• This forum is the machine-generated translation of www.cad3d.it/forum1 - the Italian design community. Several terms are not translated correctly.

warping and shear lag of a fuselage section with applied a torque moment

  • Thread starter Thread starter Cathrine
  • Start date Start date

Cathrine

Guest
Hello everyone, I just signed up for this forum. I'm an aerospace engineering student. I am finding some problems in the resolution of a project. I try to explain it in a short time.
I must carry out the analysis of warping and shear lag of a fuselage section stiffened with currents. (all this using patran/nastran)
questions:
1) I create the fuselage as a "solid" entity and the 4 currents as "curve". when I make the mesh is enough meshare only the cylinder or do I have to remake the operation even for the 4 currents? If so, how do you do it? or better, the mesh of the cylinder comes out thickened and given by elements that seem triangular (tetrahedral?), so with what criterion should I then meshare the currents?
2) to observe the exit of the section I have to apply a torque moment. At first I tried to bind one end of the cylinder and to apply one moment at the other end. In the second case I applied a pair of torque moments equal but opposite to the two ends. When, however, I go to do linear static analysis nothing happens! What do I mean? Going to analyze the f06 is as if no time was applied and then in the "results" section of the patran nothing comes out! then maybe I was wrong to apply the twisting moment? my procedure is: loads/bc - create - force - nodal - new set name - input data - force < 0 0 0> - moment <40000 0 0 > (e.g.) - ok - select application region - geometry - select curve/edge and I go to select one of the two faces of my cable cylinder - ok - apply
the moment is then applied to me on a point of the circular crown ( cylinder side) is there something wrong in this process? Maybe I'm wrong with the point where I hang the torque moment? I was also told to apply it along the cylinder axis, but how do you do it? The axis is not an entity you see. Do I have to draw two points that match the centers of the two sides of the cylinder? please help me...... there is something that I miss on the application of this moment :36_1_4:
3) question of curiosity: I remember that there is a command somewhere in the patran that allows me to choose what things I want to see at the end of my analysis as efforts, deformation, tension etc...... but I do not remember how it is done. Can someone refresh my memory?

I hope someone knows how to solve these my dilemmas especially the second because I tried to search everywhere maybe taking even a simpler structure such as a beam and applying a torque moment, but when I carry out the analysis and "attack" the results do not come out to me as if the analysis did not just make it... .
I look forward to your opinions.
thanks in advance :)
 
we see proceeding by points:
1- the fuselage must be meshata with shell elements and not solid.
2- stringers are made with beam elements to matching knots, or with shell elements with common knots, on one side. depending on the size of the stringer. I recommend shell elements, as if you do a buckling analysis you have no problem with nodes offset.
the fuselage must be bound on all the knots of the two contours, with connection elements to a single central node. use for simplicity in the first instance of rbe2 even if they introduce rigidity to the extremes in the structure.

then treats the fuselage as a ground-bound beam strictly isostatic and applying the loads to the end knot left free to rotate.
For example, assuming that the fuselage develops according to the x axis, at the center junction of the rbe2 sticks in x,y,z,rx. to the right node stick the loads and constraints in y,z.
In this way the fuselage is loaded isostatically and, removed the stiffness due to the rbe2 in the contours you can assess the effect of the torsion.
lets lose solids and solid elements. Get the midsurface from your structures and use a quad4 quadrangular mesh for your model.
 
Thank you for this.
what do you mean when you write:"the stringers do with beam elements to coincidental knots"?
I've never done the rbe2 thing, but I've heard it before. I'll see how to do it.
Thanks again, I hope to solve:)
 
I mean that the fuselage shell typically has reinforcement beams that are normally glued/revetted.
to simulate these reinforcements normally use beam elements with the properties due. beam elements are placed on the edge of the shell elements and share the knots, so precisely to coincidence knots.
in this way the rigidity of the beam element is added to that of the shell element.
modeling shell reinforcement involves the use of very small elements and therefore of a very heavy global mesh.
If you place the model (I imagine it's not secret) I can take a look
 
