both riverine and coastal environments. But in recent years, in both riverine and coastal environments, computer models and statistical characterization of synthetic records have become the more common basis for estimating flood potential. More than half of the Federal Emergency Management Agency (FEMA) flood-insurance rate maps (FIRMs for riverine systems are now based on watershed rainfall-runoff models (Will Thomas, written communications, April 2012). The ascendency of computer models for the characterization of coastal flooding has been particularly dramatic. Since the landfall of Hurricane Andrew there has been a great deal of progress in the development of computer programs that can be used to model hurricanes (Jelesnianski and others, 1992; Luettich and Westereink, 2010). Now ensemble simulations featuring a variety of historic or synthetic storm tracks, intensities, and ambient weather and tidal conditions are used increasingly to model hurricane-driven storm surge, estimate flood potential, and serve as a basis for coastal FEMA FIRMs.

Past hurricanes like Ivan, Katrina and Rita indicate the need to evaluate structural performance on both qualitative and quantitative bases. This paper briefly emphasizes the use of a structural reliability-based simulation technique combined with nonlinear pushover analysis to quantify losses. The simplified reliability-based method is introduced and used to estimate the vulnerability curve of a typical fixed platform. The verification of a 4-leg fixed platform is provided by using the damage reports from Hurricanes Katrina, Rita and Ivan in MMS Project Numbers 549, 578. The comparison of the results show the vulnerability curve is able to predict the reported damage incurred from Hurricanes Katrina and Rita within one standard deviation range. However, the damage ratios reported from Hurricane Ivan are out of the range due to the wave-in-deck factor which is not included in the model as well as the lack of correlation between wind and wave states observed. Overall, this study shows that the simplified reliability-based method is capable of predicting losses. The concept and procedure of this approach can rather simply be applied to other types of platforms.

The susceptibility of the transportation network, in particular coastal bridge, along the US Gulf Coast to hurricane damage has been demonstrated in previous hurricane events. Accordingly, it is vital to have an approach for reliability assessment of coastal bridges to evaluate the vulnerability of existing bridges, prospective retrofit measures, or even new designs. Probabilistic assessments are required in order to capture the uncertainties in the performance of bridges stemming from variability in the loading and the natural hazard itself, which include surge and wave parameters, as well as structural design and construction. The conditional probability of failure of a structure under a given hazard condition, known as the fragility, has become a well accepted tool in risk assessment of structures; yet an approach for derivation of the fragility of coastal bridges is required to support risk assessment in these regions. This paper reports on a methodology for construction of fragility surfaces for bridges with high strength connections between super- and substructure subjected to coastal storms. Although these configurations are less common than typical simple supports with minimal connectivity, they do exist among coastal bridge inventories and moreover, such provision for enhanced connectivity has been suggested for retrofit or new design techniques. Construction of these fragility surfaces requires probabilistic estimates of structural capacity and demand. This research provides a consistent methodology for capacity limit state appraisal based on the numerical simulations of bridges under hurricane induced wave and surge loads, and global performance objectives for bridges related to post event loss of stiffness or load carrying capacity. Predictive probabilistic models are developed for the demand parameters identified as primary indicators of capacity loss (e.g. columns axial strain) and the bridge fragility is constructed for each damage state to demonstrate the application of the proposed method.

Major coastal cities, which have large populations and economies, easily suffer from the losses due to hurricane wind and storm surge hazards. Although current design codes consider the joint occurrence of high wind and surge, information on site specific joint distributions of hurricane wind and storm surge along the U.S. Eastern Coast and Gulf of Mexico is still sparse and limited. In this paper, joint probability distributions of combined hurricane wind and storm surge for the City of Charleston, SC is developed. A stochastic hurricane model was used to simulate 5,000 years of synthetic hurricanes. The simulated hurricanes were inputted into the ADCIRC (Advanced Circulation) surge prediction model to compute the surge heights at selected locations. The calculated peak wind speeds and surge heights were employed to generate the joint probability distributions at each location. These joint distributions developed can be used in a multi-hazard design or risk assessment framework to consider the combined effects of hurricane wind and storm surge hazards.

This study estimates tropical cyclone-generated storm surge levels for 2-year, 5-year, 10-year, 25-year, 50-year and 100-year return periods along the U.S. Gulf Coast, using SURGEDAT, a Gulf of Mexico storm surge database, for statistical analysis. Return periods are calculated using the Pareto, Gumbel and Beta-P distributions, and the Huff-Angel and Southern Regional Climate Center (SRCC) linear regression methods. The SRCC method, which fit best with the data, estimated basin-wide surge levels of 8.20 meters for the 100-year return period. The Southeast Louisiana/ Mississippi Zone generated the highest surge levels out of 10 sub-regions in the basin. Surge levels in this zone ranged were 7.67 meters for the 100-year return period. The lowest surge levels in the region were estimated along the West Coast of Florida. These results are useful for emergency management and disaster science personnel, planners, decision makers and the scientific research community.

