Explicit dynamic analysis of a bearing

[Fabi0]

Hi everybody,

i've a problem with an explicit analysis, dealing with a 2D cylindrical roller bearing working as track roller.

The model is "simple": the inner ring is fixed in its own position, and the rigid surface is pushed against the outer ring.

My problem is that i've set the force applyed to the reference point of the rigid surface to vary linearly along time, but stress dristibution looks like if it follow a sort of sin or cos.
I've done several standard analysis, but it's my first explicit one, so i really can't understand.
Here's a link to the video:

http://vimeo.com/31391947

Thx

Fabi0

 

[rstupplebeen]

Since you are in explicit and you have defined a force it is basically bouncing, try ramping up the load instead of instantaneous. I hope this helps.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[Fabi0]

It's what i've tryed to do, at least in the last step (the one with angular speed on the rigid surface).

I've three steps, the radial load is applyed on the rigid surface in the first step, in the second i changed load modulus and in the third i've tryed a tabular amplitude like you can see here:


=============
=** LOADS
=**
=** Name: Laminazione Type: Concentrated force
=*Cload, amplitude=Tabular2
=_PickedSet256, 2, -8315.
=*Amplitude, name=Tabular2
= 0., 1., 1., 1.
=
=============


But even if it's a different aplitude, it gave the same problems as you can see in the video.

 

[rstupplebeen]

try 0,0,1,1
that way it ramps from 0 to 1 across the entire step

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[Fabi0]

It's not what i want. The radial load must be constant while the rings move.

 

[rstupplebeen]

In the video it was not clear the part was rotating unless I looked at the first and then last frame. Does the oscillations correlate to the mesh?

A more detailed picture, description or CAE file really does help a lot in speeding up help on forums.

I hope this helps

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[Fabi0]

You are right! I've attached a zip file containing cae. Hope it helps, i'm not so skilled with explicit analysis to understand the problem.

 

[rstupplebeen]

Wow you have a lot of elements! You can probably significantly reduce the file size. Also, you have every step take 1 second. In explicit analyses it is time dependent so the slower you load it the longer it takes unless you use some numerical trickery such as mass scaling.

In the first step the load is set to instantaneous. I believe that is your problem. Ramp over the first step instead.

I hope this helps.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[Fabi0]

Thks for your help!
Yes i've got a lot of elements, i'll try to reduce it, even if it's not a main problem (lot of computational power here...i like to "play" with pc network ).
Instead, what do you mean with "the slower you load it the longer it takes"? Should i use 0.5 second for the first two step i.e.?

I've just "ramped" all the loads, hope it helps and i'll let you know!

Thx again and congratulation for your curriculum vitae!!

Fabi0

 

[corus]

It seems clear to me that the bearing rotates so I'm not sure what Rob has seen. I'd look at the displacements close to the contact point to see why you get these strange results in the initial period. Factor them up so you can see more clearly what's going on. It may be due to a 'bounce' or perhaps the contact point is varying somehow so that one passes through another and then corrects itself.