PBHE, Hurricane, surge, wave, spectral analysis, joint pdf


The magnitude of damage caused to the United States (US) coast due to hurricanes has increased significantly in the last decade. During the period 2004-2005, the US experienced seven of the costliest hurricanes in the country's history (NWS TPC-5, 2007) leading to an estimated loss of ~ $158 billion. The present method for predicting hurricane losses, HAZUS (HAZard US), is solely based on hurricane hazard and damage caused to building envelopes only and not to structural systems (Vickery et al., 2006). This method does not take into account an intermediate step that allows for better damage estimates, which is structural response to the hazards that in turn can be mapped to the damage. The focus of this study was to quantify the uncertainty in response of structures to the hurricane hazards associated with hurricanes from performance based engineering perspective. The study enumerates hazards associated with hurricanes events. The hazards considered can be quantified using a variety of measures, such as wind speed intensities, wave and surge heights. These hazards are quantified in terms of structural loads and are then applied to a structural system. Following that, structural analysis was performed to estimate the response from the structural system for given loads. All the possible responses are measured and they are fitted with suitable probability distribution to estimate the probability of a response. The response measured then can be used to understand the performance of a given structure under the various hurricane loads. Dynamic vs. static analysis was performed and results were compared. This will answer a few questions like, if there is any need to do both static and dynamic analysis and how hurricane loads affect the structural material models. This being an exploratory study, available resources, research, and models were used. For generation of annual or extreme values of hazard, various available wind speed, storm surge, and wave height models were studied and evaluated. The wind field model by Batts et al. (1980) was selected for generation of annual wind speed data. For calculation of maximum storm surge height, the Sea, Lake Overland Surges from Hurricane (SLOSH, Jelesnianski et al., 1992) program was used. Wave data was acquired from a National Oceanic and Atmospheric Administration (NOAA) database. The (extreme or annual) wind speed, surge height, and wave height generated were then fitted by suitable probability distributions to find the realizations of hazards and their probabilities. The distribution properties were calculated, correlations between the data were established, and a joint probability distribution function (PDF) of the parameters (wind speed, wave height, and storm surge) was generated. Once the joint distribution of extreme loads was established, the next step was to measure the dynamic response of the structural system to these hazards. To measure the structural response, a finite element model of three-story concrete frame were constructed. Time histories of wind load were generated from wind net pressure coefficients recorded in a wind tunnel test (Main and Fritz, 2006). Wave load time histories were generated using laboratory basin test (Hawke's et al., 1993) wave height time history data and were converted into wave loads using Bernoulli's equation. Surge height was treated as a hydrostatic load in this analysis. These load time histories were then applied to the finite element model and response was measured. Response of the structural system was measured in terms of the mean and maximum displacements recorded at specific nodes of model. Response was calculated for loads having constant mean wind speed and surge/wave and different time histories. The dominant frequency in the wind load time histories was closer to the natural frequency of the structural model used than the dominant frequency in the wave height time histories. Trends in the response for various combinations of mean wind speed, wave height, and surge heights were analyzed. It was observed that responses are amplified with increase in the mean wind speed. Less response was measured for change in mean surge/wave height as the tributary area for wave forces was less compared to wind force. No increase in dynamic amplification factor was observed for increase in force time histories case.


If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at

Graduation Date



Mackie, Kevin


Master of Science (M.S.)


College of Engineering and Computer Science


Civil and Environmental Engineering

Degree Program

Civil Engineering








Release Date

May 2010

Length of Campus-only Access


Access Status

Masters Thesis (Open Access)