Editor’s Note

The special section of this issue of the Journal of Structural Engineering contains expanded versions of seven of the papers originally presented at the 17th Analysis and Computation Specialty Conference, held in St. Louis, Missouri, May 18–21, 2006. The 17th A&C Conference continued a 50-year tradition of specialty conferences that are dedicated to advances in computer-based structural analysis and design. The first such conference, called the Conference on Electronic Computing, was held in Kansas City in 1958. Since then, conferences have been held at separate venues on a regular basis (with separate proceedings), continuing to 1996, when the 12th A&C Conference was held in conjunction with the ASCE/SEI Structures Congress. Starting in 1998, the A&C Conference was merged into the Structures Congress program, and the proceedings were published together with the Structures Congress proceedings. This trend led to a loss of identity for the A&C Conference, which by 2004 had virtually disappeared from the congress program.

In 2005, the Analysis and Computation Technical Committee, led by Don Grierson, decided to resurrect the A&C Conference by giving it a separate identity within the Structures Congress program, and by improving the quality of the conference by requiring all accepted papers to be peer reviewed. Further, it was decided that authors of the highest-rated accepted papers would be invited to submit expanded versions suitable for publication in the Journal of Structural Engineering. This process, which was first used for the 17th A&C Conference, produced 44 accepted conference papers, the submission of 12 papers for publication in the JSE, and ultimately the 7 papers presented herein. This process is being continued for the 18th Analysis and Computation Conference, which will be held in conjunction with the 2008 Structures Congress.

The special section begins with two papers that concentrate on the use of hybrid analytical-experimental simulation. In hybrid simulation, a physical laboratory specimen is tested pseudodynamically. The loading protocol for the specimen is controlled in part by an analytical simulation of a larger portion of the structure. Hybrid testing is a major component of the National Science Foundation’s Network for Earthquake Engineering Simulation (NEES) program. The first hybrid simulation paper, “Large-Scale Experimental Verification of Semiactive Control through Real-Time Hybrid Simulation” by Christenson et al., reports on the use of the Fast Hybrid Test Facility at the University of Colorado at Boulder. In this pre-NEES project, the physical component of the test structure consisted of a large-scale semiactive magneto-rheological fluid damper, and the analytical structure was a one-story, one-bay building with nonlinear material modeling at the beam-column joints. Open System for Earthquake Engineering Simulation (OpenSees) software was used in the analytical simulation. Both the experimental and the analytical components of the test were performed at Boulder. Several components of the NEES infrastructure allowed teleparticipation at the University of Connecticut, more than $\mathrm{2,000}\mathrm{mi}$ from the test facility.

In the second paper “Hybrid Seismic Response Simulation on a Geographically Distributed Bridge Model,” Mosqueda et al. discuss hybrid simulation in which multiple analytical and physical models, were used in geographically distributed locations. In this example, a six-span bridge was investigated; five NEES sites were involved, two of which were experimental. A key feature of the research was the development of a distributed control strategy that was implemented into the NEESgrid. This strategy allowed pseudodynamic testing at $0.66\mathrm{s}$ per integration step, which was a substantial improvement over previous efforts.

Petcherdchoo, Neves, and Frangopol develop an optimum maintenance strategy for deteriorating bridge structures in “Optimizing Lifetime Condition and Reliability of Deteriorating Structures with Emphasis on Bridges.” Performance is measured in terms of condition, reliability, and cumulative cost. The procedure is capable of incorporating time-based, performance-based, and a combination of time- and performance-based strategies. The methodology is illustrated with maintenance data collected for an existing slab and girder bridge structure.

The next paper, “Improving Knowledge of Structural System Behavior through Multiple Models” by Smith and Saitta, presents a multimodel system identification methodology that accounts for factors influencing the dependability of identification. Data mining and its usefulness for system identification are also introduced. A study of the applicability of the multimodel approach is explained with a case study.

Scott et al. present a series of software design abstractions for use in the development of a material nonlinear and geometrically nonlinear frame element in their paper, “Software Patterns for Nonlinear Beam-Column Models.” These abstractions are key to the creation of next-generation object-oriented structural analysis programs. The element, implemented in OpenSees, uses a highly efficient force-based formulation rather than the more traditional displacement-based finite-element formulation. The element’s capabilities are demonstrated through the analysis of a reinforced concrete bridge pier.

The sixth paper, by Machado, Sotelino, and Liu, addresses a “Modeling Technique for Honeycomb FRP Deck Bridges via Finite Elements.” The motivation of the work was to provide improved vibration serviceability criteria for highway bridge rehabilitation. In the proposed modeling approach, implemented in ANSYS software, the deck is represented by a simplified model consisting of a series of boxes modeled with four-node shell elements, with six degrees of freedom per node. The paper presents mechanics-based equations used for setting the properties of the shell elements. The simplified model was validated by comparing the computed response with the measured behavior of an existing bridge.

The final paper of the special section, “Unintended Consequences of Modeling Damping in Structures” by Charney, discusses a variety of implications of the use of mass and stiffness proportional damping in nonlinear dynamic structural analysis. The effects of local and global inelastic behavior are investigated, and it is shown that artificial damping will be generated when the damping matrix is based on initial mass and stiffness. In some cases, the artificial damping leads to computed results that appear reasonable, but upon further investigation are completely invalid. The basic conclusion of the paper is that mass and stiffness proportional damping should be used with extreme caution and, if possible, avoided in favor of more realistic, nonviscous, hysteretic damping models.

