cpraveen / juliafem Goto Github PK

This code is a rewriting of the matlab code from "Remarks around 50 lines of Matlab: short finite element implementation"

http://link.springer.com/article/10.1023/A:1019155918070

https://www.math.hu-berlin.de/~cc/cc_homepage/download/1999-AJ_CC_FS-50_Lines_of_Matlab.pdf

in the julia language. For the matlab version, see

https://github.com/cpraveen/fem50

Add plotting solution for quadrilaterals. The matlab code uses trisurf which can plot triangles and quadrilaterals. There is no equivalent function I could find in PyPlot. I use contour plots in the Julia code through the tricontour function from PyPlot. If you run the example in "sample", it plots quadrilaterals by triangulating them.

You need to install the PyPlot package to plot results.

import Pkg
Pkg.add("PyPlot")

First, generate the mesh

bash> cd square

Generate mesh by running square.m in matlab (julia version not completed)

matlab> square(30)
matlab> quit

Run the fem code

bash> julia run.jl

You should get a solution like this

The simplest way to initialize an empty nxn sparse matrix is

julia> A = spzeros(n,n)

This makes use of 64 byte integers and floats. For small problem sizes, 32 byte integers are enough for indexing, in which case you can do

julia> A = sparse(Int32[], Int32[], Float64[], n, n)

This was my first attempt at writing a Julia program. Comments, feedback or criticism to improve this code is most welcome. Please email me at [email protected]