The following steps will assist in modeling CNT for force analysis:

[1] Generate the model of carbon  nanotubes using appropriate software, e.g. Material Studio. Nanotube Modeller is also useful. Demo (free) and full versions are available. For more information, visit: http://www.jcrystal.com/products/wincnt/. VMD program can also be used. To download it for free follow this link  http://www.ks.uiuc.edu/Research/vmd/

[2] Export the structure as PDB file. Visit this link for a script to convert PDB files to ANSYS APDL code using WING Ide (Python): 

http://machine-phase.blogspot.com/2009/11/convert-protein-data-bank-files-into.html. Also check this published article for more details: http://www.jourlib.org/paper/3112745

[3] Extract the xyz coordinates for each atom and the connectivity for all bonds.

[4] Use the data in [3] to develop a macro in ANSYS to build the structure.

[5] Simulate using appropriate material properties and boundary conditions.

It is not possible to fully model the CNTs in ANSYS. There are several reasons such as: a) CNT is highly elastic material (more than 30% of elongation without structural deformation). b) Normally CNTs are highly curvilinear structure.  c) In the percolation network, CNTs are randomly folded and entangled. That's why researchers tried to simplified the structure of CNTs as capped cylinder, very similar to metal nanowires. Check our conductivity simulation of nanowire networks:

https://www.researchgate.net/publication/261767953_Highly_Stretchable_and_Sensitive_Strain_Sensor_Based_on_Silver_Nanowire-Elastomer_Nanocomposite

Article Highly Stretchable and Sensitive Strain Sensor Based on Silv...

The following steps will assist in modeling CNT for force analysis:

[1] Generate the model of carbon  nanotubes using appropriate software, e.g. Material Studio. Nanotube Modeller is also useful. Demo (free) and full versions are available. For more information, visit: http://www.jcrystal.com/products/wincnt/. VMD program can also be used. To download it for free follow this link  http://www.ks.uiuc.edu/Research/vmd/

[2] Export the structure as PDB file. Visit this link for a script to convert PDB files to ANSYS APDL code using WING Ide (Python): 

http://machine-phase.blogspot.com/2009/11/convert-protein-data-bank-files-into.html. Also check this published article for more details: http://www.jourlib.org/paper/3112745

[3] Extract the xyz coordinates for each atom and the connectivity for all bonds.

[4] Use the data in [3] to develop a macro in ANSYS to build the structure.

[5] Simulate using appropriate material properties and boundary conditions.

You can consider an RVE for each Gauss point of the elements that are used in your ANSYS model. Then you need to determine the elements of fourth order stress-strain tensor for each RVE using a software such as LAMMPS and use these values for the elements in your ANSYS model.

Mohsen