AnyFE Tool for Ansys (APDL code generator)

Main Forum
21

Hi Pavel,

thanks for you reply.

I importet the workbench model to classic and looked at my named selections like this:
cmsel,s,pelvis
nlist
cmsel,s,rthigh
nlist
and for both I get different sets of nodes. So I believe that the creation of the named selection via workbench ist working.

Regards,
Patrick

22

What is this code doing exactly?

nodeNext=node(0,0,0)

23

Patrick,

It is searching for the closest node to (0,0,0), which coincides with the origin of the newly defined coordinate system.

Pavel

24

Thanks, OK, thats what I thought looking at the next lines in the code. So “node(x,y,z)” is a command that is not listed in the help.

25

Hi All,
I downloaded ANYFE2APDL converter but it states that " no valid licence found". Any possible solution?
Regards,
Shantanu

26

Hi Shantanu,

have you copied your AnyBody license file into the folder of the ANYFE2APDL executable and renamed it to license.lic?

Best regards
Daniel

27

Hi Daniel,
I got the correct license from anybody tech. support.
As mentioned in the tutorials, i exported the .xml file in the input folder of model,put the AnyFe2apdl.exe in the model folder and stated the working directory; but even after giving the paths for input xml ,output directory and apdl still none of the 5 files got generated !
Regards,
Shantanu

28

Shantanu,

You need to find the correct license file and copy it next to the AnyFE2APDL.exe to make it work. The license that you got from us will most likely be somewhere in:
C:\Users[username]\AppData\Local\AnyBody Technology\AnyBody.6.0.x\license.lic

Once the license is next to the executable - it should work. If it does not - try running it in the MS DOS command line and check what error messages you receive. Then we can help further.

Best regards,
Pavel

29

I am running a script generated with anyfe2apdl on ansys, however, I am getting the message “No *do trips needed etc”. I tried inputting the code bit by bit, and it also returned an error with the line “*get,nlocx,node,nodenext,loc,x”. I guess this means that the line "nodenext=node(0,0,0) is not working properly. What should I do in order to solve this problem?

PS, the file was working correctly, until I had to delete part of my geometry.

30

Hi Carl,

“No *do trips needed etc”. is just a warning, that i get myself, and it is due to the fact that the APDL template was written quite some time ago for earlier versions of ANSYS Classical. Probably something has changed in the APDL itself, but since it still worked we have not revised it.

Regarding the other problem - it is really hard to say what has changed since we do not know your full setup. Have you edited the template code? Line by line is supposed to work too, but it is very easy to make a mistake in such input.

Could you also specify how you use and what version of ANSYS?

And in any case you could always modify template to adjust to your needs. This is why it is a template and not a hard-coded APDL script generated by AnyFE2APDL.

Regards,
Pavel

31

I am using ANSYS Mechanical APDL v14. In the template code, I have only changed the file name to resume so that it refers to my directory. Otherwise I have changed nothing.

Thank you for your fast reply,
Carl

32

Could you reduce the input to a single force (manually deleting all other forces) of a single load step file? try to load it and provide us with the log message so I could help further? It is really hard to say what goes wrong and even though you say node(0,0,0) causes problems - it might be that something else did not work previously.

by the way, there is one more thing that was needed before calling this APDL code:
CSWPLA,1000,0,1,1
this will create a temporary working plane that will be used to return back once the loads are applied.

Regards,Pavel

33

I have loaded the first force found in the file and no errors seem to occur, here is copy of the log:

!File written by the AnyBody Modeling System to transfer forces and moments from a AnyBody Analysis to a FEA model defined in Ansys
!Please do carefully review the input and output of this analysis
!AnyBody Technology: Notice that this file is written in a test format!

!Specification of the base file (meshed geometry without boundary conditions)
RESUME,‘l3MESH’,‘dbb’,‘C:\Users\carmus\Desktop\Final Year Project\FYP ANSYS’,0,0

!Definition of the beam elements connecting the muscle insertion points with the bone surfaces
/prep7
scale =1
RadiusSel = 0.005
LocalCoord = 1000

!Balkenelement elastisch
et,100,4
mp,ex,100,1.1e12
mp,nuxy,100,0.3
r,100,0.001scale0.001scale,(0.001scale)**4/12,(0.001scale)**4/12,0.001scale,0.001*scale

!Turn on Inertia Relief
!Intertia Relief
MP,DENS,100,1.0e-8
irlf,1
FINISH

/FILNAME,loadstep9,0
! Name=Main.Studies.HumanModel.BodyModel.Trunk.MusclesSpineRight.Multifidi.MFdL1L3
forceNr = 0
posx = 2.194455942031337e-02
posy = 3.087541765673434e-02
posz = 2.194455942031337e-02
forceX = -6.057634122498727e-01
forceY = 7.442179400505726e-01
forceZ = -5.746820507817055e-01
momentX = 0.000000000000000e+00
momentY = 0.000000000000000e+00
momentZ = 0.000000000000000e+00

Thank you for your kind help

34

So does that mean that the problem was resolved? Because the code looks OK and it is not helping to resolve other problems. What I meant by ‘log’ was the output from ANSYS. Could you isolate the problematic case?

Regards,
Pavel

35

I have resolved the problem. However, now, when I go to solve, the solver says that there is rigid body motion and that my model is unconstrained. I checked the coding and reactions from joints seem to be fine. should I specify some boundary conditions such as no nodal displacements in respective directions?

36

Carl,

is that a single bone simulation? If so you could apply a kinematic constraint on one of the nodes that were created for the force application. You should observe very small residual reaction force that can be attributed to the numeric errors.

If it is a multibody case - the kinematic constraint should be on one of the objects.

I just realized that we have updated the APDL template a little. I will attach it here.

Pavel

37

Thank you for your kind help. I am doing a single bone simulation, however I have plans to extend it to multiple bones eventually, and this information should be very useful. Regarding the kinematic constraints, if I apply them on a node where a force is acting, won’t that make the simulation inaccurate? Or would it be to a very small extent?

Regards, Carl

38

Carl,

what i meant to say - there should be some 0 magnitude forces (typically not all muscles/ligaments are involved in a certain task). And i typically choose those to impose the constraint. Otherwise you could choose a node on the surface, but that might add some local inaccuracy around the selected node - just like you say.

Regards,
Pavel

39

Yes ok, I noticed some of these forces. What constraints should I apply then? displacement or rotational? or should i constrain all dof to zero?

Regards,
Carl

40

Hi Carl,

I constrained all 6 dof. But as I said you should observe very small residual reaction forces and moments in this node - so it might be possible to provide less.

Pavel