Finite Element Analyses and Cofferdam Excavations-PDF

PDF

Pdf Summary

The paper explains how finite element (FE) analysis can improve the design of cofferdam excavations compared with conventional methods based on assumed earth-pressure distributions. FE analysis can model soil–structure interaction and construction stages in a unified framework, making it particularly useful for deep or complex excavations, sensitive nearby structures, soft-clay basal stability, groundwater effects, consolidation, and three-dimensional behavior.<br /><br />Reliable results depend on selecting suitable soil models and parameters. The paper reviews elastic, elasto-plastic, nonlinear, and time-dependent behavior, emphasizing that soils respond differently to loading and unloading, drainage conditions, and loading direction. Common advanced models, such as the Duncan–Chang and Hardening Soil models, represent nonlinear stiffness and unloading/reloading behavior, but do not adequately capture strain softening or plastic strains from repeated load cycles. Soil parameters should be estimated using a combination of laboratory and in-situ testing; cone penetration testing calibrated with laboratory tests is favored, while SPT-based estimates are considered uncertain.<br /><br />The author recommends a careful modeling strategy: define an adequate model domain, use defensible soil properties, represent construction sequencing accurately, assess whether 3D effects matter, and conduct sensitivity studies. Results should be checked against conventional analyses and reviewed by an experienced engineer, since numerical output is approximate and strongly dependent on assumptions.<br /><br />A case study of a cofferdam near an earth dam illustrates these principles. Two-dimensional analyses predicted larger movements than monitoring recorded. The difference was attributed to ignored three-dimensional effects, cementation in the sands, and uncertainty about the continuity of a soft-clay layer. The analysis nevertheless informed design changes that reduced predicted movements and improved stability.<br /><br />The paper concludes that FE methods are valuable but not universally necessary. Their effectiveness depends on sound inputs and engineering judgment; continued work is needed on more realistic soil models and accessible 3D analysis.

Keywords

finite element analysis

cofferdam excavation

soil-structure interaction

construction staging

deep excavation

geotechnical modeling

soil constitutive models

groundwater effects

three-dimensional effects

sensitivity analysis