Tropical cyclones are usually considered wind events and/or coastal surge events. This is reflected in the main tropical cyclone classification systems, such as the Saffir-Simpson Scale, which is based on their wind intensity or central pressure, and in some cases, storm surge height. However, this line of thinking ignores the precipitation-induced flooding hazard related to tropical cyclones. Around the world, there are many examples of historical tropical cyclones that produced damage and losses relating to precipitation-induced flooding. In Central America in 1998, for example, Hurricane Mitch generated noteworthy amounts of rain in Honduras and Nicaragua causing nearly 20,000 flooding-related deaths. Similarly, Tropical Storm Allison (2001), which never reached hurricane intensity, caused more than USD 5 billion in flood-related losses in Houston, Texas. Relatedly, precipitation-related losses are not always covered by insurance companies. In the United States, for example, residential insurance does not cover flooding. Yet, in other regions, such as China, Japan, and Central America, flood coverage is offered under a standard policy or as an extended coverage. To comprehensively analyze hurricane risk both from economic and insurance loss perspectives, it is necessary to consider the tropical cyclone-induced flood peril, along with winds and storm surge. Tropical cyclone flood potential is not well correlated with the storm's wind speeds and thus should be modeled separately to determine accurate flood risk for a region. In this paper, extreme rainfall events from around the world are discussed in regards to the damage and losses. Additionally, similarities among these events are investigated, and the use of CAT (catastrophe) models to estimate economic and insured losses for such events are presented.

Hurricane storm surge depends on the tidal stage, barometric pressure, Coriolis effect, wind stress, and wave forcing, as well as the local bathymetry. In the past, storm surge numerical models, such as Sea, Lake, and Overland Surges from Hurricanes (SLOSH), neglect wave forcing components to conserve computational efficiency. However, since hurricane wind, storm surge, and waves are coupled phenomena, numerous situations necessitate the inclusion of waves' effects to more correctly model hurricane storm surge. This paper describes the result of a collaborative effort by the National Institute of Standards and Technology (NIST), the National Oceanic and Atmospheric Administration (NOAA), and the University of Florida (UFl) to incorporate a wave model into the SLOSH model to extend its capability for storm surge simulation. A two-way coupling methodology was developed to incorporate two wave forcing components, that is, set-up from wave stresses and mass flux transport, into the SLOSH model. The aim was to better understand the relative contribution of each effect and their relationship to both storm strength and bathymetry. To this end, numerous simulations with different forcing variations (wind-stress only, wind and wave stresses, and wind and wave stresses with mass flux transport) were performed. The wind stresses used for the simulations were based on three hurricanes with different intensities, and the simulations were conducted for two basins with contrasting bathymetry in Florida (shallow and steep slopes). The results show that the impact of wave set-up and mass flux to storm surge levels varies between locations - even for the same storm in the same basin - proving that the interaction between the wind and wave forcing components is indeed complex. On average, however, the addition of the wave set-up and mass flux raised the maximum storm surge levels 10 to 30 percent, although isolated positions experience increased well above 100 percent.

This work aims to evaluate the ocean wave statistics (i.e., significant height, peak period, mean wave direction, and spectral form) under hurricane events, at any geographical point and coastal location by employing two different approaches. First, by the use of classical fast and mid-accurate analytical engineering tools comparing three classical set of equations (Bretschneider 1990; Young 1988; and the Shore Protection Manual SPM 1984) and their further validation against instrumental data provided by the National Data Buoy Centre (NDBC). On the other hand, by using a high-accurate numerical approach is proposed, using the unstructured-mesh SWAN spectral wave model (Zijlema et al. 2010), comparing the results between the stationary and non stationary approach, and validating the nearshore ocean wave characteristics with measured data at different water depth locations. Both approaches are forced with accurate pressure and wind speed fields obtained with classical and well-known parametric equations (i.e., Holland 1980; Hydromet-Rankin Vortex 1980; and Bretschneider 1990), integrated with the NCEP/NCAP wind fields. Hurricane waves for both approaches are validated with instrumental data from the NDBC measured for hurricanes: Katrina (2005); Gilbert (1998); Mitch (1998); Wilma (2005); and Ike (2008). Comparisons between both approaches have been made. Additionally, user recommendations and a sensitivity analysis about the optimization and design of the numerical mesh are presented, taking into account an efficient design of the numerical finite element mesh of the model, and the optimal spatial distribution of the nodes and elements along the hurricane track. The present study provides an efficient and easy engineering-mathematical tool to evaluate the hurricane induced waves, identifying its application limits and derived recommendations. Moreover, an accurate numeric methodology is here described, which allows to obtain the historical ocean wave data for all the historic events and tracks for North Atlantic and Pacific hurricanes, registered since mid-50's (lat-lon tracks, and central pressures provided by NOAA and the National Hurricane Centre), by improving the hurricane wind fields of any reanalysis database. Finally providing improved knowledge for coastal and harbor design considerations / construction purposes.

This paper examines the use of unmanned aerial vehicles (UAVs) as platforms to capture imagery at the neighborhood and individual building level for use in post-disaster field studies. Applications of UAVs in recent disasters as described in the literature are reviewed, and a case study investigating UAV capabilities for imagery collection following the EF-3 tornado impacting northern Alabama on March 2, 2012 is presented. Topics addressed include the multi-rotor UAV equipment and setup, imagery devices, flight considerations and capabilities, and imagery and metadata collection capabilities of the unmanned aerial system (UAS). Sample post-tornado imagery of building damage is shown, demonstrating the order of magnitude improvement in imagery resolution compared to typical post-disaster aerial photography.

As the costs of natural disasters increase, the need to improve building performance and better understand how buildings fail during these events grows. This can be difficult given the desire by the public to clean up and rebuild quickly, often meaning that damaged buildings are available only for a brief time, thus limiting opportunities to study their failure. Investigators struggling to understand the causes of building failures are often faced with the challenge of collecting significant amounts of information quickly while having only limited preliminary data to inform their field investigations. Additionally, funding constraints often limit the number of forensic teams that may be able to gather data during a deployment. In the 20 years since Hurricane Andrew, new technology has offered myriad opportunities to aid in providing field teams preliminary information and assist in gathering data. Research by others who were not deployed may be limited by an inconsistency in the data collected, a lack of data collected in an electronic format, and the lack of a central location or data warehouse to house the data. Focusing on the collection of research-related data, this paper discusses considerations that must be made when selecting an electronic data collection device and operating system, standardizing the collection of data, developing databases, and evaluating methods of data organization to provide information for post-disaster modeling or statistical analysis. Finally, the paper outlines some of the challenges associated with post-disaster data collection, such as privacy concerns and device security.

Techniques in post-disaster assessment from remote sensing imagery have been studied by different research communities in the past decade. Such an assessment benefits everybody from government organizations and insurance agencies to individual home owners. The key objects of interest in such images are buildings, as their destruction directly impacts lives. This work attempts at building a system capable of fine-grained damage analysis by comparing before and after storm images. While much of the assessment is still done manually, we present an automated system for Hurricane damage analysis. We propose and evaluate techniques for image registration, color balancing, building extraction, and damage classification; each of which requires little or no manual supervision.

This paper summarizes the challenges encountered during design and construction of a voluntary hurricane retrofit of an occupied 1960's era four-story concrete frame structure located in the United States. Recent hurricane-related disasters provoked the owner to review their essential building inventory to determine if they could withstand large-scale hurricane events. The owner has developed their own internal wind design criterion for the site that is significantly higher than the minimum code required wind design. A detailed review of the original design standards and construction documents determined that the essential building would not meet the level of performance required of a "hurricane shelter" that could remain operational during and after a significant hurricane event. A structural retrofit was required to bring the building's exterior envelope and lateral force-resisting system up to the owner's current wind design standards for essential facilities. A conceptual design study was performed to better understand the requirements of the building retrofit. Three schemes were investigated for strengthening the lateral force-resisting system, which included interior and exterior strengthening options. Given the three retrofit concepts, the owner decided to proceed with the exterior concrete shear wall system to minimize interruption to the building occupants and operations that could not be relocated during construction. Once the concept was chosen and the decision was made to proceed with design and construction, the owner began setting aggressive goals for the overall project timeline. The design team was faced with many challenges including time constraints, high wind design parameters, specification of various architectural components, consideration of construction techniques, and scope creep. This paper discusses this unique project case study and examines some of the project challenges in detail.

The purpose of the Storm Shelter standard, ICC 500, is to provide design and construction requirements for ‘safe rooms' that provide refuge from hurricanes and tornadoes. Hurricanes and tornadoes generate high winds that produce greater wind pressures on buildings and structures than they are typically designed for. In addition, hurricanes and tornadoes often generate flying debris. This standard provides design requirements the main wind resisting structural system and components and cladding of these shelters. Basic occupant life safety and health requirements for these shelters are addressed, including accessibility, means of egress, lighting, sanitation, ventilation, fire safety, and minimum required floor space for occupants. This standard was developed through a public hearing process by the International Code Council (ICC) in cooperation with the National Storm Shelter Association (NSSA).

The electrical utilities that supply power to most buildings are much more vulnerable than the buildings they serve, and emergency power requirements in most model codes and standards are focused on the minimum power that is required to allow occupants to leave a facility safely during a fire. Designing electrical utilities that are more vulnerable than the buildings they serve may be appropriate for facilities where power interruptions are of little or no significance, but more reliable electrical power is needed in critical facilities where electricity must be available to allow them to function. The Federal Emergency Management Agency is preparing a new guidance document, tentatively titled Emergency Power for Critical Facilities, that is scheduled for completion in 2012. The document will provide guidance on identifying emergency power needs, designing emergency power systems for existing critical facilities, and selecting the type of emergency power system.

High wind events such as hurricanes are one of the leading causes of electricity transmission line failures. There are two main failure mechanisms associated with de-energized transmission systems during hurricanes: 1) failure of the super-structures or the support towers; and 2) breakage of the transmission lines due to extreme wind loads. This paper presents a methodology for analyzing the reliability of spatially distributed transmission lines under hurricane wind hazard. As a case study, the vulnerability of the spatially distributed 230kv transmission lines in south Carolina was analyzed using 40,000 years of simulated hurricane events. Since the response of transmission lines under wind loads is highly nonlinear, a simplified nonlinear analytical model, verified by a 3-dimentional finite element model, was developed to simulate the tension forces considering the geometric nonlinearity induced by the sagging of the transmission lines. The reliability analysis results are presented in a GIS map, which can be easily used in pre-hurricane planning by power companies.

Distribution networks in wind-prone areas along the Gulf Coast of the United States suffer extensive damage during hurricane seasons. This often results in loss of power in both small and large geographical areas depending on the configuration of the network and the spatial distribution of damage to the network components. Wood poles support overhead lines, which are primary components of power distribution networks, and are highly vulnerable to failure during strong winds. This paper presents the failure fragility function of typical classes of wood poles against strong winds. Using the provisions of the ASCE7 standard, an analytical model for wind load profiles along distribution wood poles is considered. Also assessed are categories of dimensions and material properties of dominant classes of wood poles. Uncertainties in applied wind loads and the structural response of the wood poles are accounted for in this framework to improve the accuracy of the derived fragility curves. This is achieved by making reasonable assumptions on the dispersion of the pole dimensions and material properties. The Latin hypercube sampling technique is used to generate and randomly pair 5,000 samples of the parameters of wind load and pole response. A Monte Carlo simulation is then used to compare realizations of the wind demand and pole capacities across a wide range of wind speeds, which serve as the intensity measure of the demand. The failure fragility curve is developed from the comparisons. The results of this research constitute a portion of risk analysis of power distribution networks against strong winds.

