Tenkai.jl

1d/2d FR solver

Setup (5-10 minutes)

Set parameters in any of the run files, for example run_const_linadv1d.jl. Some of the options we have are

Scheme: RKFR, LWFR

Correction: radau, g2

Degree: 1,2,3,4

Then, enter the Julia REPL as

julia --project=.

or by starting plain julia REPL and then entering import Pkg; Pkg.activate("."). Install all dependencies (only needed the first time) with

julia> import Pkg; Pkg.instantiate()

For the first time, to precompile parts of code to local drive, it is also recommended that you run

julia> using SSFR

At this point, in principle you can exit the REPL and always run your code directly through the shell as

julia --project=. run_const_linadv1d.jl

However, since Julia uses a Just-In-Time (JIT) compiler, this is not recommended. If you wish to run this file again or run a different file, Julia will have to compile most of the code again. The recommended way to use Julia is within the Julia REPL where, within each session, the compiled code is stored in memory.

Usage

Assuming you have already selected the run_const_linadv1d.jl file and modified parameters if needed, within the REPL, run it as

julia> include("Examples/1d/run_const_linadv1d.jl")

Visualization

For 1-D, you can see png, .txt and interactive HTML files of the final solution in output directory.

For 2-D, plot the solution using visit as

visit -o output/sol*.vtr

Printing to screen is slow, you can redirect the output to a string as

julia> using Suppressor
julia> out = @capture_out julia run.jl > log.txt # Capture output in variable out
julia> println(out)                              # Print output if needed

If you have a 4 core CPU, you can use 4 threads by starting REPL as

julia --project=. --threads=4