Karamba3D

Comment

Comment by Mahdi Soheyli Fard on February 17, 2013 at 9:54am

to Mohammad Azinkia,

hey buddy, could you pass your definition file here?

however, i think the problem will solve by a greater relation through the beams' kind and the load (how would you might use the components for defining beams?), or, as i should say firstly, it will be more understandable if you attach the file, otherwise my suggest is a bullet in the dark!

Comment by Mohammad Azinkia on February 17, 2013 at 9:09am

 Hi.

I have a problem with  optimization cross section.I connected components according the manual Karamba and I plugged it for example "IPE 100" but it doesn't work.It show me all of beams is IPE100. I decrease loads but doesn't change! please help me what should I do?

Thank you in advance

Comment by Karamba3D on February 16, 2013 at 11:19am

Hi Michel,

eccentricities are rigid: it's a kinematic relation between the endpoints and the centroid of a beam. Therefore eccentricities have no material and no cross section. 

O-Profile section: null thickness means filled profile is a good idea. I will add that in the next version - thank's.

Your box-profiles have zero height so there would be no space for the flanges. The current version of Karamba unfortunately doesn't issue a warning in case of impossible cross section shapes and instead calculates something. Either select a larger height or use the cross section values Karamba comes up with: Use a generate Cross Section Table component, plug a panel to its output, stream it to a .csv-file and open it with e.g. Open Office to have a look at them.

Best,

Clemens

 

Comment by Michel Cassagnes on February 16, 2013 at 8:05am

Hi Clemens,

I had a look to your model. I understand. Could you give more details on the formulation of the eccentricity feature? It should be fine if we had choice for a section with eccentricity between full or partial connection.

I have several questions :

- no material is defined. Which is the default material?

- I suppose that there is also a default crossection ?

-O-Profile section : null thickness should mean filled profile

-I made a test (attached) with two flat section (with weak inertia, inertia given by eccentricity) and calculate the deflection. I don't find the same deflection. Hand calc gives 0.001356m. Karamba gives 0.04409m. It means that it takes a partial connection (as you said in your answer : no sufficiently stiff shear connection). But which one ?

Best regards,

Michel

VierendeelBeam_B.gh

Comment by Karamba3D on February 9, 2013 at 4:18am

Hi Michel,

I had a look at your example (poutre_offset.gh) with the shell that is connected to an eccentric beam. 

The results look a bit surprising but are correct. The reason is, that beam and shell act together like in a Vierendeel beam. There is no sufficiently stiff shear connection between shell and beam.

I attached an example to illustrate the fact:

VierendeelBeam.gh

In order to achieve an efficient shear connection one would have to add diagonals between shell and beam - like for a truss. An alternative would be to model the beam as a shell also. 

Best

Clemens

Comment by Michel Cassagnes on February 9, 2013 at 3:08am

Hi Mahdi,

Beam joints are also in the student version. Maybe just dowload the latest version.

Best regards

Comment by Mahdi Soheyli Fard on February 5, 2013 at 12:02pm

thaank you Michel Cassagnes, for your help on the dome! i couldn't open the file actually because of my gh version that should be upgraded! as Karamba said because i have a student version, the beam joints will not be there!

real thank you by the way
i wanted to make the cross section i used for this model and define joints between the arcs!

patupa.gh

Comment by Karamba3D on February 5, 2013 at 11:07am

Hi Michel,

thank's for the example. It looks like a bug in the determination of the cross section forces. I will have to check it.

Best,

Clemens

Comment by Michel Cassagnes on February 5, 2013 at 8:20am

Hi Clemens,

I made a small model to test eccentricity of cross section (find attached). It doesn't give expected results. where am I wrong ? Could you check? See the moment curves and shell stresses. We should have stresses in the whole thickness and of another form.

Thanks!poutre_offset.gh

Comment by Michel Cassagnes on February 5, 2013 at 8:09am

Hi Mahdi,

You should find it at : Cross Section/Beam Joints (the second item).

• ‹ Previous
• 1
• …
• 18
• 19
• 20
• 21
• 22
• …
• 28
• Next ›
• Page