Journal of Structural Engineering

Civil engineering structures are part of our cultural heritage. The story of who we are can be told, in part, by what we have built. There have been pivotal moments in civil engineering design history wherein a master engineer creates a pioneering structure. One major example is Anton Tedesko’s 1936 Hershey Ice Arena, the first large-scale thin shell concrete roof in the United States. Tedesko left all his papers, including the original design and analysis calculations of the Hershey shell, to the Princeton Maillart Archives. These documents, as well as other archival materials and photographs, provide insight into the design history of Hershey, and the transfer of thin shell technology to America. In this paper, we retrace the design and analysis calculations performed by Tedesko, and compare them to modern computer models. We show that the hand calculations are sufficiently accurate, and in fact are necessary for initial form finding. We close by pointing out the enormous impact that this design had in thin shell concrete construction, and argue for the preservation of this remarkable structure.

Two geometrically identical half-scale precast concrete cantilever wall units were constructed and tested under quasi-static reversed cyclic lateral loading. One unit was a code compliant conventionally reinforced specimen, designed to emulate the behavior of a ductile cast-in-place concrete wall. The other unit was part of a precast partially prestressed system that incorporated post-tensioned unbonded carbon fiber tendons and steel fiber reinforced concrete. Hysteretic energy dissipation devices were provided in the latter unit in the form of low yield strength tapered longitudinal reinforcement, acting as a fuse connection between the wall panel and the foundation beam. The conventional precast reinforced wall performed very well in terms of the ductility capacity and energy absorption capability, reaching 2.5% drift before significant strength degradation occurred. The precast partially prestressed wall unit achieved drift levels well in excess of 3% with no visible damage to the wall panel prior to failure. Test results and performance comparisons between the precast partially prestressed wall system and the precast conventionally reinforced unit are presented.

A series of experiments was carried out to investigate the behavior of columns with high-strength concrete subjected to shear reversals into the nonlinear range of response. The main experimental variables were loading history, axial load, and concrete strength. Columns measured $0.20\times 0.20m$ $(8\times 8\mathrm{in}\text{.})$ in cross section and had shear span to depth ratios of 3.4. The transverse reinforcement ratio was 1%. Column behavior was closely related to the maximum strain that was measured in the stirrups during each loading cycle, which was affected by the displacement history and axial load applied to the specimens. The level of axial load had an increasingly adverse effect on the limiting drift of specimens subjected to axial loads larger than one half the balanced load, while the opposite effect was observed in specimens with lower values of axial load. Despite its apparent brittleness under axial compression, high-strength concrete was found to improve the toughness of reinforced concrete columns and extended the drift range within a column would not require repairs after an earthquake. In specimens made with high-strength concrete, the amount of reinforcement dictated by ACI 318-99 in 1999 provided sufficient confinement to obtain limiting drift ratios exceeding 4%.

The American Concrete Institute (ACI) 318-02 provisions for stress block parameters were derived using normal-strength concrete column tests. Azizinamini et al. in 1994, Ibrahim and MacGregor in 1996, Bayrak in 1999 have reported that the use of current ACI stress block parameters results in estimations that are not conservative for high-strength concrete column sections. This was mainly attributed to the early cover spalling in high-strength concrete columns. In this paper, a numerical procedure for simulating premature cover spalling in high-strength concrete columns $\left(66\mathrm{MPa}<f_c^{\prime }<140\mathrm{MPa}\right)$ is presented. The proposed method for nominal moment capacity calculation employs rectangular stress block parameters and an analytically derived strength reduction parameter to integrate the effect of early cover spalling into the design process. Although the proposed method for moment capacity calculation is analytically derived, its results are validated with the experimental data from 224 column tests reported in the literature. Twenty-four of the 224 tests were conducted by Bayrak as part of a comprehensive research program in which the behavior of reinforced concrete columns was studied and provided the motivation for the research reported here. ACI 318-02 provisions for stress block parameters are critically examined using the experimental data from 224 tests. It is concluded that the ACI stress block parameters do not result in conservative estimations for the moment capacity of high-strength concrete column sections failing in compression. However, the proposed technique results in conservative estimations of strength for high-strength concrete column designs as well as normal strength concrete column designs.

