Finite element method solver for heat conduction problems on flat plates with the following geometry:
top h = 0.5 * (a*x*x + b*x + h1)
________________________________
| |
| |
| |
h1 or left height | | h2 or right height
| |
| |
|________________________________|
bottom h = -0.5 * (a*x*x + b*x + h1)
To build Heat++ run make from the root directory of the source distribution.
In order to be able to run Heat++ from the command line you should include the
folder "/path/to/Heat++/bin" to your $PATH environment variable.
To run Heat++ the following parameters have to be specified:
* -A: constant a for defining height of the plate.
* --left-height: h1 dimension of the plate.
* --right-height: h2 dimension of the plate.
* -L: length of the plate, from the left to the right edge.
* -T: thickness of the plate.
* --k-xx: conductivity in the xx direction.
* --k-xy: conductivity in the xy direction.
* --k-yy: conductivity in the yy direction.
* --n-x: number of elements in the x direction.
* --n-y: number of elements in the y direction.
* --flux-location: location of flux boundary condition.
* --flux-value: value of flux boundary condition.
* --temp-location: location of temperature boundary condition.
* --temp-value: value of temperature boundary condition.
An example is here presented:
Heat++.out -A 0.0 --left-height 1.0 --right-height 1.0 -L 2.0 -T 0.2 \
--k-xx 250.0 --k-xy 0.0 --k-yy 250.0 \
--n-x 10 --n-y 5 \
--flux-location top --flux-value 2500.0 \
--temp-location bottom --temp-value 10.0
The solver can also be run with make c1, make c2 and make c3, which run
the code for three test cases.
The solution is printed in the file disp.vtk. This file can then be plotted.