Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

IFCHAR = T ?? #100

Open
fischer1 opened this issue Jan 27, 2018 · 3 comments
Open

IFCHAR = T ?? #100

fischer1 opened this issue Jan 27, 2018 · 3 comments

Comments

@fischer1
Copy link
Contributor

I notice that several examples have ifchar=T.

Some don't even use a dt big enough to warrant such usage (e.g., cyl_restart case).

I think all examples (except for short tests designed to test Characteristics) should have IFCHAR F unless we really want to show off CFL >> 1.

Otherwise, jobs will run twice as slow. People often start with one of the examples and modify it to suit their needs without realizing that IFCHAR=T.

(This happened in one of the large KTH runs that George did for pipe flow. He was running with CFL=0.5 and IFCHAR T. Turning that to F sped the code up by 2X and would have saved a million core hours---it was my fault for not paying attention earlier that time.)

Can we / should we switch all the examples to have default IFCHAR F (since the decision
to use Characteristics is typically a more advanced-user decision) ?

@stgeke
Copy link
Contributor

stgeke commented Jan 29, 2018 via email

@fischer1
Copy link
Contributor Author

Yes - we definitely want examples that demo ifchar.

I was just surprised (and I'm sure it's my fault) that ext_cyl had IFCHAR T yet the CFL was 0.5

So, I grepped "IFCHAR" in all the /.rea files and saw that there were other places
where it was true. Given that the overhead his potentially expensive, we probably
want to notify users which cases have IFCHAR T --- turbchannel is a good one.

@adampep
Copy link
Contributor

adampep commented Jan 16, 2019

Right now there are 5 examples using characteristics on purpose: Pipe_Stenosis, TurbJet, TurbChannel, TurbPipe and Vortex. I removed the rest of the flags except Os7000, as in this case delta in the test is so small, that the test crushes even though cfl=0.475. What was the source of the test value in this case?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants