Verification of deformed shape of knuckle joint. (My first FEA) 4

[nemosis]

This is simulation of knuckle joint with 50KN tensile load on rods. (One end fixed)

Here's how it's assembled, if you don't know what knuckle joint is.
assembly animation mechteacher article

Someone told me it shouldn't deform like this(and the part with two eyes should contract), without explaining why.


Any help would be highly appreciated, I have to present my work tomorrow

p.s. I am using Abaqus/standard, ask anything else you might need.

 

[btrueblood]

Take each part individually and explain why the stresses appear where they are, and what that implies for the deformation. Start from one end and do the explanations for each piece.

Hint: "Someone" doesn't know what he is talking about.

 

[nemosis]

Take each part individually and explain why the stresses appear where they are, and what that implies for the deformation. Start from one end and do the explanations for each piece.

That's pretty neat idea! Thanks.

I performed several more simulations after that(with varying parameters and general contact) here are results if anyone wants to have a look at it and comment. -

http://imgur.com/a/xLwK3

 

[Mustaine3]

Without knowing all details of the model, I would say it looks ok. The general deformation makes sense in my opinion.

You are aware that the deformation scale factor is not 1?

 

[rb1957]

I find it hard to visualise how the two lugs of the clevis could "contract" (I get what you mean, instead of splay apart).

I'm looking at the loading on the pin, and I can see that it'd bend only one way (with a tension load), as you've shown.

draw simple FBDs to understand what's happening.

another day in paradise, or is paradise one day closer ?

 

[jhardy1]

Poisson's Ratio effect

https://en.wikipedia.org/wiki/Poisson's_ratio

might tend to cause the yoke to contract laterally while it is under pure longitudinal tension, but I agree that the splaying of the yoke due to the bending of the pin looks "sensible", and I imagine this effect would probably dominate over the Poisson's Ratio effect for the geometry shown.

You should be able to load each part in isolation to get an understanding of the relative magnitude of all such behaviours to confirm the overall behaviour of the assembly is "sensible" - and you would probably also gain some insight into the possible interaction of the various behaviours, assumptions you have made about the contact interfaces, clearances, friction, etc.

http://julianh72.blogspot.com

 

[ESPcomposites]

Note that if you are interested in predicting the ultimate load capability (or comparing to a classical solution), you will need to consider the plastic material response of the lug. However, in that case, the mesh would need to be much finer at the Kt locations.

Also, for an improved deformation result, you probably should not have a contact condition between the tang and the clevis (it appears you do now). You need a gap to allow for manufacturing tolerances and to allow it to rotate.

Brian

http://www.espcomposites.com

 

[nemosis]

Thank you all!

@ESPcomposites Yeah I know but I only have to study linear static simulations.(Plastic deformation comes under non-linear, right?) BTW I have other simulation models which includes all manufacturing tolerances. (Clearance between pin and both eyes/tong/clevis AND clearance between single and double eye.) Deformation in that model looks similar.

 

[corus]

You're using Abaqus as a solver and yet you've just applied a tetrahedral mesh everywhere. You'll get better results by spending a little time on the model to get structured regions and using brick type elements, preferably quadrilateral elements. A finer mesh at points of contact and where you think regions of high stress will occur, would also give better, more accurate results. However, as others have said, the general deformation is what you'd expect.