Rbe2 or rbe3 for this application? 1

[MrK159]

Assume you have a plate with a rigid box attached to it by 4 screws. Using the location of center of mass of the box you place a node or 0d mass there, apply a load or other applied forces (direction doesnt matter). Would you use an rbe2 or rbe3 from the cg to the bolts?

 

[IRstuff]

Is there some reason you didn't post this in the forum specifically for your software, which you've not even indicated?

TTFN
I can do absolutely anything. I'm an expert!
faq731-376 forum1529

 

[MrK159]

Sorry NX 10 nastran

 

[rb1957]

if you connect the cg to the four mounting screws with an RBE, why not solve by hand ? the RBE would defeat the witchery of FEA and simplify the problem to a hand calc.

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

 

[MrK159]

Because its tranferimg load to the plate as well, not just the bolts.

 

[rb1957]

to me it sounded like the mass of the box would directly load the four mounting screws via the RBE ... but if it's the plate you're looking at then maybe ... still replace the box (and it's RBE) with a set of point loads on the screws seems enough ?

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

 

[MrK159]

I'd like to take into account the cg of the box. Cant do that with point loads.

 

[rb1957]

sure you can ... out-of-plane couple, non-uniform distribution (ie not 1/4 at each fastener) ... lick o' paint ...

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

 

[aerostress82]

I posted on your other "title" in Nastran forum too, but since you gave more explanation on this title, I'll add to my previous comment in the Nastran section:
The box with 4 screw connections if you want to be accurate:
- on the part where your screws are on your original plate, model the bottom of your box as shell elements (at midsurface plane) with regards to its "projection geometry" on the plate,
- these shell elements (as part of your box) should have the zero mass and the same exact Young's Modulus (E) and Poisson's Ratio (v) as your box material,
- have nodes both on the screw panel of your box and on the plate at the exact fastener axis locations,
- connect this bottom plane plate of your box to the plate (your plate should be modeled accurate with cutouts as well) via RBE2-CBUSH-RBE2 elements. (CBUSH will be located at your interface plane between the 2 parts and will be of coincident nodes),
- for your box, model it as a CONM2 at its center of gravity location. Calculate all mass and inertia properties via CATIA or some other FEA tool and add these to your CONM2 properties: Mass + I11, I22, I33, I12, I23, I13,
- connect your CONM2 to the massless plate via RBE3 elements to "all nodes" on your massless plate "except" the fastener connection nodes (where you had the RBE2-CBUSH-RBE2 sets mentioned above),
- Regarding CBUSH elements, you may want to enter values that are close to your fastener stiffnesses. 10000 would be good for shear, 100000 for axial translational dofs, and 10e8 for in-plane rotational stiffnesses and 100 for out-of-plane rotational stiffness,
- ALL ABOVE should help you have a really accurate model, but if your 4 screws box is too big for your plate and it has significant "contact" behavior with your plate, then in addition to all that is mentioned above, you might want to add extra "Weighted RBE3" elements between your cutout nodes of your plate and the edges of your "massless plate".

If you would need more accuracy, you could also use contact cards if you are using Nastran 2014 or newer.


Nastran made the "CONTACT" card available for SOL101 to let us model contact cases without having to deal with "Support + SPOINT + Explicit MPC" combinations for the past contact modeling (via Patran Linear GAP Elements Utility). Ignore this last sentence if it doesn't make sense. But I think your solution should be somewhere above..

Spaceship!!
Aerospace Engineer, M.Sc. / Aircraft Stress Engineer