Force concentration in steel plates when using equalDOF constraints in strengthened rammed earth wall model

[LauraQuevedo]

Dear STKO Team and colleagues,

I am currently working on the numerical modeling of a rammed earth wall strengthened with steel plates. The wall is modeled using the Damage TC3D material, while the plates are modeled using an elastic material.

During the analysis, I observed that the plates are taking a disproportionately high amount of reaction forces compared to the wall, which does not seem physically consistent with the expected behavior (i.e., confinement rather than primary load-bearing).

In my model, the plates are represented using frame elements, and I am enforcing the connection with the wall by applying equalDOF constraints at selected nodes along the interface (linking translational degrees of freedom between the wall and the plates).

However, this approach appears to overly constrain the system, leading to an unrealistic stiffness increase and force distribution, where the plates attract excessive load.

Could you please advise:

Why this behavior may be occurring?
Whether the use of equalDOF in this way is appropriate for this type of interaction?
What would be a more suitable strategy to model the connection so that the plates provide confinement to the wall without artificially increasing the global stiffness?

Any guidance or suggestions would be greatly appreciated.

Best regards,
Laura Quevedo

[STKO Team]

I don't have the model so I cannot give you an exact answer.
Anyways, you have to make sure your model reflects the real behavior:
Is the steel plate actually anchored in each and every part to the wall? If Yes, the EqualDOF is correct.

[LauraQuevedo]

Dear STKO Team,

Thank you for your response.

Following your comment, I have reviewed the physical assumptions of my model. In my case, the steel plates are not continuously bonded to the wall; instead, they are intended to provide confinement through discrete anchorage, so a fully rigid connection along the entire interface may not be appropriate.

I am attaching the model for your reference.

At the moment, the analysis runs successfully when only vertical loads are applied. However, when I introduce the horizontal loading protocol, the model fails to converge and the analysis does not proceed.

[STKO Team]

Now your model is singular.
Every strengthening beam is detached from the others at the corners.
All of them are "Linked" to the wall in Translational DOFs only.
This fixes Rigid Body Translations: the beam are translationally attached to the wall.
But it does not fix Rigid Body Rotation about their own axis!

You can try:
1) merge beams that are physically attached to each other (I immagine each ring)
2) Link them to the wall as you did in your model