Michael G Katona

Modernize and Upgrade CANDE for Analysis and LRFD Design of Buried Structures

This report documents research performed to develop, modernize, and upgrade CANDE (Culvert ANalysis and DEsign). The new version is called CANDE-2007. The report details the research performed to update the program. CANDE installation files are included on a CD-ROM (CRP-CD-69) with this report. The installed program includes integrated help files and 14 tutorial examples. The report and software will be of immediate interest to culvert designers.

Guideline for Interpreting AASHTO Specifications to Design or Evaluate Buried Structures with Comprehensive Solution Methods

Design engineers and consultants are required to satisfy the criteria set forth in the AASHTO LRFD Bridge Design Specifications when designing buried culverts within the national transportation infrastructure. Adherence to the specifications is especially difficult when an engineer uses a comprehensive solution method such as a finite element program to predict the deformations and internal forces of the culvert-soil system. Part of the difficulty is in knowing how to apply factored loads as well as to determine their values. A more difficult problem is to assess whether the culvert is structurally safe according to the AASHTO LRFD specifications. Said another way, the design criteria and the corresponding factored capacities as written in the specifications are less than well defined. A succinct interpretation is presented of the AASHTO specification for load and resistance factor design (LRFD) for buried structures with regard to factored loads and design criteria. Tabularized values are provided for load factors and load modifiers to quantify dead loads, earth loads, and live loads, which are required for a LRFD analysis to produce the factored demands. Also presented are tabularized descriptions of the design criteria and the corresponding factored capacities, which are defined and quantified for corrugated metal, reinforced concrete, and plastic pipe materials. For some culverts, the design criteria have been generalized to provide sufficient information that is compatible with comprehensive analysis methods. Although intended especially for design engineers who use comprehensive analysis tools, the information presented is also useful and educational for engineers who use the simplified design solutions in the AASHTO LRFD specifications.

Modifying Duncan–Selig Soil Model for Plasticlike Behavior

The variable-modulus Duncan–Selig soil model is very effective in capturing the nonlinear stress–strain behavior of most soils in loading environments, including the softening effect from increased shear stress as well as the stiffening effect from increased confining pressure. Accordingly, the Duncan–Selig model is used in many geotechnical applications, such as construction of culvert installations with incremental layers of soil, for which soil loading is the dominant condition. However, the Duncan–Selig model is a nonlinear elastic formulation that retraces the same stress–strain path upon unloading. Consequently, the model does not predict residual deformation upon unloading. In contrast, plasticity-based soil models are generally less accurate in simulating the nonlinear loading response of soil samples; however, they inherently include permanent deformation if plastic deformation occurs during loading. This paper introduces modifications to the Duncan–Selig model that result in permanent deformations upon unloading similar to those of advanced plasticity models. Although the modifications are based on concepts from incremental plasticity theory, the modified Duncan–Selig model remains a variable-modulus formulation without the need of defining any additional model parameters or evoking a plastic-flow rule. Thus, the large existing database of Duncan–Selig parameters remains valid for the modified formulation. Most important, the modified Duncan–Selig model is shown to satisfy thermodynamic and continuity requirements and to compare very favorably with triaxial and hydrostatic laboratory test data under load–unload–reload conditions. Finally, the modified Duncan–Selig model is used to simulate the effects of soil compaction for an installation of a long-span corrugated-steel culvert.

Seismic Design and Analysis of Buried Structures with CANDE-2007

A method is presented to analyze and evaluate buried culverts and cut-and-cover tunnels for seismic loading in addition to standard static loading from dead and live loads. With CANDE-2007, a plane–strain finite element program, the soil–structure problem is characterized by a cross-sectional slice through the structure and surrounding soil envelope. First, the static design loads are applied with a series of incremental load steps. Next, the seismic loading condition is simulated by specifying quasi-static displacements at the peripheral boundaries of the soil envelope to produce a shear-racking distortion equivalent to the maximum free-field seismic shear strain from the design earthquake. The proposed method is based on two recently completed NCHRP projects and is presented here in detail. An initial linear-elastic study investigates two basic culvert shapes, circular and rectangular, wherein moment and thrust distributions from CANDE-2007 are shown to compare favorably with closed-form solutions. A second study investigates a nonlinear reinforced concrete box culvert installation under the combination of static and seismic loading. Plots of moment, thrust, and shear distributions show how and where the seismic loading alters the response of static loading, including the effect on safety factors for steel yielding, concrete crushing, and shear failure. It is concluded that the proposed seismic method is rational, easy to apply, and fulfills an engineering need heretofore unfulfilled. The procedure applies to any culvert shape, size, and material, and the safety of the culvert installation may be assessed by working either stress or load and resistance factor design.

Continuous Load Scaling

Because of their prismatic configuration, buried culverts are often designed and analyzed as two-dimensional (2-D) plane strain soil structure systems, as exemplified by the Culvert Analysis and Design (CANDE) finite element program. One major difficulty with 2-D analysis is that live loads are not infinitely long prismatic strips but are finite footprints that induce load spreading through soil in the in-plane and out-of-plane directions. Because traditional 2-D analysis permits only in-plane load spreading, the predicted soil stresses are increasingly overestimated as soil depth increases. The traditional approach to correcting the overestimate is to reduce judiciously the 2-D strip load; this process is called the reduced surface load (RSL) procedure. The reduction factor is dependent on an assumed load-spreading theory and a selected soil depth (H). RSL analysis results in soil stresses that are correct at the selected soil depth (H) but understressed for depths less than H and overstressed for depths greater than H. The new continuous load scaling (CLS) procedure introduced in this paper overcomes such shortcomings by continually diverting live load in the longitudinal direction as a continuous function of soil depth. This diversion is achieved by continuously increasing the plane strain unit thickness as a function of depth on the basis of an assumed load-spreading theory. CANDE solutions were obtained for both RSL and CLS procedures and compared with three-dimensional solutions for free-field conditions and for flexible and rigid culvert installations. The bottom-line finding is that the CLS procedure with the elasticity-based load-spreading method is far superior to all other 2-D methods.

Fluid Jackets: New Concept to Enhance Structural Performance of Buried Concrete Pipes

To surround the circumference of plain concrete pipes with fluid jackets is a revolutionary concept that can result in maximum burial depths and live load capacities more than twice those of conventional, steel-reinforced concrete pipes. Moreover, the enhanced capacities are achievable with less than one-half the usual amount of concrete and with zero steel reinforcement. Alternatively, fluid jacketed concrete pipes may be designed with the same structural capacity as conventional reinforced concrete pipes but with a project cost savings that results from a 50% weight reduction per foot of pipe. These claims are supported by the analysis presented in this paper, which is a necessary first step before experimental validation can occur. Engineered plastic jackets encapsulate fluid (water) in a watertight configuration that encircles the buried concrete pipe. The encapsulated fluid transforms nonuniform soil pressure to uniform hydrostatic pressure around the pipe periphery. Consequently, the concrete pipe experiences hoop compression only, without the bending deformation and concrete cracking inherent in conventionally installed, reinforced concrete pipe. The fluid jacket concept has not been published in open literature. This paper illustrates the physics of the concept and presents the most promising configuration and materials determined through structural analysis studies with the CANDE-2007 computer program. Fluid jacket designs are proposed in this paper as a first step to bring this revolutionary idea to fruition. Next steps include laboratory testing, field testing, and modeling refinement.