Self weight with compression

[bhajay]

Hello Professor,

I would like to do a compression test for soil materials using hypoplastic models. Since I am solving an implicit equation related to the material model, I would like to initialize the stresses to non-zero before applying any loads. So I was trying to apply a self-weight.

• Think of a box of 1 x 1 x 1 dimension
• The sides of the box are constrained along the lateral directions
• The bottom of the box is also constrained for all motions
• To start with no displacements are applied but only gravity is considered to initialize the stresses
• Now I would like to impose a compression load on the top surface

If I use proportional loading for the compression, this does not work. It starts by imposing zero displacements on the top surface. But due to gravity load, the displacements on the top surface cannot be zero.

I also tried to use the "Mesh" command to reset boundary conditions but that does not work either.

I would be glad if you have any suggestions as to how this can be done in FEAP.

[Prof. R.L. Taylor]

You need two proportional loads: (1) for the gravity loading and (2) one for the compression loading  Use Type 2 prop loads s uch that (1) is constant and (2) is zero for the time the gravity load solution is performed and then a linear function, or whatever load variation you want.  LOAD groups (ver 8.4 only) can simplify the specification) or use FPROP for nodes where compression load applied.

[bhajay]

Thank you very much Professor. I will definitely try this.

[bhajay]

But if I would like the gravity to continue acting, this does not seem to work. Enclosed along is the input file that I am using.

Here Elmt 15 is a simple hyperelastic material model for now. I would like to change this to Hypoplastic once I have the boundary conditions working. In my element code, I have already considered the contribution of the body forces into the residual.

Would you have any suggestions, if the gravity has to continue to act even when a compression displacement b.c. is added along the same direction?

[Prof. R.L. Taylor]

If you are using your own element and it is computing the body force contribution then you need to manage the proportional loading.  You can see how we do this by looking at the routine sbodyf.f in the .elements/solid2d directory.  The array prldv(*) has the individual proportional load values whereas 'dm' has the total proportional load amount.  We use values from the d(*) array to control which proportional loading to use and how much.  This is set as MATErial parameters.

I hope this helps you solve the loading you want to do.