RWDI has provided wind loading on the majority of the retractable roof stadia in North America and specifically on Minute Maid Park. Reliant Stadium and most recently Miami Marlins Ballpark in hurricane areas. The process of providing the wind loading on Reliant Stadium and Minute Maid Park provided the challenges of assessing the risks relating to the internal pressurization of the building in the roof closed position for hurricane passage. The internal pressurization of the stadium being impacted by the potential for debris impact including large glass wall breakage leading to dominant openings and the potential failure of windward seals between roof panels leading to potential dominant openings. RWDI worked collaboratively with Walter P. Moore and Associates to find solutions to these potential large internal pressure increases impacting design for hurricane conditions.

Tropical cyclones pose a significant threat to structures in Australia, especially to the growing populations and properties of coastal communities. In order to estimate losses from tropical cyclones in Australia, we have developed a comprehensive model that employs the latest research and published investigations of both meteorological hazards from tropical cyclones and the vulnerabilities of Australian structures. The model's loss output explicitly accounts for damage from tropical cyclone winds, precipitation-inducing flooding, and storm surge. Additionally, regional and temporal variability among structures in Australia is included in the model calculations. This variability reflects local cyclone experience, as well as differences in wind load design standards and implementation. The model was rigorously validated against published post-storm surveys conducted by Australian research groups, focusing on Cyclones Tracy (1974), Larry (2006), and Yasi (2011). The modeled damage ratios and loss estimates compare well with published damage reports and observed losses from insured properties.

The nature of hurricanes poses a significant challenge for making the decision on whether to evacuate or shelter in place. Uncertainties in the hurricane forecast, expected wind field, potential flooding, and performance of particular structures under high winds make it difficult to make a decision based on quantifiable data. By accounting for uncertainties in the hurricane forecast, as well as the performance of a structure, in a probabilistic manner, information can be determined which will aid in the decision on evacuation versus sheltering in place. A process has been developed to aid in the evacuation decision making process. This decision making process is applicable for individual structures as well as on a community wide level.

This paper introduces an overview of the issues and approaches when dealing with structure and infrastructure investment in the face of hurricanes or other natural hazards. It then expands on the difference between robustness and resilience. Finally, it presents some of the risk perception issues associated with hurricanes in the Southeastern region of the United States, placing into perspective the difficulty of fostering investment for different natural hazards.

In this paper, a framework for characterizing infrastructure damage due to hurricanes is outlined. The combination of high winds, rain and possible storm surge associated with hurricanes presents multiple difficulties for the proper functioning of the infrastructure. Wind-driven rain may have significant consequences for structures already weakened by extreme winds and airborne debris. Rain also contributes to flooding and may exacerbate storm surge damage.

A higher gust factor (GF), the ratio of a short-duration gust speed normalized by a longer-duration mean wind speed, has been identified in tropical cyclone (TC) wind relative to synoptically-generated extratropical (ET) wind in a number of studies (Yu and Chowdhury, 2009; Miller, 2006; Paulsen and Schroeder, 2004). These studies have focused on a single gust length and averaging duration when making comparisons. Data collected from within fourteen landfalling tropical cyclones was compared with ET data collected during an experiment conducted west of Lubbock, TX. GFs were computed using a 2-second gust based on three mean lengths, 10-minutes, 5-minutes, and 1-minute, and differences between the TC and ET GFs were evaluated. The percent difference between the TC and ET GFs was largest for the 10- and 5- minute segments and smallest for the 1-minute segment. Histograms for the various GF populations also revealed differences in distribution shape and position for the two environments that were more significant for the 10- and 5-minute data than for the 1-minute data. These findings make sense when comparing power spectral density (PSD) plots of data from the TC and ET environments. The two environments have similar distributions of high-frequency energy, but differences emerge between the two distributions at lower frequencies. As a result, the 2-second gust taken from the longer (5- and 10-minute) segments reflects the difference in low-frequency variation between the two environments better than the 2-second gust taken from the 1-minute segment, which captures only the variation present in the portion of the spectrum the TC and ET data have in common. As a result, the distributions of GFs for TC and ET data are most alike when considering a shorter averaging duration.

The paper highlights analysis of full scale wind measurements conducted during typhoons in South China Sea. First, data from Doppler radar wind profiler were employed to examine characteristics of typhoon vertical wind profile over two types of underlying surface categories. The wind field characteristics in the regions of interest for engineering applications were obtained. These wind profiles were characterized by a three-layer flow field and the data were fitted to empirical expressions. Characteristics of typhoon wind spectra were examined and compared in detail to well established descriptions of extratropical storms. It was noted that the wind spectra in typhoon wind field showed a leveling off trend at relatively higher frequency range, which is contrary to the traditional energy cascade mechanism noted in turbulent flows. The mechanism responsible for this feature was examined and a plausible interpretation was proposed.

Methods of analysis and modeling of the distribution of surface winds about tropical cyclones have increasingly recognized the deficiency of the classical representation of the radial distribution of (e.g. "Rankine vortex", "Holland profile") of a linear increase of wind speed from the center to a single and tight radius of maximum and a sharp decay to the ambient flow. Newer analysis schemes and parametric and dynamical models now attempt to allow for more complex radial distributions including the possibility of multiple radial wind maxima and storms with indistinct radial maxima and shelf-like profiles extending far into the ambient flow. This paper demonstrates the rather high degree of skill with which a representation of the radial pressure profile with a double exponential form within the initialization scheme of the widely applied "TC96" steady state planetary boundary model (PBL) succeeds in modeling such wind field complexities as well as the large range of wind field spatial scales associated with have come to be known as "compact", "incompact" and "annular" cyclones, and in representing the life cycle of storms that undergo rapid dynamic transformation in wind structure associated with the eye-wall replacement cycle and transitory storm encounters with land masses. Examples are drawn from recent well monitored U.S. Gulf and East Coast historical hurricanes, but the characteristics demonstrated appear to apply to tropical cyclone worldwide.