A rigorous analytical procedure for the crack analysis of concrete dams is established using a newly developed numerical technique for analyzing multiple discrete cracks in concrete. The theoretical aspect of the problem solution is addressed first, focusing on the cracking modes of multiple cracks, followed by a brief review on the numerical formulation of the Mode-I type crack propagation for multiple discrete cracks. Next, a simple technique for modeling curvilinear crack propagation is introduced, and the method is examined through numerical studies on the bending tests of simple plain concrete beams with multiple cracks. Then, a well-quoted scale model of a concrete gravity dam is used to study the cracking behaviors in concrete dams; the numerical modeling involves not only the original single-crack problems but also multiple-crack problems. Employing two different types of numerical models and assuming multiple initial notches of different sizes at different locations along the upstream face of the model dams, various kinds of cracking behaviors are obtained and discussed, focusing on the crack interactions.

Existing microplane models for concrete use three-dimensional spherical microplanes in the analysis of two-dimensional planar members as well as three-dimensional members. Also, they do not accurately describe the postcracking behavior of reinforced concrete in tension-compression. In this study, a new microplane model is developed to overcome the disadvantages of the existing models. Instead of the existing spherical microplanes, the proposed microplane model uses disk microplanes involving a less number of microplanes and two-dimensional stresses and strains. As the result, the proposed model is more effective in numerical calculation. Also, the concept of the strain boundary is introduced to describe accurately the compressive behavior of reinforced concrete with tensile cracks in tension-compression. The validity of the proposed model was verified by comparison with existing experiments. In this paper, the microplane model and the numerical techniques involved in the finite-element analysis are described in detail.

Hollow structural steel (HSS) columns filled with concrete offer a number of benefits and are often used in tall buildings and other industrial structures. A mathematical model was developed and used to investigate the influence of important parameters that determine the fire resistance of concrete-filled hollow steel columns. The behavior of concrete-filled HSS columns with or without fire protection subjected to axial or eccentric loads was experimentally investigated and the results presented in this paper. The parametric and experimental studies provide information for the development of formulas for the calculation of the fire resistance and fire protection thickness of the composite columns. The main objectives of this work were threefold: First, to report a series of fire tests on composite columns; second, to analyze the influence of several parameters, such as the sectional dimension, load eccentricity ratio, and fire protection thickness on the fire resistance of the composite columns. It was found, in general, that the sectional dimension and the fire protection thickness have significant influence on the fire resistance of concrete-filled HSS columns; however, the load eccentricity ratio has moderate influence on the fire resistance. The final objective was to develop formulas for the calculation of the fire resistance and the fire protection thickness of the concrete-filled HSS columns.

It was found that a very small time step might be needed in the step-by-step solution of the shock response from an impulse. In fact, this time step might be much smaller than the time step needed for accurate representation of the impulse and negligible period distortion. In this study, it is analytically verified that this difficulty arises from the discontinuity in the external force. In addition, a technique used to estimate the relative amplitude error caused by this discontinuity is also developed. In fact, the numerical accuracy of the shock response from a linear impulse is thoroughly explored in terms of both amplitude and period. Analytical results show that the numerical accuracy might be entirely dominated by amplitude distortion for an impulse with significant discontinuity in the external force. However, it is generally controlled by period distortion if there is no discontinuity in the external force. It is also found that amplitude distortion is not significantly affected by viscous damping. Guidelines for obtaining an accurate shock response from an impulse are proposed and thoroughly verified through numerical examples.

The dynamic characteristics of a cantilevered traffic signal structure were determined experimentally and theoretically. The experimental procedures involved forced vibration of the structure by an eccentric-mass oscillator. Theoretical characteristics were determined using computer modeling. Several devices were developed to provide additional damping. The response spectra of the retrofitted structure were compared to the response spectrum of the non-retrofitted structure. Theoretical response curves were developed and compared to the experimental results. The most viable devices included a dual-strut system equipped with shock absorbers, and a tuned-mass impact damper. In the former, the damping was increased by an average factor of 6.3 in the in-plane direction, and 1.8 in the out-of-plane direction when considering the response at the box connection. In the latter, the most promising orientation yielded a 3.5 and 2.4 fold increase in damping in the in-plane and out-of-plane directions, respectively.

