How to model a pin inside a link plate?

[Deleted User]

here is an example of what i mentioned before. yet another way you could think to load such a part. this is an assembly with bolt and nut preload. the part is restrained in the middle. obviously, only half the part could have been modelled, but just wanted to show the whole thing. the material properties are ansys structural steel so this won't correspond exactly to yours. the meshing was done in ansys. element quality plot is shown. most parts are hex elements, the plate is tets. midside nodes are used. all bonded contact was used. the pins in this case are smaller than the holes in the plate. so the holes can deform in this model. i set the bolt preload based on fatigue life. then i added the force to stretch the plate and based that on fatigue life as well. so multiple analyses were performed. lastly, the whole model was analyzed, i have a slice plane on after the fact for visualization purposes.

you could also do something like this in mecway as well.

johnkent

Yes, prop_design, such a simple geometry but I'm still learning:

In order to see which elements are actually in contact, I looked at the compressive stresses on the inside of the hole. Below is the plot for sigma-3 and indeed the bottom half is in contact (the green elements are negative stress). In order to balance the 100 kN force, the normal stresses here should run around 100 MPa and they do. So all seems consistent.

But next level of physical reality would be to include the pin (with fixed constraints) in a two component model.

VMH

prop_design, thanks for you comments and just an opinion, I feel that only half of the outer surface of the pin will contact half of the inner surface of the hole leaving the other half not in contact. Bonded contacts have all applicable degrees of freedom engaged (no sliding and no separation). I think the contact between the outer surface of the pins and the inner surface of the holes should allow for separation. I'm not sure what type of contacts are available in Ansys.

johnkent, another way to see where compression only support are engaged, see link below.

[Deleted User]

hi vmh,

yes i know what you mean about the contact. i did run a number of cases like that. i just don't know enough about this particular problem to be very specific. i posted i think three different loadings but ran many more. but i know what you mean. it could be that is what he is trying for not really sure. the results do change depending on boundary conditions. so it will be important to figure out the ones that match reality, unless this is just an academic study.

J_Marc

Victor, VMH & Prop_Design,

I read through all the post and watched the video. I have learned thanks to you a better way to model this problem on Mecway (yes, these holes were meant to be circular not hexagonal). I am very grateful for the time and effort you've put into this problem.

PS: I have downloaded a copy of FreeCad which I intend to use to model more complex parts in the future. I am still trialling Mecway on know problems until I am satisfied I can trust the results. So far I am very happy with the intuitive nature of the graphics an the ease in electing nodes and surfaces. Keep up the good work Victor!

[Deleted User]

Hi,

Just wanted to say that my screencasts have been updated and are located at;
http://propdesign.jimdo.com/download/

The Netgen screencast uses the latest version, 6.1 Experimental. Same results though.