Learning Forum

[h395523899]

During the simulation model contact process, I found that the result of the gap is negative. How can I set it to change to a positive number?

[peteroznewman]

I assume a negative gap is a penetration.  If you want less penetration, you can adjust the contact settings to limit the penetration. Please show the contact definition details and the mesh on each side of the contact so we can see the target and the contact sides separately.

[h395523899]

My model is two squares that are symmetrical up and down.

The irregular object in the middle of the square is the support.

It mainly simulates the change of the gap between irregular objects.

Model size: length 4m, width 0.5m, thickness 1m and 5m

The bottom surface is fixed, and the top surface of the model is applied with a pressure of 20 to 100 MPa.

Grid data:

The upper and lower surface mesh size of the model is set to 100mm

 

The intermediate contact surface grid is set to 50mm

The thickness of the intermediate support is set to a surface mesh size of 2.5.

[peteroznewman]

Try typing in a small value for Penetration Tolerance instead of Program Controlled.

You can also try to change the formulation to Penalty Method.

[h395523899]

Instead of seeing the Penalty Method, I saw the pure Penalty option.

Penetration Tolerance instead is set to 0.1mm~0.01mm

But I tried both ways and still didn't change the sign of the result.

[peteroznewman]

Instead of Rough, try Frictional with a high Coefficient. Is there some constraints on the top block that prevents all other motion except for downward displacement?

Try a much finer mesh.  It seems the contact areas in red and blue are below the top surface of the other colored bodies. Is that correct?

Yes, Pure Penalty, I wrote the reply from memory. Sorry about that.

[sathya]

Hi First make sure the CAD is without penetration.you can do this in DM using Analysis tool.If no penetration in CAD you can play with contact settings.
You can confirm same in contact status in mechanical.

[h395523899]

1. Because I am simulating the simulation of the formation and proppant. So I didn't set the constraints around the upper and lower baffles. Only the upper surface of the upper baffle is loaded with pressure, and the lower surface of the lower baffle is loaded with a fixed constraint.

2. So the surface of the red and blue areas is indeed lower than the surrounding surface. The specific enlarged view is as follows.

3. The original gap of my model is 5mm. If I want to study the deformation of the gap, do I need the grid size to be less than 5m? Or the mesh size in the direction of the vertical deformation plane needs to be less than 5mm?

[sathya]

Hi,

Sometimes higher Pinball radius will result in Penetration.

So I guess in this model your pinball of 5.1mm is too high considering the CAD in contact.

if you had given contact between 1&3 and 2&3 then reduce the pinball radius to auto or magnitude(just enough for closed condition)

 

[peteroznewman]

h395523899, I agree with sathya that the faces that you want to be in contact should be touching in the CAD model before you start solving.

If you do that, then the red and blue surfaces from the earlier post will be touching and the colored bodies will be interfering. I don't understand what you are trying to simulate here, it is very confusing to me why you are making contact at the lower surfaces.

If I had your geometry and wanted to press them together, I would create one contact selecting the upper surfaces to make contact, not the lower surfaces. That contact definition would be what defines the position of the upper block. If I wanted to know what the Gap was between the lower surfaces, I would put in a second contact definition (same as the one you have) to keep track of the gap between the lower surfaces. In this case, the upper surfaces should be touching in CAD. Neither contact definition would include the side walls of the colored bodies.

In the zoomed in image above, if you want contact at the lower surface, why do you have the vertical wall to the higher surface selected?

 

[h395523899]

Sorry, maybe I didn't make clear my intentions.

(1) The model is divided into three parts

1 is the upper platen, 2 is the intermediate support, and 3 is the lower platen.

The upper and lower plates are the same material. The intermediate support is another material.

The support is not completely filled with the gap between the upper and lower plates, and there is a missing portion between the supports.

In other words, the support is a patchy distribution.

(2) I want to simulate the change of the gap between the upper and lower plates under pressure.

So what I want to simulate is the change in the gap without the support. The initial state of this gap should be 5mm. Under the action of pressure, the gap gradually becomes smaller and becomes 0.

The smaller the gap is mainly caused by the deformation of the board, not because the support is crushed.

That is, the support is harder than the board.

The following is a simplified schematic

(3) The part I chose green is just to show that there is a support, that is the thickness of the support. The contact relationship I am currently setting is represented by a simplified diagram

So how do I set it up in this situation?

According to what you mean, should I establish 3 contact relationships?

The first contact relationship is the upper and lower surfaces of the gap

The second contact relationship is the contact surface between the support and the upper plate.

The third contact relationship is the contact surface between the support and the lower plate.

Simplified diagram

[h395523899]

sathya,Thank you for your answer, please take a look at my new reply. About the Pinball radius Since my initial state of the gap is 5mm, it is not closed, so this is set.

[peteroznewman]

Yes, create 3 contacts:  1-2,  2-3 and 1-3  using the Body numbers. The last one is the gap you want to measure.

It would be better if you had named the contacts with letters A, B, and C rather than use numbers again which make communication less clear.

 

[h395523899]

When I want to establish a contact problem of 1-2, 2-3, I can't establish contact relationship after selecting A B surface. For example, by selecting the face name, it is displayed in yellow.

[peteroznewman]

I'm not sure why that is yellow. Please show the collection of faces in each Named Selection.

Another approach is to have no A or B contact pairs and use Shared Topology to connect the bodies 1, 2, 3 into one mesh with shared nodes at the A and B interfaces.  You would only have contact C to measure the gap. You have to bring your geometry into DesignModeler or SpaceClaim to create Shared Topology. Which one are you using?

[h395523899]

I use SpaceClaim, and I originally used Shared Topology.

[peteroznewman]

Thanks for the screen snapshot.

In SpaceClaim, go to the Workbench tab and click on the Share button and show a screen snapshot of that.

Once you have Shared topology, you would delete (or suppress) the A and B contact definitions.

[h395523899]

I didn't understand the meaning of your sentence.

In SpaceClaim, go to the Workbench tab and click on the Share button and show a screen snapshot of that.

Is this Share button in SpaceClaim or on the workbench work surface?

I have not found both of these work surfaces.

[peteroznewman]

[h395523899]

Maybe our version is different. I am using 18.0.

[peteroznewman]

Then you did the only Share you can in 18.0. The Workbench tab is a 19 improvement.

You can create a Workbench Project Archive .wbpz file to attach after you post your reply and I will take a look.

[h395523899]

I didn't find the button to upload the file

[peteroznewman]

The Attach button only appears after you post.  Look again.

[h395523899]

The file has been uploaded, please take a look

[peteroznewman]

I looked at your model in ANSYS 19.2 Having contact between faces on the same part doesn't seem to give a graphic to see the contours of the gap.

I opened the geometry in SpaceClaim and set Shared Topology to None, then added bonded contact to the top and bottom of the thin plates, and added a frictionless contact between the large flat surfaces on dc1 and dc2.  I don't know why Gap results show as negative values.