Learning Forum

[Mustafa Fahri Karabulut]

Hello everyone;

I've tried to solve static structural using nodal displacement. However; when I give displacement to the nodes (the nodes on the same surface but have different values), the displacement is equal to the entered values on that nodes, while this value is transferred to other points on the surface in small amounts. What should be done to make the nodal displacement value extrapolation or interpolation to the nodes that on the same surface?

[peteroznewman]

It is better to apply loads and supports to areas not nodes when you have a solid element mesh. The reason is a point load represents an infinite stress because the area is zero.

Are there other constraints on the body? Are the material properties properly defined?

[Mustafa Fahri Karabulut]

Thank you very much for your reply. 

I'm so sorry to hear that. Because I think the data I have is convenient for nodal displacement. It would not be appropriate to calculate a single displacement on the entire surface from the data at the points I have.

Except for the nodal displacement constraint, I gave the model fixed support and frictionless support from the bottom. I defined the material parameter myself and defined the density, young modulus, and Poisson ratio values.

Best regards ...

[Rahul Kumbhar]

You should use tabular data to define spatially varying displacement. Using the APDL commands, you can then apply this table as displacement.

[peteroznewman]

Please reply with an image of the geometry in Mechanical. If you select Static Structural in the Outline, it will label the loads and boundary conditions so we can see which face you applied the fixed support, the frictionless support and the nodal displacement.  Is the Nodal Displacement on the same surface as the Frictionless Support?

[Mustafa Fahri Karabulut]

Thank you Rahul.

I will try tabular data but I don't know how to do. Do you have any suggestions (lesson, video etc.)?   

Mr. Peter, I will share the images of my model. 

A and B are the nodal displacement, B is the frictionless support and D is the fixed support. There is 9 part. I defined bonded contact to the horizontal surfaces and to no separation contact to the vertical surfaces (I'll also try the frictional, and frictionless contact until the test nodes RMS that is calculated by using known and model result displacement values is minimum). 

Best regards...

[peteroznewman]

It looks like the Frictionless Support is on the same surface that the Nodal Displacements are on. That is why just the individual nodes poke up, the surrounding nodes are prevented from doing so by the Frictionless support.

It looks like this is a slab of terrain, hundreds of kilometers long. So the surface is not flat. I have only ever used Frictionless Support on Flat or Cylindrical geometric surfaces. I don't know how it behaves on NURBs surfaces.

Please explain what you are trying to model and why you are using a Frictionless support.

[Mustafa Fahri Karabulut]

Mr. Peter;

The frictionless support was applied bottom of the model to restrict the vertical movements. It is not on the surface that applied nodal displacements. 

3D model is Earth structure, vertical surfaces are tectonic faults and the nodal displacements are GPS velocities.  The velocities (per year values, so I use them as displacement) were determined relative to the northern part of the model from long-term GPS coordinates. I've tried to find the stress (per year) value on the fault lines.

[peteroznewman]

Ah, okay, now I understand.

You can't use nodal displacements to pull single nodes on the Earth surface to match GPS measurements of the Earth surface.

You need to create a model of the Earth's crust that applies the pressure from below and from the sides to cause tectonic movement and adjust those boundary conditions until they have an acceptable match to the GPS measurements.

[Mustafa Fahri Karabulut]

You are right. However I think that the result will be more reliable if I can perform the analysis by using some of the GPS data as boundary conditions and some as test data. The applied pressure can change according to the entered material parameter or contact type. The material parameters I use are the values obtained through various approaches and assumptions so I also aimed to test the material parameter in analysis.

[peteroznewman]

I believe the results will be more reliable if you only measure the deformation of the nodes at the GPS coordinates and don’t try to pull the surface using displacements on those nodes.

[Mustafa Fahri Karabulut]

Thank you very much your comments. I will try what you said. Finally confirm my understanding, do I need to give the pressure to the side surfaces as the boundary condition?

Best regards...

[peteroznewman]

You would apply the same loads that the plate gets. I don't know what those are, I haven't done any research on plate tectonics. From what I have heard, plates get pushed on their sides from other plates. But sometimes the edge of one plate dives under the edge of another plate and that pushes up mountain ranges. I don't know what your GPS data show, but there is some elevation change. What is causing that?