again, "minor" trouble in detect.F90 & Re: [FLASH-BUGS] "minor" glitch in flatten.F90 (part 2)

[Tomasz Plewa]

Markus -

> Yes, my case may be a bit special, it only really occurs at the beginning, 
> were I have a point discontinuity in the center of the domain and this may 
> somewhat "break" the algorithm. With a point it is of cousre important to 
> detect it propperly, wheras with a front the is a broader range.
> Unfortunately I cannot just refine more in order to spread this point out a 
> bit, since refining just leads to this situation occuring earlier in time. 
> Once it is spread out by time evolution it is not as severe anymore, but of 
> course if you "ruin" your data in the first sweep you may end up with some 
> interesting "instabilities" but the whole thing is a bit questionable. I 
> rather add a small pertubation myself once I know that everything works okay 
> ;-)

Right. :)

This was my feeling that we might be expecting a little too much from
the code which in the end, even if formally correct, relies on finite
arithmetics. FLASH is not free from such deficiencies.

Even worse situation plagues some other AMR implementations which
ultimately require extended precision to work properly. This happens
in the limit of high spatial resolution, and by high I mean > 1e7.  I
think actually James Quirk was the first one to explicitly discourage
people from using floating point where integers are not only just
sufficiently accurate but exact.

I should also mention that this spring we became aware of some
inconsistencies in grid geometry in FLASH. Some of them were removed
and corrections are available in FLASH2.3. Artur Gawryszczak, Warsaw,
pointed to me some more very low-level problems and these fixes will
become available in FLASH2.4.

I think that in your case, as you already pointed out, the important
fact is whether the solution reacts in consistent way to the imposed
perturbations. Idealized problems are excellent in identifying formal
errors, or verifying the code, but not necessarily adequate for real
situations.

Thanks for detailed explanation. I am currently traveling in Europe
and will give a closer look into your problem later. But it seems to
me that you took the right approach and your diagnosis is fine.

Cheers -

Tomek
--