Why is there no descending section of the skeleton curve when I do the cyclic analysis of the column?

[jiayanghnu]

Hello, I am using STKO for the cycle analysis of a reinforced concrete column. I found that when I used ASDConcrete (6p) concrete material and steel02 material, my skeleton curve did not have a descending section. Could you please help me look at my model? Any help would be greatly appreciated!

[kesavapraba]

Hi, The response does not soften since the rebar material (steel02) continually increases when the strain increases. In actual case, there will be fracture at a certain strain value (here I assumed it to be 5%, you should play it to get a good match in your cyclic curve), and the stress response has to be zero after the fracture. So, you can set a MinMax and Parallel material for Steel02 to obtain proper softening response. Please check the revised SCD file attached here and modify it accordingly.

[jiayanghnu]

Hi, The response does not soften since the rebar material (steel02) continually increases when the strain increases. In actual case, there will be fracture at a certain strain value (here I assumed it to be 5%, you should play it to get a good match in your cyclic curve), and the stress response has to be zero after the fracture. So, you can set a MinMax and Parallel material for Steel02 to obtain proper softening response. Please check the revised SCD file attached here and modify it accordingly.

In addition, in my simulation, I found that when ASDConcrete (9p) and DamageTC3D were used respectively for concrete's constitutive materials, even though their parameter values were the same, hysteresis curves were very different. When ASDConcrete (9p) is used, there is no falling section of hysteresis curve, and when DamageTC3D is used, the falling section is more obvious. Is this because ASDConcrete (9p) does not consider damage, while DamageTC does? Can you explain this phenomenon?

[STKO Team]

In addition, in my simulation, I found that when ASDConcrete (9p) and DamageTC3D were used respectively for concrete's constitutive materials, even though their parameter values were the same,

The DamageTC3D tends to overestimate damage in cyclic analysis.

Note that the cyclic damage you see in the experimental test is due to multiple effects you may want to investigate.
First check steel slip and fracturing

Also, you should use a finer mesh for concrete. And the mesh for embedded rebars should be at least equal to the concrete or smaller, but not larger.

Finally, do not apply a displacement only on one edge, otherwise during the right-disp of the cycle it will pull your model, and since there you have nonlinear concrete, it may fracture in tension.
Apply a disp to the whole perimeter

Try first these modification with a monotonic test and then do the cyclic.

[Karishma]

Hello All,
Thank you for sharing your problems and solutions here, we can learn from each other this way.
My question is how can we calculate Gc, Gt, and Lcn for materials properties simulation of Masonry micro modeling as DamageTC3D?

Thank you,
Karishma

[kesavapraba]

Gc and Gt should be calculated from the "stress-displacement jump" or "stress-strain" curves that come from the material testing in labs.
Lcn is characteristic length of the finite element. It depends on how you mesh your model. You do not have to worry about Lcn when you input the Gc and Gt (calculated from stress-disp jump) and enable "auto-regularization" in the material input window of damagetc3d. Following webinar discusses about how to consider fracture energies for masonry material (watch from 20 sec for material definition)

https://youtu.be/lPNd33zOnBE

[Karishma]

Thank you for your reply.
I learned from the analysis that the Gc/Gt doesn't affect the maximum strength. I am trying to test the compressive strength of the masonry prism, the maximum strength of the masonry prism by STKO was obtained twice or more than twice what we got in the test.
I attached the results of both STKO and the test here. Can you experience a compressive strength test of prism and say why the results are so high? (See attachment)

[STKO Team]

Did you try what I suggested in this other post?
viewtopic.php?f=54&t=2929&p=9100&hilit=Karishma#p9100