In the recent Corpus Christi bridge delay issues, it was claimed that the initial analysis methods did not consider the pipe cap stiffness, and this led to underestimation of the maximum pile cap load. To prove this effect out, we prepared a conceptually similar pile cap model and analyzed it with a PY springs and 3D finite element methods. To keep things simple a single load was assumed at the center of the 18ft (6m) thick pile cap. Figures 1 and 2 present the two models for PY and FEM respectively as prepared with our DeepFND software.
In the PY model, 3D axial effects are incorporated with conventional reduction factors on the pile resistance. In the FEM method such effects are computed automatically. Figures 3 and 4 compare the computed pile axial reactions for the PY and FEM models respectively.
Both models confirm that pile reactions are not equally distributed and that the pile cap stiffness plays a role. The PY model produces greater axial reactions at corner piles and smaller reactions at center piles in this case vs. the FEM analysis. This is most likely because of how group interaction factors are imposed in the PY method that typically reflect near ultimate loading conditions.
The bottom line is that you always need to consider pile cap stiffness and pile group interaction effects.
Figure 1: PY 3D Pile Cap Model
Figure 2: 3D FEM Pile Cap Model
Figure 3: Vertical Pile Reactions from PY Model
Figure 4: Vertical Pile Reactions from FEM Model