Add karamba.dll (which is the wrong file) first. Then use a text editor on the MSVC .sln-file and change karamba.dll to karamba.gha.

Best,

Clemens

OK, I renamed .gha to .dll and it worked. Thanks a lot !

Last question : did you developped yourself all the classes needed to add radio buttons or, was it already available ?

The classes were developed by Christoph Zimmel.

Best,

Clemens

How can i define joints in my models?

regards

Use the 'Beam-Joints'-component.

Best,

Clemens

does "beam-jonits"- component have another name! :)

i cana find it by the way!

It is part of the trial and pro-versions of Karamba.

Well, 

Thanks a lot Christoph, alias "the genius behind the customized user-interface elements". But you don't mind if we use your classes to improve our customised components (which are not to be sold of course) ?

As Karamba is not free, although Grasshopper is, it is not clear...

Hi Mahdi,

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

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

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

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

Hi Mahdi,

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

Best regards

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

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

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

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!

Thank you Mahdi for attention.

here is the file.

1.gh

Hi Mohammad,

there are two things you have to change in your definition:
- flatten the element input at the assemble-component. Otherwise you get three different models instead of one
- Feed a list of cross sections into the cross section optimization component. 

Best,
Clemens

Thank you Clemens!! Best help as always.

It works well.

another question: we should analyze structure in x and y by force to analyze what is structure reaction and cross section.Is it any way to have maximum cross sections between x and y direction?If yes how?

In huge project, if we want to construct our structure how can we have output of the cross sections optimized?here I just see a cloud of fonts.

You get the optimized cross section for each element by disassembling the model and then disassembling the elements. In case this does not work as intended with Karamba 1.0.2. send me an e-mail (info@karamba3d.com).

Regarding your previous post: the cross section optimizer computes the cross sections in such a way that their utilization is less than 100% for all load cases. If you have e.g. wind blowing from two directions create two load-cases (see manual).

Best,

Clemens

How can we use double cross section in karamba for e.g. we want use double "IPE" for column or beams.

At the moment there is no direct way to define double IPE cross sections.

You could set up a cross section table with the exact cross section properties of double IPEs, but set their type to e.g. Box-cross section. When displayed you would see box cross sections which have the physical properties of double IPEs.

Best,

Clemens

Thanks Clemens!

Tony: thank you for uploading the video.

Best,

Clemens

Hi Clemens,

I don't understand Princ Stress results. I made a simple wall loaded in its plane, downwards loads on each mesh. I expected to have princ Stress nearly horizontal at bottom and top of wall. Instead I get something that seems to depend on triangle orientation : see attached. gh file attached also.

Thanks for your help.

Michel

wall01-princStress.JPG

wall01-Supports_Loads.JPG

shell_test.gh

Hello. I am new to using karamba so I am fairly unfamiliar with how the components work. I am interested in using karamba to evaluate stresses for various surface forms, and then using the results to place more, or larger, structural members where there are the highest stresses. Is there a good example file for how to do this? Or can anyone point me in the right direction? Thanks.  

Hi Adam,

welcome to Karamba,

you can figure out what all you need by checking out the examples below:
http://www.karamba3d.com/category/examples/

by any kind of analysis, if you check the model rendering to figure out the Utilization or Displacement, you can handle the structure you've designed, and absolutely the definition you've written is important also,
so just write down a path-controled defining for your elements (as there are few examples with this solution in the link above) to find out the point you looking for.
Good Luck Buddy!

Hi, I have a problem when I try to install karamba on my pc.

I have Rhino 4, my pc is 64bit  resolution and I have grasshopper (december updated). When I've installed it, the plugin doesn't appear on the grasshopper screen. Why?

Which version should be ok for me? I tried a lot of version of karamba but they did'nt work. Well, before the updating (with the previous version of karamba) it worked.

thank you

GS

Dear Michel Cassagnes (four posts below),

the shell structure you analyze in shell_test.gh is essentially a simply supported beam. If the loads would be placed only at the upper nodes of the structure the first principal stresses on the bottom would be aligned with the boundary there and the second principal stress would be zero and at right angle to the boundary.

The fact that there are loads on all points of the structure makes the first  principal stress in the center of the bottom triangles deviate from the horizontal direction.  

Best,

Clemens

Dear Clemens,

Thanks for your answer. I modified the file to have loads at the upper nodes. Result on princ stress directions is the same. What do I make wrong ?shell_test.gh

Best regards,

Michel

Dear Michel,

thank you for your example. The first principal stress direction of the bottom elements should be horizontal - that is for sure. I haven't found an explanation yet, why Karamba does not display them correctly. If one selects the boundary conditions of a cantilever (shell_test_cp.gh) the directions come out as expected. As soon as there are horizontal or vertical supports on the other end the directions turn. I have to look into that.

Three things regarding your definition: You support the shell at two points only. This leads to theoretical stress singularities there and the solution does not converge with decreasing mesh size.

The second thing is, that the shell elements in Karamba have no in-plane rotational stiffness. In case of flat shell patches it is sometimes necessary to lock the rotation of one node about an axis perpendicular to the patch.

Your shell is very thin (1cm) relative to its span. That is the reason why it does not compute for some mesh settings.

Best,

Clemens

Hey,

New problem in reinstalling Karamba happened, what this runtime error is about? and how can we fix this up!?

thanks in advance!

hi I wanna ask how to apply a surface in "shape optimization irregular structure 5" example which ı draw manually

In order to install Karamba 1.0.3. with the latest release of Grasshopper 0.9.0052 or higher you need to change the Karamba installation path to the folder where Grasshopper resides. This can be e.g. ‘C:\Users\admin\AppData\Roaming\McNeel\Rhinoceros\5.0\Plug-ins\Grasshopper {B45A29B1-4343-4035-989E-044E8580D9CF}\0.9.56.0′ for Grasshopper 0.9.0056.

Best,

Clemens

Hey All,

immediately with no introduction,

i wana calculate optimized support position on a supposed waffle ceiling, adapted to Persian Girih's Geometry,

the definition attached here got involved the beams rounded the base module but i cana define the lines of the girih,

here we have 2 question:

1- how can i make a range out of my item list to be checked by galapagos for selecting the best position of supports?

2- what's wrong with my modeling method (as i know there is something i haven't noticed), that it could not define as beams for the base of girih's module?

P.S: the definition has been written by Paneling Tools in some parts

thanks

surface%20subdivision.gh

Hi all!

I am new to this magic world of Grasshopper and its plugin, so here it is my question:

How can i assign a Curve element such an arc to the "line to beam" panel? It only takes the end points of the curve and makes a straight line..

Or what's the right way to have a curved "finite element" to work on!?

Cheers

n order to install Karamba 1.0.3. with the latest release of Grasshopper 0.9.0052 or higher you need to change the Karamba installation path to the folder where Grasshopper resides. This can be e.g. ‘C:\Users\admin\AppData\Roaming\McNeel\Rhinoceros\5.0\Plug-ins\Grasshopper {B45A29B1-4343-4035-989E-044E8580D9CF}\0.9.56.0′ for Grasshopper 0.9.0056.

for me only worked ...Plug-ins\Grasshopper\Libraries

in case anyone is struggling with the install like me

Hello,

Having an install issue with 0.9.0061, with Karamba version 1.0.4.  It seemed to install just fine, but upon opening GH, the components do not load.  No errors are thrown, it's just not there.  Perhaps I'm missing something simple?

Cheers,

n

Hello Nick,

did you install Karamba in the right directory? It should be the same as that where 'Grasshopper.dll' resides. You can find out the location of Grasshopper by selecting 'Tools' in the Rhino menu then 'Options...', 'Plug-ins' on the left, 'Grasshopper' on the right and the button 'Properties'. 

Another source of the problem could be that the bitnesses of Karamba and Rhino do not match.

Best,

Clemens

Hi Clemens,

I gave your suggestions a shot.  I plugged in the location for GH as noted in the Plug-in properties, but without success.  The GH .dll is in (x86)Grasshopper for Rhino 5, so I tried to re-install there, ans had the same results. 

As a note, I selected the 32 bit Karamba version, since my Rhino 5 is installed as such. 

Thanks again for your assistance.  I'll keep trying ...

n

Hi Clemens,

I’ve been testing Karamba for a while now, trying to optimize a simple form based on its load transfer. Using a cylinder as a basic form, I try to put a horizontal force on it consisting of five circles. I want these circles to scale in a manner (in Galapagos) so I can get an optimized force flow as a result. In this case the cylinder would become wider at the bottom and narrow at the top. Instead of receiving this form I always get a straight cylinder with the smallest radius for all circles. Do you know where I’m going wrong with my settings? In the attached example I minimized the mass and displacement. Is that correct resp. what other settings do I have to make? I’d be very thankful for some input about this.

Marco

diff_radius_scales.gh

Hi Marco,

on first inspection I think there are two possible reasons for the behavior of your structure:

• The value of the horizontal load is constant but the area of the mesh-load increases with increasing size of the structure. Thus a smaller structure means less load.
• You minimize mass and maximum deflection. These have different physical units. Therefore you need to apply a weighting factor to one of them in order to combine them into one meaningful objective value. 

Best,

Clemens

Hi Clemens,

Thanks a lot for the input. I didn’t really know how to apply a weighting factor to mass and deflection so I only changed the loads in order to have the same amount and intensity in any case. This works fine when I only optimize the max. displacement although I’m not sure if the result is 100% correct. Could you please take a look at the attachment one more time and maybe give me another hint of how to apply the weighting factor? Thanks in advance.

Marco

diff_radius_scales_02.gh

Hi Marco,

the necessity of weighting factors is the main shortcoming of single objective optimization algorithms in case you have more than one design target. There is no right or wrong - setting this factor is a matter of trial and error.

You could use e.g. Octopus which can do multi-objective optimization without the need of pre-set weighting factors.

Best,

Clemens

Hi Clemens,

Thanks a lot for your reply. It's good to know that Octopus might be a better tool for multiple optimization. Once again - thank you for your help.

Best,

Marco