Learning Forum

[OguzhanA]

Have you modeled a concrete like in the picture before? (on workbench or apdl)

[Sean Harvey]

Hi OguzhanA,nIn the other forum post (https://forum.ansys.com/discussion/comment/96549#Comment_96549) this was brought up and we discussed reinforcements, so I'll comment on another method here where we have many fibers.nYou commented there will be many reinforcements in random directions. One could consider a micromechanical approach. One option is to use Ansys Material Designer to compute the homogenized material properties of the material. It can include nonlinearties in that model to generate non-linear stress-strain curves. You would not be making a model of the entire structure (macro scale) but rather a micromechanical model. This could help you understand the stiffness behavior (including nonlinear effects)nNow, In fact the Ansys MAPDL database model that Material Design generates can be used for further studies, so if say you wished to just study how the material behaves in bending, or you wished to add more mesh, parts. The nice thing is you can quickly create the models using Material Designer. Keep in mind Material Designer is a templated version of SpaceClaim (for easy and automation) and Ansys APDL solver running in the background to perform homogenization. It will not permit you do load in arbitrary directions, so if you loading had arbitrary loads, etc. It is really to study material behavior, not macroscopic model scale behavior, but again, maybe the fact that it is creating these complex geometries you could leverage that.nI still believe the reinforcements might be a good option, even with thousands of reinforcements, one could script in APDL to create the beams with random orientations to embed in the matrix, so getting to understand APDL could be highly beneficial.nnThere are some 3rd party tools that use Ansys as the solver and our partners have created such as.nSome of these will do the localization which basically is to impose the macroscale deformations back on the micromechanical model to better predict failure.nI hope this helps and others are welcome to comment on this thread.nnRegards,nSeann