Bolt Pre-load with Force on 3D Elements

[Plankis]

Hello,
I'm trying to figure out how to apply a pre-load on a model I've drawn. The bolt is affected by shearing stresses which is the reason I want to use the 3d Elements method instead of the 1D method that most people seems to use.
What is the correct way of defining the cross section? I've been trying two ways and none of them does it the way I thought it would. Here's how I've been doing it:

Create the model, sketch a "half I-beam"-shaped geometry and revolve it to get the screw and nut in the same body.
Use Split body in the middle to get a cross section.
Create a fem and simulation.
Mesh the split bodies.
a. Use "Mesh Mating Condition" to link the bodies together. Or
b. Add "Surface-to-Surface Glueing" between the split surfaces.
Fixed constraints on the head and nut surfaces.
Add "Bolt pre-load" with "Force on 3D elements" and choose the split surface.
Also add a Force so the screw is bent a bit.
Solve it.
Result: Forces are applied, however the mesh mating condition seem to have disappeared because the meshes now intersect each other nor are the meshes co-linear with each other. I don't like that.

Any ideas if this is intended behavior or am I doing it completely wrong?
I've attached the example model I've used with NX 10. There's also pictures showing the displacement for those who don't want/can't to open the files.

 

[Plankis]

Bump, no one has had any experience with the 3d version of the bolt pre load?

 

[PJRatFCCC]

Hi,
I don't have any experience using pre-loaded fasteners but wonder if you are controlling your mesh size in the mesh mating area? I typically use the 1-D connection or bolted connection after creating separate but "united" geometry for my washer(s)mating to my parts. Using mesh control in the area of the washer and the bolt hole should connect your meshes node-to-node if done correctly. This will clean up the "Co-linear" issue. I also use "free coincident" mating in such areas since it's not actually a "glued" condition. I don't know if this will solve your entire problem but I think it might help you get there.

 

[Plankis]

Hello,
I think I have not been clear where I am applying the preload. In this example it is in the middle of the model, this means I have split the bolt in half. I am not really concerned how the nut is behaving so I just revolved it in the same body as the bolt. If you look on the pictures you can see how the meshes are going through each other, even though the surfaces are supposed to be glued to each other.

The approach I did is very similar to this Abaqus CAE video:

https://www.youtube.com/watch?v=aoxWHFo6Ln8

Thanks for the mesh control suggestion though! I have been wondering how to do this type of specific finer meshing.

 

[Tuw]

Hello Plankis,

I have a rough check on your files. To me, your way of applying the loading is correct. To further strengthen my statement, I also check the total vertical reaction force on 1 side, which come close to the value that you applied, 500N.

So, this is a small matter on post viewing only. The exaggerated displacement is the default deformation shape for Sol101(linear), because it sometimes help the analyst to understand the shape of the deformation clearly. If you prefer to have the exact deformation shape, you will need to change the following settings:

1. Result Tab > Post Processing Group > Click Edit Post View
2. In the dialog box, notice the Deformation is checked. Click the Result... button beside Deformation.
3. In the dialog box, change 10.0000 to 1, change %Model to Absolute.
4. Click OK twice.

Hope this helps to clarify.
Tuw