Dear Adonis,
I am sending you an attached file with three folders: ADO, PLX and PNT.
In the ADO folder, there is an example of a retaining wall calculated with Adonis.
In the PLX folder, there are the graphical results of the same problem calculated with Plaxis8.
In the PNT folder, there are the graphical results of the same problem calculated with the PANT program, based on elastoplastic springs, that I developed myself and that can be seen on my website: https://cqtpya.wixsite.com/retaining.
My query is because the results of terms of ground pressures and forces in anchors, agree sensibly. But where they do not coincide is in the displacements or in the bending moments of the wall.
Plaxis and PANT, make sense, because as you excavate, the greatest displacements happen to be in the middle of the excavated height of the retaining wall, but Adonis does not coincide and predicts that the greatest displacements always occur in the upper part of the wall.
I would appreciate if you can analyze this problem and tell me if I am making a mistake in entering the data in Adonis.
Greetings
Nestor

Hi Nestor,

thanks for the comments and the examples. unfortunately I don't have plaxis or the other program to make the comparison but by simply going through your script realized that length of grouted tiebacks are only 1% with low sbond and kbond comparing with pretens force and this is not enough to hold back the retaining wall and my guess is that the tiebacks are keep moving with the wall and that's why you are getting such displacement on top. I'm no sure the reason for using tiebacks for this case but it does not seem to be working.

in plaxis, did you use exact same element with 1% grouted length and the same kbond and sbond?

unfortunately I don't have the whole picture from plaxis such as mesh size, type of elements to make a fare judgment.

Hi Roozbeh,
In the Adonis model, to avoid interface problems, the anchor goes from the retaining wall to the end of the model, which is on the border, which is a fixed node. In this way I can accurately establish the stiffness of the anchor which is ExA / L. That is why the grouted length is 1%, because it does not act, the anchor is fixed to the right end of the model and does not move dragged. What is more, the force absorbed by the anchor turns out to be the same as absorbed by the anchor in the other two models (Plaxis and Pant) and the active pressure is also the same. Only the moments in the structure and the maximum displacement zone in the middle sector of the wall do not coincide.
In the case of the plaxis anchor, I use anchors that behave the same as the one I put in the Adonis model, establishing their axial stiffness, avoiding interface problems.
Greetings
Nestor

even when you put the end of anchor at fixed point it does not guarantee fixed anchor. the anchorage between tieback element and soil element is simply coming from coupled spring which is controlled by kbond ans sbond value. link below explains the formulation used in cable/tieback elements.

http://docs.itascacg.com/flac3d700/common/sel/doc/manual/sel_manual/cables/cables.html?node852#mechanical-behavior

also equal force in the tieback does not mean that the tieback is behaving similar comparing with plaxis. mesh size and level of displacement has major impact on the forces in the tieback/cable elements.

I still believe you are not comparing the same models. If I have time I'll be able to make the comparison with FLAC which is available for me. but from displacement pattern the moment force in the beam elements look reasonable in ADONIS.

instead of tieback you can use free beam element like struts and put it in front and fix the other end to make sure fixed node. tieback is not a good choice.

Thanks Roozbeh for your answer,
I will try to install bar elements on the excavation side as you indicate, the query would be using this bar element, how can I apply the prestress load?
Thanks for your help
I will try if I can make the same model in Flac
Greetings
Nestor

Hello All,

I found this page when I was looking for a program that allowed me to verify a retaining wall with a tieback. This seems to be exactly the case. However, before I get into learning how to use the model I'd like to check if the model allows me to compute safety factors for sliding, bearing and overturning.

Please, could you tell me if this is possible. I have to compare this factors with the ones of an old verification.

Thanks a lot.

Jorge

factor of safety calculation is formulated based on shear strength reduction (SSR) method so it could be used for any problem regardless. Please note that FOS calculation is only activated for solid elements (soil/rock) not for structural elements. which means in the FS calculation the strength parameters for the supports are not reduced.

Thank you very much Roozbeh. As I'm still not familiar with the model, I get from your answer that the only FOS I can compute is the sliding one. And that would be by adding the reduced shear stress of each element. No overturning or bearing FOS. Is that right? By any chance, do you have a step by step example for this kind of calculation?

Again, thank you very much.

Jorge

no, you can use it for any kind of models overturning, bearing capacity and etc. the only thing is that FOS is calculated for soil not structural elements (like tiebacks, struts and ..., if you have any).

generally speaking, FOS calculation can be done for soil and structural elements in the same model by reducing the strength parameters. for this case FOS is automatically calculated for soil only. the SR method have not been applied to structural elements yet.

instead of tieback you can use free beam element like struts and put it in front and fix the other end to make sure fixed node. tieback is not a good choice.

I have tried to use a free beam element like a strut in the front, as you suggested. It does not seem to connect to the vertical beam element. Also, tried to make a portal with all free beam elements, the horizontal and vertical elements do not seem to connect. Probably I'm missing something here. It would be useful if the software has more tutorials.

end nodes of horizontal beam should be selected as one of the existing nodes on the retaining wall. this way you can make sure they have same exact information like displacement moment and etc. this is easily doable in ADONIS by simply selecting the "From Struc. Node at Point" or  "From Struc. Node" option when creating the beam element.

Thanks a lot, Roozbeh. Now its working.

Best Regards

Tahsin