The most appropriate statistical technique to determine a peak pressure coefficient from wind tunnel data is not a settled issue. This lack of standardization can lead to inconsistent definitions of the peak pressure coefficient. Recent papers have proposed a peak pressure coefficient cumulative distribution function (CDF) estimation procedure which utilizes the entirety of a short time history rather than sampled peaks from a long time history to derive the peak distribution. This method relies upon the accurate probabilistic modeling of the time history. The first part of this paper will present a robust PDF model developed to represent fluctuating pressures over the entire surface of a low-rise building. The existing short time history peak CDF estimation method is then modified using this new PDF model. Very long duration wind tunnel data is used to provide an empirical peak CDF to benchmark the performance of the existing and proposed new method.

This paper proposes a modular framework utilized to assess the risk of building envelope failures due to hurricane wind hazards. A component-based approach is taken to develop an integrated building envelope model that is based on previous research of individual component and system capacities. Key modules of the proposed framework include a wind-borne debris generation module and an impact-tracking module that will interact with a hurricane simulation module capable of simulating synthetic hurricanes for various return periods and historical hurricanes. One major difference between the proposed framework and other risk assessment models is that the proposed framework is extremely flexible in allowing the user to define the building stock within the area of study, which will provide the user with the ability to investigate an unlimited number of "what if" scenarios. Another distinction of the proposed framework is that it is driven by a three-dimensional probabilistic debris trajectory model developed by the authors, rather than using damage curves developed from observed post-hurricane assessments or insurance claim data. Debris impact risk plots presented in a polar coordinate system are developed using the framework and can be utilized either pre- or post-construction to mitigate the damage to the building envelope of homes within a subdivision during a hurricane event.

A probabilistic analysis is applied on full scale data and numerical simulations to characterize extreme values of pressure coefficients on low-rise structures. The full scale data analyzed here, provided by the Florida Coastal Monitoring Program, are obtained from measurements made on the roof of a house in Florida during Hurricane Ivan in 2004. The analysis provides a probability of non-exceedence of any given load coefficient. This probability is discussed in the context of ASCE 7 GCp values. The same analysis is applied to data obtained from numerical simulations of the flow over a surface-mounted prism. The objective is to quantify the effects of inflow disturbances on the distribution and probability of non-exceedence of peak wind loads on the roof of the prism.

The dynamic properties of the cantilever sign structures such as its mode shapes and frequency for different modes are being studied under wind load. The analysis considered the dead loads of the structure principally the self-weight of the truss, sign, and walkways, and the wind load. Finite element analysis are conducted to evaluate the modal stresses in the truss and compared to the total stresses. The first three mode shapes in the wind load direction and the gravity load direction are modeled. The modal participating mass ratios of the sign structure for different mode shapes are also calculated. Results show that in chord members with 24 inch length, 50 percent and in chord members with 48 inch length, almost all the stresses (bending + axial) are modal stresses. In interior diagonal members 40 percent, in horizontal diagonal members 75 percent and in column part almost 75 percent of stresses (bending + axial) are modal stresses. The purpose of the study is a modal and static analysis in critical members of a cantilever sign structure under wind loading using finite element method.

Large scale wind flow simulation over complex topography requires a tremendous amount of computational resources, both in terms of CPU hours and storage requirement. Use of complex turbulence models such as large eddy simulation (LES) significantly add to the computational demand. In practice, a relatively less accurate turbulence model such as Reynolds Average Navier Stokes (RANS) equations are solved due to computational resource limitations. If computing power is available, parallel simulation can be carried out to cut down simulation time without degrading the quality of flow modeling. Coarse grained parallelism in a distributed memory cluster is traditionally achieved by domain partitioning strategies. The whole domain is partitioned into smaller sub-domains which are then assigned to a processor in a cluster of computers. In this work a non-overlapping type of domain decomposition method is implemented where information such as pressure and velocity is exchanged at the boundaries of the domain through ghost cells. The solution of all sub domains goes on simultaneously. The message passing interface (MPI) is used to exchange information between sub-domains. The second focus of this work is to examine suitability of general purpose graphic processing units (GPGPU) for solving fluid flow problems. Nowadays GPUs are becoming a much cheaper alternative for speeding up CFD simulations significantly. Speed ups of up to 100 times compared to a single CPU implementation have been reported in literature for simple flow cases. In this study, NVIDIA's CUDA programming is used to exploit the fine grained parallelism offered by GPUs, and use it for practical fluid flow problems. Both methods of parallelization are combined at the solver level so that a mixed CPU-GPU computation is possible. Finally the speed up numbers obtained for different size problems are compared.

