I am aiming to simulate a deep excavation near an existing tower with 4 levels of below grade parking (shaded in grey). The approximate loading of the building is 10 ksf. Therefore I have simulated the material (shaded in grey) as a non-porous Mohr Coloumb material with stiffness and strength parameters equivalent to concrete. The density is0.187 kips/ft3 in order to achieve the 10 ksf vertical pressure at the base of the building. I have also added a negative interface on the outside of the building with Rinter of 0.6. The soil layers are simulated with the Hardening Soil model. I was wondering if this approach makes sense and if not, what is the best way to simulate a nearby burried structure.
That doesn't look bad for me. I usually model the basement walls with plate element (and add horizontal beams on possible slab locations) and add the total building load as a distrubuted load at the bottom. Of course plate elements have zero unit weight. With this, I can take the small deformation of the building into account too while in your case it is very rigid. But, of course, all of these are rough approximations to buildings that we have not designed. Rinter is all other subject and nobody can give you the correct answer. Just one note, I would make sure that building does not have any piles etc. to support that excavation since it is not very shallow. If they have piles, you might have a little trouble with the anchors. Lastly, click on embedded beams of anchors' bond length and select the behaviour as grout body, as far as I can see it is not correct right now - you can make sure that by checking if there is any small green rectangles at the beginning or end of the embedded beam.
P.S.: The contractors usually do not like the first anchor that you have there crossing through other anchors. I can imagine that in 3D, they do not intersect, but considering tolerances and other possible construction problems, if they catch the strands during drilling, you might have problems since anchor will lose its force.
Hello Boris,
I believe that Berk Demir explained it very well, including some important modelling aspects to be aware of (e.g., anchors crossing).
So, it comes down to how exactly you want to simulate the building, as different approaches can be used.
Typically it comes down to engineering decisions and if you feel unsure, try to run both cases and see if there are big differences so that you can decide which approach gives better results.
In principle, modelling the building as a polygon gives a good representation, even though many times we select the Linear Elastic material model to simulate concrete as we expect failure to occur on the soil side and not so much on the concrete.
Thank you Berk! I will give it a try with the plate elements and compare the two results. Regarding the grouted body, the color is misleading because I chose a green color for my embedded beam rows. I changed it to light blue now and there are no green rectangles. I simulated the grouted body as massive circular beam with a diameter of 6".
Answer Verified By: Boris Kolev
Thank you Stefanos. I just checked the simulation with the linear elastic properties as you suggested. The displacements/FoS look fairly similar. I will run a second simulation to try out the method Berk suggested. Appreciate the prompt response.
Hi Berk,
I have a follow up question regarding using plate elements with 0 weight to model the building. I am seeing some unexpected deformations that don't correspond to what we've observed in the past with similar deep excavations - we usually see higher deformations at the top of the wall and lower towards the bottom. I have a suspicion that the interaction between the soils in the wedge and the building foundation walls is not modeled correctly. When you use this approach (stiff plate elements) to model parkade levels, how do you model the interaction between the plates and the soils? Thank you in advance for your help.