Search

Today, culverts are often analyzed and designed using plane strain finite element analyses in which a transverse slice of the structure is simulated and assumed equal to any other slice. Although this works well for soil loads, in reality live loads are discrete pressure loads from tires, and the out-of-plane effects can be significant, especially when cover depths are relatively low. While current design specifications define minimum cover that precludes significant localized pressure effects on culverts, many culverts are currently in service that were constructed under very shallow cover depths. In this study, analyses of corrugated metal culverts under truck load were carried out using two-dimensional (2D) (plane strain) and three-dimensional (3D) finite element analyses to assess the difference between structural responses predicted by these two approaches. Thirty-six culvert models with different shapes, spans, and cover depths were created using both 2D and 3D meshes. A significant difference was found between the results of these two analysis methods. Specifically, the thrust from the 3D models under shallow cover can be as much as three times the thrust results from the 2D model, which is unconservative from a safety standpoint. An innovative explanation was provided to explicate the differences. Finally, to address the issue, a reduction ratio method was developed to correct the 2D results to match the corresponding 3D results so that the relatively simple plane strain approach can be used.

Departments of transportation are required to inspect and rate bridges that span more than 20 ft (6.1 m). One frequent challenge is rating shallow-buried corrugated metal culverts, since there are no standard methodologies deemed applicable to all sections and depths for defining the resistance when moment or combined moment and thrust controls. The commonly used Culvert ANalysis and DEsign (CANDE) finite element software was developed specifically for culvert design; its companion program, CANDE Tool Box, enables computing of load rating factors. However, owing to nonlinearities associated with soil–structure interaction and yielding of the corrugated cross-section, extracting accurate representations of available resistance for shallow-buried corrugated metal culverts may require a cumbersome iterative procedure. Furthermore, the resulting rating factor from not properly implementing nonlinear procedures may indicate an available resistance that is unconservative from a safety perspective. Recent improvements adopted by AASHTO and incorporated into CANDE include combined thrust–moment criterion, but are deemed suitable only for culverts with deep corrugation cross-sections. The objective of this research was therefore to assess flexure criteria for load rating shallow-buried corrugated metal pipe and arch structures, including standard corrugation cross-sections, to simplify the process without significantly sacrificing accuracy. Challenges with available methods and recently implemented combined criterion for the shallow-buried culvert scenario are first discussed. A modified flexural capacity criterion based on load and resistance factor rating is then developed and presented. Finally, the accuracy and applicability of the methodologies are assessed using a range of practical corrugated metal culvert dimensions and parameters.

A stratum of cemented sand can cause numerous geotechnical design problems. Extensive research has been conducted to determine both the properties of cemented sand and their effects on the engineering response of soil. A cone penetration test module was developed at the University of Florida to detect the presence and strength of these materials. After design and construction were complete, the prototype module was field tested in sand without cementation and again at a site with possible cementation. Development, calibration, and field testing of the module are discussed, as are preliminary recommendations for its potential use.