Civil Engineering March 2021 | Vol 29 No 2

Civil Engineering March 2021 65 which reduces the internal friction angle with appropriate partial factors. If grouted anchors are locked-off at values resulting from lower assumed internal friction angles in the baseline model, the pressure should not reduce at the facing. The PoF decreases to 0.051% and RI in- creases to 3.3 when considering the resulting increased pressure due to the soil structure interaction for the “φ– σ” analysis. SHOTCRETE INFILL PANELS/LAGGING The following question was asked: What pressure do you assume to act on the shotcrete infill panels/lagging and use in the structural design of the section? Three respondents indicated that they use triangular earth pressures, another three use apparent earth pressure diagrams with appropriate reduction for arching, and one uses pressure derived only from geotech- nical FE analyses. Caltrans recommends using 60% of the theoretical apparent earth pressure (AEP diagrams), with Douglas Fir timber infill panels/lagging. The difference in relative stiffness between the timber lagging and the steel piles, typically used in the USA, results in horizontal soil arching. Timber lagging thickness is normally chosen based on semi-empirical rules and tables such as the Goldberg-Zoino table. In South Africa, and other regions such as Australia, concrete piles are typically used with shotcrete infill panels and integrated into the permanent structure (Figure 9). Building slabs provide horizontal resistance after destressing of grouted anchors. Shotcrete infill panels are typically placed in South Africa, between piles, using an arch shaped section in line with the back of the piles. In some instances, the client might request linearly one-way spanning infill panels. Shotcrete is typically stiffer and placed thicker than timber with sufficient shotcrete cover to the welded wire mesh; therefore, less soil arching is expected. The authors of this article recommend that the designer assess the pressure (p) that develops using FE or BOE analyses, while comparing these to FHWA pressure using DA1-1 and DA1-2 (i.e. Ultimate Limit State (ULS)). The maximum pressure during all stages should be considered in designing the shotcrete infill panels (connected to piles with dowels or in some instances only in shear behind centreline of piles) for a bending moment of: M = pl 2 8 for linearly placed one-way spanning shotcrete infill panels. Arched shotcrete infill panels can be assessed using structural FE suites with p (ULS) acting normally on the arch. For FHWA stressed solutions, the pressure distribution that develops behind the wall would tend to be more rectangular and for a non-stressed solution (passive soil nails) the pressure developing will be more triangular (maximum at the bottom of the excavation). For the example above the unfactored pressure diagrams derived using the various constitutive models are provided in Figure 10. When modelling a one-way spanning infill panel in structural FE, moments as detailed in the formula above develop in the shotcrete infill panel (Figure 11a). When modelling a flat arch with earth pressure applied normal to the arch, the bending moment reduces significantly, and compres- sion load develops in the shotcrete infill panel (Figure 11b and c). In this scenario the bending moment in the arch is small enough to add only nominal steel in the arch while a double layer mesh is required in the one-way spanning shotcrete infill panel. In Australia, non-square rectangular design fabric meshes (rectangular) are typically used to increase the area of steel spanning between piles. The reaction forces at the pile positions, normal to the centreline of the wall, need to be checked and the requirement for dowels assessed. The displacement of the one-way spanning flat plate was modelled to be eight times larger than that of the arched shotcrete infill panel (0.4 mm for arched infill panel). This could possibly imply that a reduced earth pressure Figure 8 Effect of lower internal friction angle on pressure that develops over the wall’s face Wall movement / wall height (∆/H) –0.049 –0.025 –0.09 0.025 0 Dense sand, φ’ = 45°, K p = 0.17 MD sand, φ’ = 37°, K p = 4.0 Loose sand, φ’ = 30°, K p = 3.0 K 4.0 2.0 1.0 0.5 0.25 0 Dense sand, φ’ = 45°, K a = 0.17 MD sand, φ’ = 37°, K a = 0.25 Loose sand, φ’ = 30°, K a = 0.33 Lock-off at 80%–100% of FHWA Lock-off at 1.5T from wedge analysis K o = 0.50 K o = 0.40 K o = 0.29 ∆ Figure 9 Placement of mesh and shotcrete infill panels