Energy balance

[kolozvari]

Hello, is there a way so record and plot energy ballance (strain energy, damping, etc.) from the dynamic analysis of a SSI model?

Thanks

[STKO Team]

Actually the only materials that have a built-in support for energy computation are uniaxial materials. And subsequently beam elements via fiber cross section can integrate those results.
But in general you would need a tcl custom script to compute energies element-wise and integrate them.

[kolozvari]

Ok. Thank you.

[kolozvari]

Hello,

I am currently trying to understand the amount of energy dissipated by the soil during an earthquake in a full SSI model of a tall building we discussed before. Based on our previous response, it can be quite difficult to calculate the amount of energy dissipated by different components of the model as it needs to be done manually, and I might try to do this later. At this point, I am thinking about saving some of the responses by the soil to understand if the soil goes nonlinear or how much damage it experiences. When I am looking into available options under "element results" in definition of MPCOrecorder, there are many options to consider. What would be the recorders that you would recommend (e.g., strain, stress, damage, plastic strain etc.)? Also, what does the "damage" recorder record and what is the difference between "damage" and "material.damage", and which one would be more appropriate for me? The model is quite large and I am trying to save only quantities that are necessary.

Looking forward to hearing from you.

Thanks

[STKO Team]

The damage can be used with the DamageTC3D or ASDConcrete3D which gives the damage output.
As you know, the output you can ask to OpenSees depends on the element and/or material model you are using.

For soils, if you use one of the PressureDepend(independ)MultiYieldXX variants, you can simply ask for
material.stress (if you have multiple gauss points in an element) or stress (if you have only 1 gauss point).
The stress tensor will have a last extra component that tells you if you are in a nonlinear state:

PDMY.png (33.77 KiB) Viewed 2562 times

[kolozvari]

Thank you for your response.

Here are the responses I can obtain for my soil elements model with pressure-dependent/independent materials you mentioned (see the first screenshot). Can you please explain what are the C1, C2, C3, ..., C7 are the parameters? Why is there 7 of them?

Regarding the second screenshot, which shows different stress components, can you please explain the meaning of all parameters other than Sigma,i,j (Principal Min, etc., Principal Mid, etc.)? Which parameter is the stress ratio parameter you mentioned from all these parameters?

Finally, do you have any additional advice on how to interpret the strain/stress behavior of the soil? I would primarily want to know if the soil goes in the nonlinear range and what is the magnitude of the nonlinearity.

Please let me know. Thank you.

[kolozvari]

I have a follow up question. After running the analyses with my model in which soil is represented with Pressure dependent/independent nonlinear materials for soil, I realized that all stress results are zero except C1, C2, .., C7 parameters (see screenshot below). I changed the material to Elastic Isotropic and re-run the analysis and all results were there. Is there anything I need to do in particular when defining recorders for nonlinear soil materials to obtain the results? You can also see a screenshot of how I defined my recorders.

I have send you the model via email so you can take a look.

Please let me know. Thank you.

[STKO Team]

It's because the developers of those material models decided to put in the "stress" result, 7 components.
The first 6, C1 to C6 are the stresses (s11, s22, s33, s12, s23, s13)
The last component C7 is the eta parameter in the description we posted previously. It tells you if the material went into a nonlinear range.

[kolozvari]

I see. Unfortunately, when I plot C1, C2, .., C7 parameters, they all show results except C7, which is the most important one that I can use to figure out if the material goes into the nonlinear range (see screenshot below). As you can see there are no results and the scale is all zeros.

Also, how do I get the strains? Strain results (eps11, eps22, etc.) are also blank. Is there anything I should be defining differently in my recorders? Please let me know.

Thank you.

[STKO Team]

Unfortunately, when I plot C1, C2, .., C7 parameters, they all show results except C7, which is the most important one that I can use to figure out if the material goes into the nonlinear range (see screenshot below). As you can see there are no results and the scale is all zeros.

It's because your material is never updated to plastic.
Those material models are supposed to be used like this:

1. By default they start as elastic material (they need an initial state of stress before being used as nonlinear)
2. Add gravity loads and run a gravity analysis (keep load const and reset time to 0)
3. Update materials to plastic
4. Re-run the same gravity (without adding extra loads, it's just used to correct the previous analysis in case some element are not anymore in the assumed elastic regime)

This is the workflow:

UP_1.png (7.64 KiB) Viewed 2394 times

Were the update is done like this (custom command):

Code: Select all

set nl_soil_tags {1 2 3 4 5 6 7 8}; # put here the tags of your PDMY physical properties
foreach mat_tag $nl_soil_tags {
	updateMaterialStage -material $mat_tag -stage 1; # 0 = elastic (default), 1 = plastic
}

Also, how do I get the strains? Strain results (eps11, eps22, etc.) are also blank. Is there anything I should be defining differently in my recorders? Please let me know.

Unfortunately, those material models give some inconsistencies in the XML tags of the setResponse method. A simple workaround is to ask for "material.strain" (this will work for other materials), and "material.strains" (this will work for the PDMY)