ssylvest
Structural
- Oct 12, 2010
- 15
We have a question about how to properly model a sheet pile wall for a total stress analysis, particularly where the tension zone is to be ignored.
We have a cantilever sheet pile wall with the soil at the same height on both sides that is to be designed for holding back a still pool of water for a limited durcation. When checknig the short-term condition, we are using the undrained laboratory strength properties. But there is some confusion about how to use the sheet pile program to do a total stress analysis (instead of effecive stress) for this case.
For the specific case of short-term loading on a cohesive soil (with cohesion but phi = 0), it might have been a moot point how the water was input to the program since Kactive and Kpassive are 1.0. However, the specific criteria we are working with (and the required program) is hard wired to ignore tension in the upper part of the active zone. Therefore, it seems like to correctly model total stress, the water table should not be defined on either side (or set below the sheet pile tip), soil should be input as gamma wet or gamma sat, and the load from the retained watrer on the flood side should be input as loads (vertical surcharge and a lateral varying load on the wall).
Otherwise, if the water table is input on both sides (to the ground surface on the protected side and to the full poll height on the flood side), the program uses the effective unit weight of the soil to divide into 2 x cohesion to arrive at a much larger tension zone to ignore, therefore giving a differnt overall result.
Any comments or references to clarify what is correct will be appreciated.
We have a cantilever sheet pile wall with the soil at the same height on both sides that is to be designed for holding back a still pool of water for a limited durcation. When checknig the short-term condition, we are using the undrained laboratory strength properties. But there is some confusion about how to use the sheet pile program to do a total stress analysis (instead of effecive stress) for this case.
For the specific case of short-term loading on a cohesive soil (with cohesion but phi = 0), it might have been a moot point how the water was input to the program since Kactive and Kpassive are 1.0. However, the specific criteria we are working with (and the required program) is hard wired to ignore tension in the upper part of the active zone. Therefore, it seems like to correctly model total stress, the water table should not be defined on either side (or set below the sheet pile tip), soil should be input as gamma wet or gamma sat, and the load from the retained watrer on the flood side should be input as loads (vertical surcharge and a lateral varying load on the wall).
Otherwise, if the water table is input on both sides (to the ground surface on the protected side and to the full poll height on the flood side), the program uses the effective unit weight of the soil to divide into 2 x cohesion to arrive at a much larger tension zone to ignore, therefore giving a differnt overall result.
Any comments or references to clarify what is correct will be appreciated.
