Development of Design Properties for the Seawall-Sheet pile

FEA Calculation
A FEA model was created using SolidWorks Simulation that mimics the full section test. The model uses shell elements at the neutral axis of the laminate with beam elements to represent the heavy interlock feature. A single panel was modeled with boundary conditions applied to represent a continuous wall. Linear static stress and Eigen value buckling analyses were perform using the model. A load of 1000 lb was applied to the model at the wale location and linear scaling was performed for comparison to the pre-Standard analysis and test results.

Difficulty occurs in the modeling process in representing the panel interlock feature. The interlock feature has a measurable amount of clearance in the interface, which will allow for some rotation at the interface.

The Eigen value analysis was solved using two different sets of boundary conditions. One analysis was performed with the longitudinal rotation fixed at the interlock location. A second run was performed with the longitudinal rotation free. Both predict local buckling in the compression flange, but with slightly different modal shapes. The Eigen values predicted by the model are 2.8149 and 2.4939 for the rotation fixed and rotation free runs, respectively. These values correlate to an applied load of 11,260 lb and 9.976 lb for the full section test.
The linear static analysis predicted a displacement at the bottom edge of the waler of 3.107” for the 1000 lb load. This correlates to a displacement at the predicted failure loads of 8.75” and 7.75” for the rotation fixed and rotation free models, respectively. The peak compression stress predicted by the model is 12,044 psi. This correlates to a factor of safety of 5.35, which suggests the part will buckle prior to material rupture.

Experimental Testing
Full scale testing was conducted to determine the accuracy of the analysis. The test was performed using four pieces of the profile interlocked to make a structure 72” wide. The panels were 18ft long and embedded in compacted gravel at a depth of approximately 6ft. Wale sections were mounted to the panel at 9ft-7 in above the ground line to create a rigid location to apply load and prevent distortion of the wall section. Load was applied by running cables from the wale section to a backhoe bucket. West Virginia University Constructed Facilities Center was on hand during the test to perform the instrumentation and data recording and to oversee the testing.

A buckling failure of the panel occurred at a load of 10,374 lb. This value correlates to a ground line moment of 16,569 ft-lb/ft of wall and a flange compression stress of 15,201 psi. The deflection at the wale location was 18.7 inches at failure. It was observed the compacted gravel gave way on the front side of the panel resulting in a 1-2 inch gap on the back of the panels. This soil displacement may have contributed to additional measured deflection of 8.5inches resulting from rigid body motion. Strain gages mounted on the panels indicate the section performed as a well-behaved cantilever beam with uniform load distribution across the width of the structure.

The Pre-Standard analysis results in a conservative prediction of failure for the 1580 Series sheet pile. The value prediction of failure for the 1580 Series sheet pile. The values predicted is 15.5% less than the full section test result. This error is easily explainable with the conservative assumptions made for the Pre-Standard Calculations. In particular, the material property values were assumed to be the minimum value from the coupon level testing. If these values are changed to median values of the test results, the Pre-standard is only 4% conservative.

The FEA showed good correlation with the full section test results. The two boundary conditions bounded the test results with the rotation fixed analysis over predicting the test result by 8.5% and the rotation free analysis under predicting the test result by 3.8%.
The deflection prediction did not correlate with the full section testing as well as the failure prediction. The boundary condition of the soil is the most likely cause of the poor prediction. The soil provides an elastic foundation which allowed the panels to move. The Pre-Standard calculation was performed as a cantilever beam, which assumes no displacement at or below the ground line. The FEA was performed with the normal vector of displacement fixed for each face. This allows for some shear deflection below ground line but no bending displacement.

Material strength comparisons are not sufficient to determine the capacity of the FRP sheet pile. Buckling calculations must be performed to properly predict the failure of these thin wall, low modulus structural profiles. This paper shows the Pre-Standard buckling calculations or FEA are both options for determining conservation strength capacities of the assembled structure.

This article comes from CP edit released