The reduction of the flanges of steel wide flange beams near beam-column joints in special moment resistant frames is currently an acceptable alteration of a connection which connects a beam and column with complete-joint-penetration groove welded flanges and a welded or bolted web connection. Herein, the closed-form stiffness matrix of a two-dimensional frame element with radius flange reductions symmetric about the centroid of the element is derived using virtual work theories. Tip translations of a cantilevered $W24\times 103$ beam containing a maximum reduction in flange area of 44% are obtained using two methods: The analytically derived stiffness matrix of the reduced beam section (RBS) frame element and finite-element analyses of the beam discretized with solid elements. The differences between the results of the two analyses range from 0.3 to 3.9%. The stiffness matrix terms of a $W24\times 76$ beam with and without flange reductions are compared. Again, the frame element includes a 44% maximum reduction of beam flange area. The resulting reduction in stiffness terms ranges from 3.6 to 15.1%. The rotational stiffness decreased by 10.4%. The effects of changes in the dimensions and location of radius flange reductions on the stiffness matrix of the beam are investigated. Given a specific cut geometry and location, as the length of the beam increases, the effect of the cuts on all stiffness terms decreases. However, the relationships are not linear. The relationships between the reduced flange width and the stiffness terms are also nonlinear. As the length of the beam or the width of the reduced flanges increases, the reduction in all stiffness terms decreases at a decreasing rate. As the cuts are moved away from the ends of the beams, the reduction in shear and bending stiffness terms also decreases at a decreasing rate. The location of the reductions has no effect on the axial stiffness of the frame element. The results also show that as the length of the cuts is increased, the reduction in the shear and bending stiffness terms increases at a decreasing rate, and the reduction in axial stiffness increases at practically a constant rate. The results of parametric studies conducted on six-story two- and three-bay moment frames subjected to seismic base shears are also presented. Lateral story drift coefficients of moment frames with and without RBS beams are compared. The greatest reduction area investigated consisted of 40% maximum reductions in all the beam flanges. When compared to frames without cuts, these reductions resulted in a 10.6% increase in the story drift of the two-bay frame and a 10.3% increase in the story drift of the three-bay frame.

Under the action of wind, tall buildings oscillate simultaneously in the alongwind, acrosswind, and torsional directions. While the alongwind loads have been successfully treated using quasi-steady and strip theories in terms of gust loading factors, the acrosswind and torsional loads cannot be treated in this manner, since these loads cannot be related in a straightforward manner to the fluctuations in the approach flow. Accordingly, most current codes and standards provide little guidance for the acrosswind and torsional response. To fill this gap, a preliminary, interactive database of aerodynamic loads is presented, which can be accessed by any user with Microsoft Explorer at the URL address http://www.nd.edu/∼nathaz/. The database is comprised of high-frequency base balance measurements on a host of isolated tall building models. Combined with the analysis procedure provided, the nondimensional aerodynamic loads can be used to compute the wind-induced response of tall buildings. The influence of key parameters, such as the side ratio, aspect ratio, and turbulence characteristics for rectangular sections, is also discussed. The database and analysis procedure are viable candidates for possible inclusion as a design guide in the next generation of codes and standards.

This study combines a new hysteretic model for dynamic analysis of wood shear walls, ground motion suites, and an extreme value distribution in order to estimate the seismic reliability of a wood shear wall for varying levels of displacement at various sites around the U.S. Existing suites of ground motions for Los Angeles, Boston, and Seattle, believed to have the same probability of exceedance over a 50 year period, were used to introduce some level of uncertainty in the loading. Multiple wood shear wall experiments were performed at the Peter Grant Timber Engineering Laboratory at Michigan Technological University, Houghton, Michigan, to calibrate the new hysteretic model and introduce some level of uncertainty into the resistance, or hysteresis. Time domain analysis was used to simulate the seismic response of the shear wall. The peak drift of each time domain analysis was recorded and the data points fit to an extreme value distribution (Weibull). The probabilities of failure, and subsequently the reliability indices, were then easily determined from this distribution of the extremes, based on various limit states of excessive drift. The uncertainty in the hysteresis was found to have a negligible affect on the reliability estimates for the Boston site, but had a significant effect for the Los Angeles and Seattle sites. The reliability indices ranged from less than zero to over seven, depending on the site and limit state.