Civil Engineering March 2021 | Vol 29 No 2

66 March 2021 Civil Engineering would act on a linearly one-way spanning infill panel, reducing structural forces due to soil arching effects. It is, however, advised that the designer should not reduce the pressure predicted due to soil arching as shotcrete is stiffer than timber lagging, unless this can be proved to be different by testing. This reduction could be confirmed on site by placing load-pressure cells behind the shotcrete infill panels. CONCLUSIONS AND SUMMARY This article forms the second part of a three-part series (Part 3 to be published in the April 2021 issue of Civil Engineering ) which summarises the results from a survey conducted to assess the current design prac- tice of multi-anchored walls in South Africa. Both BOE and geotechnical FE pack- ages are used by designers in South Africa to model multi-anchored pile walls. The designer should practice caution when using BOE packages that do not incorporate cumulative displacements due to staged excavations and use only linear springs. The passive resistance of largely spaced soldier pile walls reduces per meter length due to 3D effects; this reduction should be kept in mind when designing these walls. The HS and, in particular, the HS-small constitutive models are better suited to modelling embedded pile walls as shear displacement is dominant, unload-reload behaviour important, and increased soil stiffness is modelled with depth due to increased confinement. Displacements are best predicted using geotechnical FE analyses but require suffi- cient information on the soil stiffness from site and laboratory testing. Probabilistic design of multi-anchored pile walls is complex as structural elements and grout- ground bond also come into play. Limit equilibrium analyses can be undertaken by applying stress on a vertical face. This stress should be equivalent to the applied prestress but could increase should the internal friction angle be lower than that assumed in the baseline analysis. Flat arched shotcrete infill panels generally develop lower bending moments compared to a linear one-way spanning shotcrete infill panel. This typically results in reduced mesh reinforcement requirements in the shotcrete section. The pressure assumed to act on the shotcrete infill panel should be the maximum of the FHWA pressure (if pre- stressed), geotechnical FE and BOE analyses. ACKNOWLEDGEMENTS The authors would like to thank Fernando Pequenino for allowing access to DeepEx, Rocscience for the provision of trial software versions, and Andries van der Merwe from Naidu Consulting for his assistance. NOTE Reference details are available from the authors. Z Y X 15 129.38 14 129.96 13 129.92 12 129.03 11 129.00 10 129.90 9 129.97 8 129.97 7 129.90 6 120.03 5 120.01 4 128.09 3 120.96 2 128.38 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 X Z –2.90 –3.66 –3.79 –4.56 –3.02 –3.66 –3.82 –4.07 –3.99 –3.00 –3.79 –3.80 –3.90 Y Y Z 1 2 3 4 5 6 7 8 9 10 11 12 13 14 X 15 –7.95 –14.58 –20.05 –24.30 –27.34 –29.16 –29.77 –29.16 –27.34 –24.30 –20.05 –14.58 –7.90 Figure 11 Prokon Frame Analysis modelling a linearly one-way spanning (a) and arch shotcrete infill panel (b and c) (unfactored loading used) Figure 10 Pressure developing behind soldier pile wall/shotcrete infill panels using various methods (to be factored for ULS) Depth (m) Shotcrete pressure (kPa) 0 10 30 40 50 60 14 12 10 8 6 4 2 0 20 Mohr Coulomb HS HS Small FHWA