A study was conducted to determine the in-place withdrawal capacities of nails in existing residential wood roofs. A field investigation project carried out by the Florida Coastal Monitoring Program collected data from 1400 nails tests in situ at 17 Florida homes, using a portable device to measure withdrawal capacity. The results reveal nail withdrawal capacities well below the values determined by the National Design Specification (NDS) empirical equation. A modified portable nail extractor (m-PNE) was developed and used to determine the withdrawal capacity of residential roof decking nails, and to provide added data to explain low values. Concurrently, laboratory tests were conducted on approximately 1300 6d common galvanized nails installed in southern yellow pine (SYP) framing members, using the m-PNE and the standard test protocol method (ASTM D1761) for a benchmark comparison. The m-PNE mean nail withdrawal capacities were 30% lower than the withdrawal capacities obtained following ASTM D1761. Test results suggest that the measured nail capacity is influenced by the method of localized sheathing removal to access the nail head. While some loss in strength may be due to environmental factors and aging, it is suggested that driving nails through sheathing may contribute to lowered capacity. Preliminary results suggest that actual factor of safety for nail withdrawal is likely smaller than the 5.0 assumed in the NDS, indicating less conservatism than intended.

Internal pressure can be a significant component of the overall net pressure the cladding would experience during wind storms. These internal pressures are in the order of about 10% to sometimes even 50% of the overall pressure exerted on the cladding elements. The higher percentages are usually the result of a dominant opening on the façade. Such a dominant opening in the middle of the façade of an apartment with single face exposed to the exterior could in fact decrease the loading. In such situation, instead of assuming the dominant opening, RWDI consider nominal leakage suggested by the code as the internal pressure. For corner apartments, where orthogonal building faces have been exposed to wind, the opening in one façade could exert high internal pressure to the orthogonal façade in phase with the high external pressure acting on it. This could result in very high overall pressure. Note that the area of internal pressures die to wind has received very little attention in the past. Also, the available past research by others has very little touch to the practical application. In the past, RWDI has developed a novel methodology accounting for the above aspects of internal pressures assuming a certain probability of breakage. Although there has been an increase in the amount of research into wind borne debris and the resistance of various glazing materials to an impact, application of these results in relation to the probability of breakage and the design for wind loading needs to be investigated. The aim of this paper is to propose a link between the research considering wind borne debris damage and the approximation of internal pressures from the results of wind tunnel pressure tests. The random and chaotic nature of many of the variables involved in the cause of an opening in the building façade, and the subsequent formation of internal pressures makes a deterministic solution almost impossible to achieve. As such, the paper outlines a probabilistic approach to merge the two lines of research. The results of research presented on the subjects of damage due to wind borne debris are used as the basis for determining the probability of a dominant opening occurring on a building façade. Differences in relationships between this probability of an opening and the positive and negative internal pressure scenarios are also investigated. Internal pressures are estimated corresponding to a selected reasonable risk level, or return period, using the probability of an opening applied to the results of wind tunnel tests.

Hurricane is the most financially devastative natural hazard. However, the prediction accuracy of the current two public hurricane damage prediction models for low-rise residential buildings is constrained by the modified ASCE code version and the pre-defined load distribution among the load bearing components. The current study proposed a rigorous Finite Element model that enables not only to accommodate the wind tunnel test or field test database, but also to update the progressively damaged building portion during the wind-structure interaction. The building envelope, including the sheathing panels and the nailing connections that were witnessed as the most vulnerable portion of the building, was modeled with extensive details. Shear failure of the roof sheathing was found to be one initiative failure mechanism and the hot spots agree well with the damaged roof sheathing by the post disaster survey. Under the same uniform uplift pressure, the building did experience the progressive failure after the initiation of the envelope damage.

This paper presents preliminary results of a study of building damage caused by Hurricane Andrew as a function of building age. The information came from a 68,000 insurance claim records. The building locations were subdivided in regions to reflect wind speeds intensities. In the region close to the hurricane track, the newer buildings on average performed poorer than older buildings. In regions with milder winds, the newer buildings performed somewhat better. It has also been observed that the relationship between year built and damage ratio is neither linear nor monotonic at all distances from the hurricane track. Future refinements are expected to this study when appropriate wind speeds estimation become available.

Accurate estimation of risk to structures due to hurricane wind hazards is critical for emergency planning and post-event recovery. A widely used model for assessing risks due to hurricane winds at the county or zip code level is the HAZUS-MH4 suite of programs developed by the Federal Emergency Management Agency (FEMA). Very few studies have attempted to assess the predictive accuracy of HAZUS wind risk models at levels of geographical specificity below counties and zip codes. In this paper, we present a methodology for computing HAZUS predictions of wind damage aggregated to the level of one kilometer square blocks. We use detailed damage data on over 700,000 single-family residences in Harris County collected in September 2008 after Hurricane Ike for evaluating the predictive performance of HAZUS. We show that at this level of aggregation, HAZUS has a predictive accuracy of 29.5%. This finding raises the question of how to improve the predictive accuracy of HAZUS-MH4 models at this level of spatial specificity. It is difficult to manually analyze errors made by HAZUS over a data set with 700,000+ structures, so we devise a statistical machine learning approach to distinguish structures whose damage is correctly predicted by HAZUS from those that are incorrectly predicted. We identify ten potential explanatory variables at the level of each building. These variables include number of floors, terrain, age, whether a building was remodeled, building and land value, proximity to highway, and more refined wind measurements including wind speed, wind swath and wind direction. Our machine learning classifier reveals the relative importance of these variables for improving HAZUS's predictive performance. In particular, we find that HAZUS's fragility curves for one and two floor residences in urban terrains need to be refined by incorporating building and land value, age, proximity to a major highway, as well as wind direction and wind swath. These results can be used to build the next generation of HAZUS fragility curves for accurately predicting hurricane wind damage to structures at a spatial specificity of one kilometer square blocks.

