Dear Hamid,

Thanks a lot for your answer! One more question, for some hyperelastic material such as rubber or artery, the sectional property is not constant but a function of the axial loading. Now we can figure out this kind of beam stiffness by VABS. But how can I implement this hyperelastic beam stiffness in 1-D FE analysis in ANSYS or ABAQUS?

Fang

Dear Hamid,

VABS has a nonlinear version now which provide nonlinear beam-like structural constitutive relation. Please take a look at the attached axial force-strain curve for a simple-extension test. What I am trying to do is to implement this relation directly into ANSYS or ABAQUS, which is actually my question, that it, is there a method to do it in this way? The reason for this intention to investigate the effect of nonlinear sectional properties in the 1-D global analysis. 

Conference Paper Non-linear Sectional Analysis of Composite Beams with Finite...

Dear Hamid,

One more question: for the 1-D ANSYS or ABAQUS simulation, after I provide the stress-strain data of the material and the geometry of the member, does the software consider the sectional deformation as the simulation shown did in my previously attached graph? 

It can be seen the sectional deformation is huge when I use hyperelastic material and the sectional area is changing during the deformation, and it will have impact on the global beam behavior.

IF you turn "large deflection" option on in ansys, it takes the effect of poisson's ratio on the cross-section (i.e. it changes the volume and area  of elements depending on the strains, even for 1D elements. 

Dear Mahmud,

Thanks a lot for your answer!

Question 1: By "large deflection", you mean the nonlinear option of the solver, right?

Question 2: I am wondering, if the 1D elements counts the sectional deformations (including in-plane and off-plane ones), it should result in the same accuracy in the global beam displacement field (extensional strain, twist, and curvatures), when the same magnitude of loading and same type of boundary conditions are applied, right? However, this is not our observation. In addition, some very famous nonlinear beam phenomena, such as Poynting effect,  trapeze effect, and Brazier effect, can not be simulated by 1D element in ANSYS or ABAQUS. Am I right in this point?

large deflection is indeed activating nonlinear geometric effects in the solver. 

However, i think what you are looking for cannot be solved using the 1D elements found in ansys, because these elements at best employ first-order shear theories (basically, Temoshenko beam elements). You would need higher order beam theories to account for all those effects, which as far as i know are not available in ansys. Best think you can do then is to try it with solid elements. Good luck.

Dear Mahmud and Hamid,

We have developed a validated model to do the beam sectional analysis including both the geometric and material nonlinearities, which means the sectional warping (including lateral contraction and Saint-Venant warping) are finite and also means the material model is not unique (actually we have checked Hencky strain energy model, neo-Hookean model, Mooney-Rivlin model, Saint-Venant Kirchhoff model...). And the Poynting effect, trapeze effect, and Brazier effect can be observed in our nonlinear sectional analysis.

Now what we are trying to do is to integrate the results of our nonlinear beam sectional analysis (that is, the nonlinear sectional stiffness which is not constant anymore) into 1-D global beam finite elements, in order to see the global beam behavior with such nonlinear sectional properties.

So I would like to confirm it again: is there a way to do this in ANSYS or ABAQUS?

Thank you very much for your answers!

Best Regards,

Fang 

Dear Fang, If you have the internal behavior already derived in terms of nodal behavior (i.e. stiffness and mass matrices, etc.), you might want to look into manipulating user-defined elements . Ansys offers many elements that can be manipulated  "or custom defined" like userelem.f  and user300 elements in ansys, where in you can put in stiffness matrix and integration points..etc.. This requires good skills in programming with Fortran.

There is also the element "Matrix27" in ansys where in you can input 12x12 stiffness matrix. And then define each index of these 12x12 stiffness terms as "real constants" . However, to make things nonlinear can be more tedious, either by looping and updating the matrix at each load step, or go back to user-defined elements wherein you have more control over what happens inside the element.

Good luck

Thanks  a lot for all of your helps and answers!

I personally feel what Mahmud indicates is the right direction on solving this problem!

Again, than you very much!

Dear Fang and Mahmud,

I read your conversation and I have the same problem,

I have two nodes in Ansys (two remote points connected to many nodes) and I need to use MARIX27 to define the stiffness between these two remote points, can you please help me? How to use this type of element in workbench where I need to wright script in a command? And how to specify the two nods which we will define the stiffness between them?

I would be appreciate your help.

Dear Yasir,

I personally think APDL in classic version of ANSYS may assist you more. In fact, I have no experience in workbench version....

Dear Yasir,

Another way to do this is to use the code of GEBT (Geometrically Exact Beam Thoery) in cloud, of which the website link is provided below.

https://cdmhub.org/

Dear Fang,

Thank you very much for your replay and for your information, actually I’m not familiar with APDL, so can you please write me a brief outlines on how to define the stiffness matrix for the beam section in APDL?

It would be very helpful if you have a file already defined the matrix, then I can open it in workbench and understand how it works.

Or even if you have some files of GEBT.

You can send me on my email: [email protected]

Thank you for your time and I appreciate your help

Dear Yasir,

Sadly I haven't got a chance to try the matrix27 in APDL. But in the near future, one of our master students will visit this issue. I will let you know then.

Now, to solve this problem, I think GEBT is a better way.

It is a free and open-source stuff now. 

What you might want to do is to register an account in the cdmHUB which is funded by US government and managed by Purdue University. And search GEBT as a tool on the cdmHUB. Then what you need is there~

You can also join in the group of VABS which is a very advanced beam sectional analysis code. Many scholars are changing their ideas on cdmHUB~

Good luck~!

Dear Fang,

Thank you very much for your replay, actually I need to use Ansys because I’m doing a fluid structure interaction (FSI) for a wind turbine blade.

I would be appreciated to inform me if you get in progress with your master student

Thanks!

Dear Yasir,

Sure! Let's keep in touch!

Best Regards,

Fang JIANG

Dear Sir

Now, I want to apply Matrix 27 element to shaft analysis. Specifically, the shaft is connected to a bearing but there are many balls in bearing, it makes some problem when analysis so I want to replace its ball by Matrix 27 element. Is there anybody know how to apply it? 

Thank you very much