How to output yield stress curve

ASPECT
1

Dear all
How to get the yield stress curve of the initial phase of the model

like this curve in the manual

image

Best
Xu

2

Dear Xu,

One could generate a curve like that for any model output field with the Paraview “Plot over line” feature.

However, what I think you are asking is how to generate a strength profile (i.e., stress versus depth) for the very first nonlinear iteration (i.e., on the first time step) when a constant (fixed) strain rate invariant is given to the material model.

There are a number of ways to do this:

  1. Run the simulation for just one time step and restricting the maximum number of nonlinear iterations to 1. You can calculate the stress invariant from the strain rate and viscosity output, but there is also an option in ASPECT to output the stress invariant.

  2. Run the simulation as normal, but output the solution for each nonlinear iteration (probably not ideal).

  3. Calculate the strength profile independently using the constitutive equations and material parameters specified for the simulation (what I typically do).

Let me know if I can help further with any of these points.

Cheers,
John

3

Hi John

Thanks for your response

If I want to output the stress invariant in ASPECT,what I need to input in the List of output variables

What’s the difference between the first method and the third method

Best Xu

4

Hi Xu,

If I want to output the stress invariant in ASPECT,what I need to input in the List of output variables

You need to add “stress second invariant” to the list output variables for visualization in the Postprocess subsection. See the online documentation link here.

Search for "stress second invariant on that page.

What’s the difference between the first method and the third method

The first method obtains the stress profile from ASPECT on the very first nonlinear iteration where a constant strain rate is used everywhere, and the third method would involve calculating a stress profile analytically in python or matlab using the same constitutive equations, material properties, and background strain rates.

Theses approaches should give effectively the same result.

Cheers,
John

5

Hi John

Thanks for your answer

I got the curve I wanted

Cheers
Xu