A review of wind mitigation benefit/cost analysis (BCA) results for an initial set of 60 residential dwellings analyzed for the Coastal Retrofit Mississippi (CRM) project is presented. An overview of the three mitigation packages available in the CRM project is presented, along with a discussion of the physics-based modeling approach implemented in the FEMA BCA methodology for hurricane wind mitigation projects. Benefit/Cost Ratios (BCRs) for the initial set of 60 projects are examined as a function of the mitigation package type, project cost, building exposure category, and other characteristics. The cumulative BCR for the first 60 projects is 1.60, indicating an expected net present value of $1.60 in avoided building losses, contents losses, and loss of use expenses over the useful life of the retrofits for each $1.00 in net present value of project inspection, analysis, construction and maintenance costs.

Typical North American wood frame houses have trusses that are toe-nailed to the wall top-plates. In order to develop models of roof failures for such structures, considering the effects of wind speed, direction and duration for a storm, as well as the structural details, knowledge of the load sharing and how it is altered by the nail slip at each of the connections is required. The objective of the current paper is to quantify the changes in load sharing during nail/connection slip. This is achieved by simplifying the roof structure to the essential components of a beam, which bends elastically in the direction normal to the trusses, and real, toe-nailed connections. Test results for a single "roof stiffness" and truss spacing show that significant load sharing is not observed until permanent displacements (i.e., damage) to a connection occur. Dynamic load tests, with the same time-varying load applied to each connection, have shown that the load sharing increases with the damage accumulation to each of the connections and changes continuously. Ultimately, this information will be used in the development of analytical models that are able to consider storm duration effects on the performance and failure of wood-frame houses in extreme winds.

A new, large-scale, hurricane simulator is under construction at the Powell Family Structures and Materials Laboratory at the East Campus of the University of Florida. This facility was designed to evaluate the performance of building components and cladding systems under extreme windstorm conditions in a controlled laboratory environment. The simulator will be capable of reproducing Saffir-Simpson Hurricane Scale Category 5 hurricane wind loads on full-scale building mockups with a maximum area of 40 m². The simulator was designed to replicate dynamic wind pressure waveforms up to 23 kPa. The facility should be able to test specimens that exhibit large air leakage (< 47 m³/s) and deflections (< 1 m). The paper provides an update on the development of the facility, which will be operational in summer 2012. Details about the final configuration, validation and initial experiments will be discussed during the conference session.

A low-rise wood building equipped with pressure and force-monitoring equipment was used to study the wind-induced response during strong wind events. The main interest in this study was to examine the wind load path, i.e. how the dynamic wind load is transferred through the building superstructure and down to the foundation. By investigating the load transfer mechanisms it was possible to identify, and most importantly, quantify the degree of wind load attenuation, as this load reaches the foundation level. The results identified a significant degree of attenuation of wind-induced load. For most cases the predicted peak forces (FEM) exceed the observed forces captured by the load cells (full-scale). This behavior indicates that although the applied load should generate predicted responses at various components, e.g. roof to wall, wall to foundation interfaces, actual dynamic force monitoring reveals attenuation of peak forces and reactions.

The Insurance Institute for Business & Home Safety (IBHS) research facility is capable of subjecting full scale structures to high wind loads. In order to validate the facility, comparisons of the mean velocity, turbulence intensity profiles and turbulent spectra are made to both field measurements from Texas Tech University and theoretical profiles and spectral estimates. The match of the mean velocity profiles and turbulence intensity profiles was good in both the longitudinal and lateral directions. The match of the turbulent spectra from the IBHS facility is good, with a small spectra gap between wave numbers of 0.01 and 0.1 and too much energy at small scales, similar to model scale wind tunnels. The match of point pressures on a building compared to model scale and full scale experiments is good, with the results from the IBHS facility generally falling between the full scale and model scale results in the separated regions on the roof.

Parapet walls around the perimeter of a building serve several functions including changing the wind-induced pressures on the walls and roof surfaces of the building. To quantify the effects of parapet walls on a building's surfaces full-scale data was collected on the Texas Tech University's (TTU) Wind Engineering Field Laboratory (WERFL) test building with and without a parapet around the entire building perimeter. The WERFL test building is instrumented with 204 pressure transducers connected to taps flush mounted on the surface of the building providing the capability to observe the pressures on the building surface and parapet with a high degree of detail. The pressure taps are positioned on the building surface with two planes of aerodynamic symmetry. The positions allow the pressure data to be reflected about these axes of symmetry thus quadrupling the data set for better resolution of confidence intervals. Three data sets were collected for comparison; WERFL without a parapet, WERFL with a 1.2 m (48 in) parapet, and WERFL with a 85.1 cm (33.5 in) parapet. The data was analyzed based on the summary statistics of the validated pressure coefficients. The pressure coefficients studies are based on 15 degree increments of wind angle of attacks from 0 thru 359 degrees. The effects of the parapets on the roof pressures as a function of the parapet height is presented. Generally the parapets reduced the magnitude of the negative mean roof pressures while the leeward wall parapets contain the airflow creating slight positive pressure on the flat roof of the test structure.

This paper summarizes results from a study directed at developing design wind speed maps for use in the Caribbean. The study is described in detail in Vickery and Wadhera (2008b). The design wind speeds were developed using a hurricane simulation model similar to that used in the development of the ASCE 7-10 wind speeds. Maps were produced for return periods of 50, 100, 700 and 1,700 years so that the countries adopting the IBC Codes, or using ASCE 7-10 to develop design loads, would have design maps consistent with the design wind speed map used in the United States.