The remaining papers selected for this issue complement the special section. Five additional papers that deal with the theme of analysis and computation follow in the next section, while the remaining four papers address issues related to seismic effects on structures.

Vo et al. present findings from “Buckling Analysis of Moderately Thick Rotational Shells under Uniform Pressure Using the Ritz Method.” In applying the Ritz method, displacement components of the shell are approximated by the product of 1D polynomial functions, and the boundary equations are suitably adjusted to ensure the satisfaction of geometric boundary conditions. The validity of the method and convergence and accuracy of the solutions are demonstrated using spherical shells. A parametric study is conducted on spherical and parabolic shells, considering the effects of height-to-base-radius ratios, thickness-to-radius ratios, and different support conditions on the buckling solutions.

“Analytical Approximations to Large Hygrothermal Buckling Deformation of a Beam” are developed by Yu et al. The solutions are established by combining Newton’s method with the method of harmonic balance. In contrast to the classical method of harmonic balance, the linearization is performed prior to proceeding with harmonic balancing, thus resulting in a set of linear algebraic equations instead of nonlinear ones. The proposed approximate solutions show excellent agreement with the exact solution and are valid for a small as well as a large angle of rotation at the end of the beam.

In “Explicit Inelastic Stiffness for Beam Elements with Uniform and Nonuniform Cross Sections,” Leu et al. propose a quasi-plastic-hinge approach in an incremental form, wherein the elastic-plastic flexibility coefficients are explicitly given. The procedure is directly extended to nonuniform beams through the use of different moment-curvature relations for the two ends of the beam, or indirectly through subdivision of the beam into a number of inner segments. Three steel frames composed of nonuniform cross sections are studied to validate the accuracy of the proposed approaches.

A “Model for Cyclic Inelastic Buckling of Steel Braces” is developed by Uriz et al. The model consists of a force-based frame element with distributed inelasticity and fiber discretization of the cross section. Even though the element only accounts for small deformations in the basic system, a corotational formulation enables consideration of large displacement geometry. It is shown that two elements for each brace are sufficient to yield results that match an extensive set of experimental data for braces with different cross sections and slenderness ratios. A simple smeared-crack, one-parameter model that describes the triaxial stress-strain behavior of concrete is extended by Lykidis and Spiliopoulos to enable “3D Solid Finite-Element Analysis of Cyclically Loaded RC Structures Allowing Embedded Reinforcement Slippage.” The proposed procedure is applied successfully in a long anchorage rebar test as well as two cases of bond-critical exterior and interior beam-column joints, and numerical results are shown to compare well with available experimental data.

In “Seismic Fragility and Confidence Bounds for Gravity Load Designed Reinforced Concrete Frames of Varying Height” by Ramamoorthy et al., five structures are used to represent generic RC frame buildings 1 to 10 stories tall. For each building, fragility estimates are obtained by assessing the conditional probability that the drift demand reaches or exceeds the drift capacity for a given earthquake spectral acceleration. Confidence bounds are developed to represent the epistemic uncertainty inherent in the demand models used in the fragility estimates. Bivariate fragility estimates, formulated as a function of spectral acceleration and the fundamental building period, are developed from the fragility estimates of the individual buildings.

Kwon and Elnashai introduce a “Seismic Analysis of Meloland Road Overcrossing Using Multiplatform Simulation Software Including SSI.” The embankments, abutments, and pile groups are modeled using finite-element software that contains an appropriate nonlinear material model for the soil, while the bridge is modeled using a different software that incorporates fiber-based frame sections. The dynamic properties of the model are compared with those evaluated from system identification techniques using recorded ground motion. Results from response history analyses of the system using recorded free-field motions compare well with measured data.

In “Effect of Building Nonlinearity on Seismic Response of Nonstructural Components: A Parametric Study,” an extensive parametric study is conducted by Chaudhuri and Villaverde on eight code-designed steel moment-resisting frames and a series of linear and nonlinear single-degree-of-freedom nonstructural components. It is found that, in general, the nonlinear behavior of the supporting structures reduces the seismic response of the nonstructural components in comparison with the linear counterparts. In a few cases, however, the nonstructural component response is amplified by a factor that can be as large as 5.2. These cases correspond to components located on the lower building floors, with a natural period equal to the second or third natural period of the building and subjected to a narrow-band excitation with a dominant period close to the building’s fundamental natural period.

Results from a combined experimental and analytical investigation are presented by Chang et al. in “Shaking Table Study on Displacement-Based Design for Seismic Retrofit of Existing Buildings Using Nonlinear Viscous Dampers.” It is shown that the addition of nonlinear viscous dampers to the structure results in displacement and force reduction by about 68 to 80% under El Centro earthquake (magnitude scaled by a factor of 0.3) and that higher-mode response is significantly diminished. The displacement-based evaluation procedure adopted in the study tends to underestimate the responses of the damped structure in the elastic range and overestimate them in the inelastic range.

Also included in this issue is a discussion by Rovňák et al. of a paper by Dezi et al. that appeared in the September 2006 issue of the journal. The discussers draw attention to conditions under which expressions derived by other researchers for estimation of the effective width can be applied for prestressed composite girders without FEM analysis. They also raise questions on some of the conclusions presented in the paper. The writers of the original paper demonstrate that the simplified formulas presented by the discussers are unable to depict the phenomenon studied in the paper with the same level of accuracy. Additionally, they point to the need to consider shear lag in the cable anchorage region and disagree with the discussers on the issue of superimposing dead-load effects.