Learning Forum

[WooMin Jang]

Hi i wanna do slope stability analysis in LS-DYNA. 

But there's a problem with gravity initialization.

Dynamic Relaxation or Implicit initialization both give geometries that is not same with slope i modelled...

Is there any way to initialize gravity with "original geometry " to be undeformed...

Because it is slope stability analysis, it is essential to apply gravity load... 

Thanks in advance.

P.S.Upper is original(h=20m) and Down is after DR(Initialziation)(h=17.xxm)

[Jim Day]

Use the results you already have to estimate an initial geometry that, when settled due to gravity, gives you close to the geometry you want. You can iterate using this approach a couple of times to get the accuracy you seek.

[WooMin Jang]

I can't fully understand... Just iterate to get the results close to my intial geom??

Well.. the results were almost same as lower figure when i did that

[Jim Day]

Sorry to be unclear. What I meant to convey was to remesh your initial geometry so it better represents the unconsolidated geometry (like what it would look like if there were no gravity load). In other words, scale up the geometry so the initial height is around 23.0 so that when it consolidates, the height becomes closer to 20.0. You may have go through some trial-and-error to get the right initial geometry. Hope that's clear.

[peteroznewman]

I'm not an expert at LS-Dyna, but I can share some information about Ansys Static Structrual that is relevant.  I'm responding here rather than your new post since this post has the images.

In a nonlinear (large deflection ON) static structural analysis, you can turn on the Inverse Option. That takes an input mesh shape that assumes a load (gravity load in your case) is present at Time=0 and solves for the shape of the mesh that would exist when the gravity load is removed at Time=1. To perform inverse solving, simply set the Inverse Option property to Yes in the Advanced category of the Analysis Settings.

Another way to have the shape of the soil surface remain unchanged when a gravity load is applied is to use the INISTATE command. This command applies a stress to every element in the model at Time=0. If you apply the correct values of stress and a gravity load, then the mesh will be in equilibrium and the mesh won't move at the end of the first step at Time=1.

[Jim Day]

LS-DYNA does not have an Inverse Option.   Your issue seems to be you have a precise target geometry after settlement due to gravity but don't know the associated stress state, and further, you don't know the geometry prior to application of gravity.   As for the aforementioned associated stress state, *INITIAL_STRESS_SOLID or *INITIAL_STRESS_DEPTH would allow you to impose an initial stress field.   Perhaps the latter would provide a state close enough to equilibrium to suit your purpose.