Python wrapper for FEM2D program from J.N. Reddy. Use this to generate the problem data cards or alternatively solve straight from python. Postprocessing uses a polynomial fit to smooth the solution.
Shear stress field from a torsion problem.
I highly recommend initializing FEM2DProblemData with the arrays elem_connectivity and node_cords generated with another software. Pass the problem data as a dictionary to **kwargs. After running postprocess, the variables u, qx, and qy can be accessed using get_var.
elem_connectivity = ...
node_coords = ...
problem_data = {...} # dictionary
fea_mesh = FEM2DProblemData(elem_connectivity, node_coords, **problem_data)
fea_mesh.save_card('example.inp') # to produce the data card file
solu = fea_mesh.run() # to run the problem and get the output card as a string
fea_mesh.postprocess()
points_of_interest = ... # points where we want to find fea solution
u = fea_mesh.get_var('u', points_of_interest)
qx = fea_mesh.get_var('qx', points_of_interest)
qy = fea_mesh.get_var('qy', points_of_interest)
Variables follow same names as given in the FEM2D Input Description. Some data is generated automatically when the information is redundant.
I couldn't verify the license for FEM2D by J.N. Reddy, although the program is freely available on the McGraw Hill website. Before running the code, download the .exe and place it in /fem2d.