MPC Forces Not Making Sense 1

[CDH007]

Hi all, I have an interesting problem that I've run out of ideas on.

I have a model in SI units that includes a large RBE3 with a mass element on the dependent node, and four equally spaced independent nodes connected to zero length pbush spring/damper elements. Each pbush element is then connected to the dependent node of another, smaller RBE3 element which spiders out to a post. So essentially I have four posts with a mass in the center which I then subject to GRAV loading.

When I go to look at the MPC resultant forces, the forces on the independent nodes of the large RBE3 with the mass are equally distributed and add up to my applied load at the dependent node. However, when I look at the other side of the spring connected to the dependent node of the smaller RBE3's (which are then connected to the posts) the MPC force value is about 7% less than it is on the other side (independent nodes of the large loaded RBE3). Any ideas as to why this might be?

The pbush translational stiffness is 1E+12 and its rotational stiffness is 1E+9. I have tried other stiffness values to no avail. I have also tried making the larger RBE3 into an RBE2 just to see if there was any correlation, but with the same results.

Any help would be appreciated. Thanks!

 

[Onda]

Could You post the *.bdf So we can get run it?

 

[CDH007]

Here is a self contained .bdf (this is a small section of the model as not to give away any proprietary information etc...) that you can run and see the results.

 

[Onda]

I run your model with MSC Nastran2007 and the difference is only 1.18% (418 vs 423)
I don't know if could help...
Onda

 

[CDH007]

Onda,

Could you post the .f06 or .op2 file? I'm curious to look at the MSC Nastran 2007 results. I've been using NX Nastran 6.0.

Thanks
CDH

 

[Onda]

Here both files. Let me know.

Your model look right and is very interesting.

Onda

 

[CDH007]

Onda,

Thank you for posting those.
When you say the difference is only 1.18%, what values are you looking at and where?
Are you looking at the Bush forces vs. applied MPC force or are you looking at just the large RBE3 grid point MPC forces?
Subcase 2 and 3 don't add up when you look at the total MPC forces on the large RBE3. For subcase 2 I see a value of 496N on each independent node of the large RBE3 and for subcase 3 I see 507N. That doesn't really make sense considering the total load should only be 1692N... Thanks for your help!

CDH

 

[Onda]

I have only look on subcase 1, really..
I did not look at all at subcase 2 and 3.

On Subcase 1 the difference on SPC Force (vector plot in patran) but I think You use Femap is 1.18%

Onda

 

[CDH007]

I created a group in Patran of just the pbush elements and moved nodes so I could see results on both sides of the pbush. I then went to plot options and selected Coord 100 for my coordinate transformation. That way all results are in the loaded coordinate system (100).

For all cases, the bushing forces match perfectly. (423N for case 1, 496N for case 2, and 507N for case 3). However, this is not the case for the MPC forces (which does not make sense to me).

For case 1 (GRAV in X Coord 100), the MPC force is 423N (423N Fx, 2.93e-6 Fy, -1.18e-7 Fz) on the large RBE3 side and 419N (-419N Fx, -2.28e-2 Fy, 1.88e-4 Fz) on the small RBE3 side. I would say this is acceptable and close enough.

For case 2 (GRAV in Y Coord 100), the MPC force is 496N (-260N Fx, 423N Fy, 3.58e-4N Fz) on the large RBE3 side and 381N (257N Fx, -281N Fy, -4.08e-4N Fz) on the small RBE3 side.

For case 3, the MPC force is 507N (280N Fx, 0.0301N Fy, 423N Fz) on the large RBE3 side and 395N (-277N Fx, -0.0355N Fy, -282N Fz) on the small RBE3 side.

Any idea why the MPC forces don't balance across the bush for Case 2 and 3 (GRAV in Y and Z)?

 

[CDH007]

Another interesting thing to note is that the loads into the posts through the bush element for the Y and Z cases (and their reaction loads) have a significant X force which actually gets translated quite well through the bush. Just the Y and Z forces which get reduced drastically. Something to consider I suppose…

 

[Onda]

I've try to change the Analysis coord of the grids of Cbush and set both to coord 0. the MPC force change. maybe here is the solution..

Onda

 

[Onda]

The dependent grid point of the spider RBE is also a grid point of the solid. I think that part of the load is going thrught the solid .
If you sum the MPC force of all indipendent node and compare with the MPC force of dipendent grid the two forces are the same.

The force is going from the indipendent point of "princioal" RBE3 to the CBUSH and remain the same and then part of the force goes to the RBE3 spider and part of the force goes to the solid elements.

Onda

 

[CDH007]

Onda, you are absolutely correct and this is something that was corrected on a later iteration.

That explains why the SPC forces were correct even though the MPC forces were not. That "disappearing" load was simply taken by the solid elements and therefore not visible on an MPC force plot. Probably doesn't effect the analysis because the load path is similar (still through the posts).

Thanks again for your help!

CDH