Unfortunately I don't have a model to post as I'm having problems in applying force, I'm trying to make some more simplified models. I couldn't find where this command is. so I thought that instead of putting a torque moment applied to the center of the circumference, I could put two equal and opposite forces on the two points of the diameter, but alas the program does not read me any strength. when I analyze the f06 is as if the force applied in these two points is not applied, it does not exist. ♪
 
I wanted to send you the f06 plus the paran file, but they are two extensions that do not charge me....so the f06 I converted it into pdf while the other I don't know how to send it to you:(
 

Attachments

to load a file you have to zip it.
do not send patran, bensi' the bdf (input file of nastran).
loads are not there, you probably applied them but not selected to download them.
the analysis turns, but having no load, not from any result.
 
ok .bdf file loaded.
what do you mean when you say I applied the loads but not selected to download them?
I define loads in the load/bc's window then define the material and properties and then start the analysis. Why if I stick the load to the two points of the diameter does not read them and if instead I attach a load to the entire circumference reads it to me? In the end I do the same operations....
 

Attachments

if you open the bdf with a test editor see that there are no loads.
to apply the load to the circumference do a knot at the center of this, then create a rbe3 that has as dependent node the central knot bound from 1 to 6 and as independent knots all knots of the circumference, bound from 1 to 3.
then apply the load to the central node.

I believe that Patran does not transmit the load because you associate it to a geometry and not to a fem entity and that geometry is not in turn associated with an entity, so the load does not move.
work only on fem entities so avoid these problems
 
I had already seen that there were no loads for this I said k the structure had not read the loads I had applied.
However I have remade the model paying more attention to the names of the entities I created (points, curves etc.) and this time the operation went to good end.
the only thing I wonder now is: following the warping the section should go out of the plan and instead if I visualize the deformations it shows me only a change of colors and not the exit. Do I have to put some other parameters to display in the analysis? (I only have stress tensor, displacement and constraints) can you tell me where I can view the list of parameters I want to analyze?
I would also like to try to follow your advice and that is to apply the twisting moment in the center of the circumference, but how does a rbe3 create? :confused:
 
try to make a quick plot and put together deformation and movements, you should see the colored deformed with the colors of the deformations.
to create the rbe3 go into elements ->create->mpc->rbe3
What do you mean by list of parameters?
 
I remember that once a teacher showed us that there was a window in which to go to select the parameters I wanted as an analysis output (type shifts, stress tensor, constraints) only that these 3 parameters the patran li "caccia" by default, but I remember that if they could choose many others too. Maybe I remember badly....? ! ! ? !
According to you in the circumference are only 8 knots to be connected to the central node with the rbe3? Moreover, do you think that as knots I must also consider the points where I put the currents?
the currents I put them every 60°, while the knots I wanted to create them 45° for a total of 8 knots, but if the central node do not connect it to the knots coincident with the currents, the sollectation is transmitted to them too?
 
those you call parameters are the output requests you find under the analysis->subcases->output requests tab.
the load would apply it on all the nodes of the outer shell of the contour, no one excluded.
for a fuselage 8 knots on the contour seem very few. .
depends then that you have to see. the distance between the center of the element and the fuselage, supposed cylindrical is very high. I would put at least 16 if not 32.. You don't make them by hand!
 
Unfortunately, problems continue to arise. in the creation of the rbe3 gives me two errors:
user fatal message 4284 (rbe3d)
user fatal message 1250 (bioers)

I tried to find them on the internet and on the nastran guide but I found nothing.
What do you want? ? What do I keep wrong??? :frown:
 

Attachments

Unfortunately it doesn't even work like that. this time the mistake that gives me is:
user fatal message 5289 (wrgmtd)
 
It works. to me the model turned. I uploaded the results.
Pay attention to the fields. must be eight 8 characters.
then you have to delete your rbe otherwise they are double!!
 
In the end I solved the problem in another way and instead of using the rbe3 I used the rbe2 because the prof told me that in my case it was equivalent to using one of the two instead with the rbe2 was easier. at the end the mistake I made was in the definition of knot weights (which I did not really do and I still have no idea how to do it) xdxd.
 

Forum statistics

Threads
44,997
Messages
339,767
Members
4
Latest member
ibt

Members online

No members online now.
Total: 1 (members: 0, guests: 1)
Back
Top