Load rate controlled UCT

3DEC
,
1

Hello,

I am trying to model an unconfined compression test on a mortar cube to replicate to what I got from my lab test, which followed the ASTM procedure (a load rate controlled test, instead of strain controlled). I have two main questions.

  1. Should I still use a velocity based approach to apply my load or applying an increasing stress, on the boundaries, using tables is okay too?

  2. What is better, using platens (rigids or deformables?) on top and bottom of the sample or to apply stress directly to the sample like the examples found in the documentation?

Thanks for the help!

2

You have two end cases. Applying velocity to the boundary is like an infinitely stiff platen. Applying a stress is like a very soft platen. Obviously platens are not infinitely stiff, but if the platens are, say, steel, and the mortar is relatively soft, then this approximation is probably OK.

I’m not sure what you mean by “load rate controlled”. If it is applying increasing levels of stress, you could do the same in 3DEC. But again, this means the platens will be “soft” so you may want to model the actual platen in this case.

3

Ah, interesting — thanks for the explanation, Jhazzard. By “load rate controlled,” I mean in the lab test I prescribed a load/stress rate (for example, an increase of 75 psi/s) rather than a displacement rate, as is typically done for soils in triaxial tests. I am following ASTM procedure for concrete-type materials.

In the lab, my mortar cube was loaded under an applied stress rate of 75 psi/s, and I tried to reproduce that in 3DEC using a stress boundary controlled through the table function.

One more question: if I want to accurately replicate the lab behavior, would I necessarily need to implement some form of servo control with the velocity approach? I’m using 3DEC 5.0 and I’m not sure what the best approach would be in that version.

4

I think if you are using a stress boundary, you probably want to add a stiff platen, and apply the load to the platen. If you use a velocity, then maybe you want some kind of servo to ensure that the rate of stress is constant. Probably it would be easier just to add the platen and apply the stresses directly.

5

Edgar,

There are various published studies that have looked at loading rate sensitivity on UCS/triaxial test simulations. The first two that spring to mind are FEM-DEM studies and are of rock, not mortar, but the same principles may still apply and help give you ideas to follow a similar procedure and have confidence in your research. There are likely also other relevant studies already undertaken in 3DEC.

Mahabadi, O., Lisjak, A., Grasselli, G., Lukas, T., and Munjiza, A. (2010). Numerical modelling of a triaxial test of homogeneous rocks using the combined-finite discrete element method. In Zhao, J., Labouise, V., Dudt, J., Mathier, and J.F., editors, Rock Mechanics in Civil and Environmental Engineering, pages 173–176. Taylor and Francis Group, London.

Tatone, B. (2014). Investigating the evolution of rock discontinuity asperity degradation and void space morphology under direct shear. PhD thesis, University of Toronto

1 Like
6

Thank you both for the help!
I will take a look at the papers and then decide what could work better for my problem.