Learning Forum

[SriVin]

Hi all ! I have the following problem

I have a block of soil mass, 60m*60m*35m. E of soil is 35MPa, poisson ratio is 0.3, Density is 1800 kg/m3, Mohr-coloumb model with angle of internal friction 30 degree. Now under normal static analysis with base and all lateral faces fixed under self weight top settles by 171mm. at middle value eventually reaches zero at boundary.

Now I am working on soil structure interaction, If i place a building at center of top face on this soil mass, it settles further as expected, I can find the settlement due to building by subtracting values of initial and final settlement. But I want to simulate something close to reality, as soil mass in field dosen't sink under self weight and deformation starts after construction starts.

I am guessing Inertia relief is not an option. I tried putting on elastic mat support on top face is is correct ? If I turn off the non linearity in geometry will the mohr-coloumb model be use less ?

How do I go about it.

above is picture of soil mass with NON linear effects in geometry turned off and with elastic mat on top surface

Non linear effects on and without elastic support.

Non linearity off and no elastic support.

Thank you

[SriVin]

Elastic support is also not the way. It is is supporting the face as well as the the load from top also. I want to do a time history analysis to the same model later

[peteroznewman]

1. You have two planes of symmetry. Cut the model on those two planes and make a 1/4 symmetry model, unless the building load is somehow not symmetric.

2. Don't used Fixed Support anywhere. Use a Displacement support with just the normal component set to 0 and the other components Free. The same applies to the bottom of the cube of soil. It should have a Y = 0 displacement support only.

3. With the BC change in #2 done, when you Solve for the Gravity load only, the top surface should remain flat and subside uniformly.

4. The state of stress from solution #3 could be used with an INISTATE command to set that stress up in the model at the start of a solution that includes a gravity load. Then the nodes will not move until you apply a new load, such as a structure.

[SriVin]

yes, If i use friction less support as you said in 3rd point the settlement of entire soil face is uniform but I don't want it to settle down until I place building on it. As I understood INISTATE after reading few articles, I think we can have a new load on same geometry but my geometry is going to change if I place my building on the soil mass, for which I need to stop settlement of its top face. OR Can we have a new geometry ?

This is my full geometry. What is happening is the top face of soil mass settles by huge amount under gravity load. Now I need to stop the settlement somehow and continue the analysis settlements and other answers from the model.

[peteroznewman]

You have a model with a soil body and a structure body.  You apply the INISTATE command to the soil only, so that it is prestressed by the effects of gravity. In this model gravity will act on both the structure and the soil, but the soil will only deform according to the load created by the structure, because it is already prestressed to support the gravity load.

[SriVin]

I've tried applying INISTATE command but I was not successful, may be I did it the wrong way.

Is is possible that you could assign the required commands, If I send you the model ?

I could use it as reference copy. It would be a huge help.

or

Could you send me the code and step wise procedure ?

Thank you !

[peteroznewman]

I have only seen INISTATE used with success on beam/link models for pretensioned cables. In the case of a pretensioned cable, all the elements have the same state of initial stress, so it is very easy to apply.  You can download some examples that worked from the list of posts found by this Google search.

You can search on Google for examples of INISTATE that don't use the word cable. I don't have a reference copy or any code other than what can be found on this site.

A soil block with gravity has a stress state that varies with depth. How have you accounted for this effect?  What happens if you just have soil in the model, will it converge? Show us what goes wrong. You can attach your model, but I can't promise anything. I learned about INISTATE from jj77 last year, but there are others on this site who have been successful.

[SriVin]

thank you sir, I plan to simply increase the stiffness of soil and decrease the dimensions to get my soil's top face deflection into control