Civil Engineering April 2021 | Vol 29 No 3

46 April 2021 Civil Engineering engineers in South Africa. Seismic loading does occur throughout South Africa, with the peak ground acceleration with a 10% probability of exceedance in 50 years provided in Figure 10. One respondent indicated that they do not allow for seismic loading on walls. Another only allows for seismic loading on permanent walls, while the others allow for horizontal acceleration in FE analysis and a sixth always for horizontal and vertical acceleration of 0.1 g. Normally, a seismic coefficient of between 0.66 and 1.0 times the PGA should be used in analyses. Vertical acceleration can normally be ignored in the design of these structures as vertical loads are not considered to add significant loads to grouted anchors (FHWA, 1999). Pseudo-static methods such as Mononobe and Okabe’s (Okabe, 1926) reverse trapezoidal pressures or Wood’s rectangular pressure (Wood and Elms, 1990) can be used in Beam on Elastic Winkler spring analyses. The walls should be sufficiently designed to prevent ultimate limit state failure, even in a temporary application. LIMITING CRACK WIDTHS IN CONCRETE PILE SECTION The following question was asked to respondents: Do you design pile sections to limit crack widths? Four respondents indicated that they do not allow for crack control of concrete piles while two others stated that they do if requested by the client. SANS10100 calls for crack width control to 0.3 mm or 0.004 times the nominal cover in highly aggressive environments. This serviceability design (i.e. crack width design) of the permanent pile section would typically result in an increased requirement for reinforcement compared to the ultimate limit state design (M ULS = M WSD × FoS = 1.5M WSD ) of roughly 25%. Using EN1997-1 DA1-1 and DA1-2 would typically result in a smaller percentage increase in reinforcement to limit crack width than 25%. This is due to FoS typically being larger than 1.5 resulting using partial factors (van der Merwe & Chang, 2022). Increasing steel quantity to limit crack widths for the quasi-permanent SLS load situation is required according to code but may result in congested steel and there- fore requires careful consideration. For water retaining structures and to prevent water seeping through the face tighter control on crack widths would be required (typically 0.2 mm). FAILURES/LESSONS LEARNED The following question was asked to respondents: Have you had any major disputes or failures during construction of deep basement excavations? All respondents, except for one, indicated that they have had failures on walls designed by themselves or indirectly through their involvement in dispute resolutions on lateral support contracts. To summarise, the following disputes and failures were noted: Q Q Over-dig of excavations, resulting in increased displacements of a cantilever wall. Q Q Over-excavation during staged excava- tion before installing adequate grouted anchors at the level above. Such move- ments are difficult to reverse. Q Q Collapse due to excessive blasting and overbreak (charges exceeding those allowed and not adhering to minimum blast distance). Q Q Collapse due to inadequate manage- ment of stormwater runoff above the lateral support wall. Q Q Insufficient investigation not ex- tending deep enough and the use of incorrect investigation methodologies. Q Q Vibration damage to adjacent build- ings using high vibratory pile instal- lation methods in a saturated sandy profile. Q Q Flow of soils between soldier piles in saturated sandy profile undermining neighbouring properties. Q Q Incorrect grouted anchor stressing procedure followed, resulting in eccen- tric stress distribution in the strands. Q Q Incorrect grouting procedure followed for soil nails with grouting from top of hole and not bottom with no grout placed over the lower half. CONCLUSION AND SUMMARY The article forms part of a series of three articles* which summarises the results from a survey conducted to assess the current design practice of multi-anchored pile walls in South Africa. SDAs are used in South Africa as both soil nails (hollow bar soil nails and stressed grouted anchors). These bars cannot be provided with double corrosion protection measures and therefore cannot be used in permanent applications when stressed according to EN1537. The use of non-stressed SDAs or soil nails in embedded pile walls will necessitate a decrease in pile spacing (i.e. increased wall stiffness) when compared to a prestressed solution, to limit displacement. The pressure distribution resulting from a non-stressed solution is more triangular and will have an effect on the structural forces that develop in both the piles and shotcrete infill panels. The use of SDAs as permanent nails is questionable when allowing for grout cover as grout body quality is not guaranteed. When designing with sacrificial thickness allowances inside the mating coupler, the couplers’ ability to transfer load if corroded is considered potentially problematic. Additionally, corrosion of any bar in coastal areas in permanent application, without DCP at the facing, is considered highly problematic. Further research is required to assess the effect of corrosion on mating coupler connections by undertaking accel- erated corrosion tests and tensile load tests at various stages of corrosion. Seismic loading and limiting crack widths in concrete pile sections are often not considered in the design of these walls but should be considered to comply with the applicable codes.  ACKNOWLEDGEMENTS The authors would like to thank Fernando Pequenino for allowing access to DeepEx and Rocscience for access to trial versions of RS2. *Part 1 of this series was published in the April 2020 issue of Civil Engineering , Vol.28 No.3. Part 2 of this series was published in the March 2021 issue of Civil Engineering , Vol.29 No.2. NOTES The concerns raised in this article are solely that of the authors and raised to seek clarification. Suppliers and industry are encouraged to participate by responding to this article, particularly the concerns raised on coupler effectiveness due to the effect of corrosion inside the connecting couplers on SDA bars. For a full list of references please contact the authors