The Caribbean Coastal Ocean Observing System (CariCOOS) is the observing arm of the Caribbean Regional Association for Integrated Coastal Ocean Observing (CaRA). This effort, funded by the NOAA Integrated Ocean Observing System (IOOS) office, is one of eleven coastal observing systems and regional associations which along with federal agencies constitute the national coastal component of the US Integrated Ocean Observing System. CariCOOS observational assets provide a real time network to monitor wind, waves and currents and consists of four ocean data buoys and thirteen hurricane-hardened land based weather stations, distributed along the coastal areas of Puerto Rico and the US Virgin Islands. In this paper we present a detailed description of our observational infrastructure and provide analysis of wind and wave observations during the passage of Hurricane Irene over Puerto Rico and the US Virgin Islands during August 21-22, 2011.

Hurricanes routinely cause the loss of functionality of referral hospitals in the Caribbean, North and Central America. This is unacceptable and an avoidable inconvenience for the affected communities. This has been clearly recognised by the Pan American Health Organisation (PAHO). Since 1985 there has been a concerted effort on the part of PAHO to rectify that situation. PAHO created a Disaster Mitigation Advisory Group (DiMAG) in 2003 and members of that group have been assisting PAHO in developing procedures, guidelines, assessment tools and courses towards the achievement of that end. In addition, eight governments in Latin America and the Caribbean have formally committed themselves to promote and support and implement processes leading to safe hospitals. Specific programmes have been developed to lead to the elimination of disasters in the health sector and the reduction of risk. Some of these programmes reaffirm what already exists, some strengthen existing weak programmes and some are new in at least some of the countries of the hemisphere. The programmes include: education of engineers and architects;post-graduation training of engineers and architects; development of building standards; development of code "overlay" documents dealing specifically with the requirements of referral hospitals; implementation of independent reviews of designs and of quality assurance systems during construction; attention to the non-structural components of hospitals; vulnerability assessments of existing hospitals; retrofitting actions for vulnerable, existing hospitals; maintenance programmes.

The paper summarizes recent research and development of computer-aided and computational methods in wind engineering at the National Institute of Standards and Technology (NIST). Specific topics to be covered include: 1. Time-domain database-assisted design of tall, flexible structures excited by turbulent winds inducing dynamic structural responses. 2. Computational Fluid Dynamics (CFD) estimates of aerodynamic pressure on low-rise buildings with geometrical dimensions comparable to the across-wind integral turbulence length scale. 3. Development of synthetic directional wind speed databases covering on the order of 10,000 years, from measured wind speeds over periods of approximately 30 years. 4. Probability distributions of hurricane wind speeds, Gulf and Atlantic coasts. 5. Extreme wind speeds in non-hurricane prone conterminous United States. 6. Estimates of hurricane-borne missile speeds. 7. Disaster and Failure Events Repository - a new research tool for archival and analysis of post-disaster information and data.

This paper introduces a probabilistic Performance Based Hurricane Engineering (PBHE) framework for risk assessment based on the total probability theorem. This methodology disaggregates the risk assessment analysis into independent elementary components, namely hazard analysis, structural characterization, interaction analysis, structural analysis, damage analysis, and loss analysis. The proposed PBHE framework accounts for the multi-hazard nature of hurricanes. The procedure for the probabilistic risk assessment is illustrated through an application example consisting of a residential house subjected to the combined hazards of windborne debris and hurricane strong winds.

In 2008, ABS Consulting performed a hurricane wind vulnerability assessment for the new Marlins Ballpark that was recently constructed in Miami, Florida. The assessment was intended to provide a probabilistic loss analysis of the ballpark due to direct damage from hurricane winds and subsequent potential damage related to the storm surge that may be produced by the event. Loss assessments were developed at selected points during construction, and after stadium completion. The probabilistic loss analysis was performed for the recently completed Marlins Ballpark in Miami, Florida, using ABS Consulting's proprietary computer software program WORLDCAT Enterprise^TM to develop an aggregate damage loss estimate for the site. A stochastic hurricane event set was used to estimate damage and insured losses due to the probabilistic occurrence of hurricanes. Project specific damage curves were used to represent unique features related to the design and construction utilized for the new ballpark. In addition, input was provided by the design team and the contractor to further refine the potential vulnerabities during the construction phase of the project

In traditional wind engineering design, the response of a building is essentially at or below the first significant yield point in some members of the lateral load resisting system under ASCE 7 Strength Level load combinations. Engineers have generally not considered pushing this threshold primarily because there is no established framework for consideration of inelastic behavior under wind loading. In buildings subjected to seismic loads, controlled inelastic response is permitted by the building codes for a safe and economically feasible design. The concept of allowing a controlled inelastic response could also be extended to the analysis and design of tall, slender buildings subject to wind loads, where the dynamic wind component is a significant portion of the total wind load. In this paper, a framework for performance-based wind engineering is proposed. Selection of building performance objectives, quantification of wind hazards and methods for explicit consideration of nonlinear behavior are outlined. The proposed methodology can be applied to evaluate the performance of existing buildings or to economically design new buildings subjected to wind loads.

Buildings subjected to wind forces are designed using elastic analysis and equivalent static lateral loads. When existing buildings are subject to alterations, conformance under wind provisions is currently evaluated using the same approach. However, significant benefits can be obtained with the adoption of a performance-based technique, where member inelastic behavior and dynamic effects of the natural hazard are explicitly considered. Such strategies have been successfully adopted in seismic engineering. This paper presents a performance-based evaluation of an existing building with concrete core walls as the lateral resisting system. Nondestructive testing, nonlinear material models and dynamic loads from wind tunnel testing were utilized to determine the adequacy of the lateral load system. Building responses at the target performance level were compared with owner-ratified performance criteria. The evaluation highlighted deficient elements in the lateral system and enabled the owner to make well informed decisions regarding retrofit strategies.