An automated OpenCV tool to detect circular patterns in gray-scale images subject to over-segmentation and generate unstructured triangular meshes for use in finite element analysis.
This is still a work in progress.
Follow the instruction in these links in order to intall OpenCv and FreeFem++ as well as Cmake. Launch the Cmake build process once the program is installed :
Cmake .Written by: D.Sousa-Pontes, A.Dakri
In order to use the program to generate the particale detected mesh on the image Img.png:
./launsher.sh Img.pngExample of use
The geometry detection tool in action:
Topology detection:
The result mesh:
Written by: M.Ettaouchi
A C++ Eigen implementation of different linear sparse matrix solvers for use in a finite element composite linear elsticity problem. The algorithms implemented include: Gradient conjugate, Minimal residual, Optimal step Gradient, FOM and GMRES as well as an SDP-modified GMRes variant.
Edited by: A.Dakri The automatic geometry and meshing tool was used in a linear elasticity simulation of Carbon/Carbon composite using the finite element method, source code in the /FEM directory.
This is mostly a verified work, at least numerically. The user interaction is reduced to a datafile.txt where all the simulation parameters are defined. To change the studied problem to another field of study (thermoelastics, magnetics, rupture mecanics, CFD ? ..) the main file to modify is Model.cpp, the boundary consitions are mainly constructed in InitialCondition.cpp. For more advanced users, the liberty is all yours to discover/modify/erase or develop any part you seem fit.
Example of Results
Applied efforts on a fiber (model to a circle in a background of a carbon matrix):
Example of post processing using the appended .vtk paraview tool:
