Simple, but a problem... 1

[duckmatch]

Can someone please help...
this is a case of getting the model fixed in place, but F06 is is constantly showing me

*** SYSTEM FATAL MESSAGE 6466 (SBUT1)
FOR THE 101 INPUT ALL TERMS IN COLUMN NUMBER = 4 ARE NULL.
0FATAL ERROR
1 * * * END OF JOB * * *

As you can see, cylinder is fixed by back plane, loaded by temperature... every node is bounded in all translations and rotations... where is the problem??
I can solve it, but with AutoConstraints?!?

 

[rb1957]

1) look into the manual to see how they set up a thermal analysis

2) call the help desk

3) add a small load to see if it'll work ... i think it's complaining that you haven't applied any loads

maybe it's enough to run with mulitple loadcases, one being a dummy axial load, the other being the desired thermal analysis

 

[duckmatch]

1) Believe me, I looked up on a lot of them so far... and by now, I know the procedure by heart. If you have any suggestion tutorials, I'll gladly look at them;
2) to be honest, don't know what is 'help desk';
3)about loads... the current load case contains displacement on the back wall(all translations and rotations) and thermal loads(exactly results imported from thermal to structural analysis)... am using SOL101 and reasonable thinking - that should do it, but no... if I do not turn ON AUTOSPC, it's always popping out with an ERROR I already talked about!
With them of course am getting results, but not good... basic hand calculation is saying values around 200MPa, but software is saying 300MPa!!
I believe I tried almost everything and am pretty sure that wrong results are because of AUTOSPC, but without them I GOT NO RESULTS at all! :/
I know it's a stupid question, but have to ask it... SOL101 is able to work without AUTOSPC, right?!?! :/

 

[rb1957]

sure sol101 will work without autospc.

how large are your autospc forces ? if they're small then your results are ok-ish.

a hand calc gives 200 MPa and the model 300 MPa ... well, depending that might be good ... is there an appreciable stress gradient around this number ?? is there a simplification in the hand calc ??

what is the physical meaning of a displacement of all translations and rotations ? i think displacing the back wall, which is supporting the tube, will confuse NASTRAN ... won't this mean that the tube is "just" being displaced bodily in the x-direction ??

how are you dealing with rigid body motion ? your tube is constrained onto the back wall, yes? but you're applying an enforced displacement to this ...

you don't know what is "help desk" ? ... you have a NASTRAN license from someone ? they provide support ? ask them (they're the "help desk").

see if your model is doing what you think it should ... apply a test load, something simple, and check the results.

 

[duckmatch]

- So I thought.

- You can see for yourself.
0 MAXIMUM SPCFORCES
SUBCASE/
DAREA ID T1 T2 T3 R1 R2 R3
0 1 1.6064515E+03 2.3271157E+03 2.2659985E+03 0.0000000E+00 0.0000000E+00 0.0000000E+00

- There is no special simplification in hand calc.

- OK, now... I think that maybe I didn't explain it good enough, because reading your post, I saw we maybe didn't get each other pretty well. Sorry about that.
So, in model there is ONLY cylinder... there is no SEPARATE back wall, it's just the face of the cylinders back side... and its nodes are LOCKED in "environment" by setting their displacements in translation < 0, 0, 0>, as in rotations also.
Isn't with those "locked" nodes a question of rigid body motion settled?? What ever loads I place on cylinder??

By the way, thank you for helping... really, it means a lot.

 

[rb1957]

ok you've fixed the face of the brick element; note, there are no rotational freedoms on the nodes of bricks, only translation. you'll need AUTOSPC 'cause NASTRAN has 6 dof at each node and you're not using any of the rotational freedoms. if you look in the f06 you can see NASTRAN finds these freedoms (with zero stiffness) and effectively deletes them from the stiffness matrix.

but if you enforce a displacement on these nodes, aren't you just bodily moving the tube ?

what happens in you model when you apply a tension load at teh other end ?

but you're getting a wack of load coming out of one node (unless there are more SPC reactions) ... looks very odd

 

[duckmatch]

Rb1957, thank you very much for all your help.
I had trouble figuring out some stuff, but @ the end I got it right.
With your help of course... thanks a lot!