Dynamic Response - Large deflection

protontim

I am currently doing a 3D non-linear analysis on a rotor blade. Non-linear is needed because the tip deflections are large and the centrifugal force offers a large "stiffening" force that increases with blade deflection, due to the increase in the angle between the deflected blade and the centrifugal force (the centrifugal force is computed elsewhere and entered in the model as a static force ). The simulations work very well, but now we need to add dynamic lift loading to the blade. I plan to use the 3D dynamic response, and I was wondering if the 3D dynamic response will still handle the large deflections and the varying angle of the static centrifugal force properly? Thanks for your help.

Victor

Hello protontim.

No, the dynamic response analysis type in Mecway is currently linear only. With v3, you could export to .inp and solve with CalculiX though. In that case you'd probably need to make some edits to the .inp file manually but the mesh, material definitions and some load types would get exported automatically.

In case you just need to use a fixed, deformed blade for dynamic response, you could use Mesh tools -> Transfer displacements from solution to deform the mesh.

[Deleted User]

hi victor,

i'm just curious, being new to mecway, is the transfer to displacements option new in v3 or did it exist in early versions of mecway? i am trying it out in the v3 beta and it is awesome.

Victor

Oh yes, it's new to v3.

protontim

Hi Victor,

I exported the file to a Calculix .inp file. Mesh and material came over fine, but it looks like the loads are not making it over. Working on coming up to speed on Calculix. Thanks.

protontim

Victor,

I realize this is beyond the scope of this forum, but any help would be appreciated. I ported the file to Calculix but I can only display displacement. Stress is not being saved to the file. Any ideas why this might happen. Below is the STEP section. "ext_force(10)" is defined earlier. Deflection appears to work properly, I just can't get stress data to be saved.

*STEP,NLGEOM,INC=100
*STATIC,DIRECT
1.,1.
*CLOAD
ext_force(10),2,100
**NODE PRINT,NSET=NALL
**U
**EL PRINT,ELSET=EALL
**S,E

*NODE FILE
U
*EL FILE
S,E
*END STEP

protontim

I looked at the .frd file and the stress data appears to be there. It just does not show in the Calculix Graphix window. I'll try and track down why.

protontim

Figured it out, sorry to bother you.

Victor

Glad you solved it OK.

protontim

Victor,

Thanks for pointing me to Calculix. I wrote a script to generate my mesh and my amplitude profiles. The input deck turned out to be about 47,500 lines. The simulation seems to be yielding reasonable results although the graphical display I am using is quite quirky. Any chance you can add dynamic non-linear to the Mecway package?

VMH

Victor, is it possible to incorporate Calculix solver into Mecway so we can view the analysis results from Calculix solver in Mecway instead of viewing the analysis results in Calculix graphix? That way, we can also have other capability like material non-linearity plus a lot more all in Mecway. I tried Autodesk Simulation trial version (didn't like it as much as Mecway), and we were able to select between simulation previous solver or their new incorporate Nastran solver.

Victor

Hello VMH

It can't yet import results from CalculiX but I expect to have that in v4. There is an option in the labs menu for solving directly from Mecway though. This just streamlines the process if you don't need to make manual edits to the .inp file.

Tools -> Options, Tick the "Labs" check box in the bottom corner.
Tools -> Labs -> CalculiX solver
Tools -> Options -> CalculiX, fill in the form.

Now the Solve button also has a CalculiX option

Victor

Protontim, I don't plan to add a dynamic nonlinear solver, but instead further integrate with CalculiX to take advantage of that. It should become smoother in the next version.

I realize the amplitude tables can be saved quite inefficiently (duplicated for each node, etc). Do you think this will cause problems in your work?

VMH

Victor, thanks. I got it to run using CalculiX and results were shown Calculix screen.
For v4 if Mecway can read the results of CalculiX, can we be able to define material non-linearity? I bring waiting for the feature because most of the time when vonMises stress exceed material yield, the maximum stress is no longer real for linear material. With material nonlinearity, we can say what is really the max. stress

Sergio

VMH, normally for nonlineal material simulations what we postprocess is strain and not stress. When you reach the nonlineal of your material the stress value will keep almost constant for the next increases of load (or displacement), but strain will increase

Regards

VMH

Sergio, thanks. I understand that. What clients asked is how much did it overstressed instead of how much is the plastic strain. I usually tell them the stress and the strain together.

Sergio

VMH, can I ask you why you want material nonlineal in your simulation? I meant, if you must desing your part for normal use lineal analysis should be enough, I switch to nonlineal when my customers ask me for the behaviour of the part after abusive or crash loads

VMH

Sergio, on most of my works, it's not new design. It's more like investigation of a certain connection to withstand existing loads and what if we have more loading to it. At that point, the connection stresses may be well over yield. Then we report the results to the client showing that it's no good for them to make a decision. We use FEA consultant to assist on these works. I was trying to see if Mecway can generate similiar results and for my own learning curves.