kaarthiksundar / gastransim.jl Goto Github PK

If we can change delta x for each pipe, then ideally, can decrease problem size. Of course, will need to rework code, calculate local delta t for each pipe and store it. Problem -- different edges at a junction will end up having different time steps. Solution -- find min delta t over all edges and use that for uniformity. Can think of then setting delta x to be length of each pipe, so that each edge has ends (junctions) as dof, each junction is one dof.

On further reflection, not sure if this feature is a good idea. Right now we choose a large time step that imposes some constant spacing satisfying CFL for all edges. If we instead have non-uniform spacing that imposes a common time step, need to evaluate which is more efficient or faster.

can think about having new boolean field in ic.json file that indicates whether initial conditions are given as constant values or a profile.
This can inform the parsing routine on how to handle it when read at start of simulation.

maybe based on a flag in ic.json, read scalar or array of values. Then initialize pipe state depending on whether ic is scalar or array.

get text files of solutions from AZ and VG. Read in and construct interpolant. Then interpolate at given solution time points and compute error norm. Can set up success or failure of unit test based on this norm value.

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!

@JuliaRegistrator register