[FLASH-BUGS] bug in hydro flash2.3

[Tomasz Plewa]

On Tue, Jun 10, 2003 at 05:40:39PM -0500, Mike Zingale wrote:
> Markus, I believe that that loop in interp is too big, and should only
> have to go from 2 to nzn6 (maybe nzn5 actually).  I will look into this.

The loop in question should run from 2 to nzn6. This preserves
symmetry between available date to near the left and near the right
end of the stencil.

In case of pure hydro PPM requires profiles only for zones from 3 to
nzn5 (one below the leftmost interface and one above the rightmost
interface). This may vary depending, for example, if one wants to
interpolate potential (instead of force) and get accelerations by
differentiating interpolants.

I do not think it is the case in FLASH so in principle one can narrow
construction of interpolants to the interval [3,nzn5]. I did not try
it, though.

Tomek
--

> 
> ------------------------------------------------------------------------------
> Michael Zingale
> UCO/Lick Observatory
> UCSC
> Santa Cruz, CA 95064
> 
> phone:  (831) 459-5246 
> fax:    (831) 459-5265
> e-mail: zingale at ucolick.org
> web:    http://www.ucolick.org/~zingale
> 
> "Once something has been approved by the Government, It's no longer immoral."
>    -- Reverend Lovejoy
> 
> 
> 
> 
> 
> 
> 
> On Tue, 10 Jun 2003, Markus Gross wrote:
> 
> > Hi!
> > 
> > I presume my earlier mail now went through fine.
> > 
> > On another topic (maybe I am lucky and I catch two birds with one stone ...):
> > 
> > 
> > line 96-line 100:
> > 
> > !!****f* source/hydro/explicit/splot/ppm/interp
> > ! assemble the left and right interface values using the quartic polynomial,
> > ! Eq. 1.6 from Colella & Woodward
> >     
> >     do i = 2, nzn6+1
> >         ar(i)  = a (i) + coeff5(i) * scrch1(i+1) + coeff3(i) * dela(i+1)
> >        ar(i)  = ar(i) + coeff4(i) * dela(i)
> >        al(i+1)= ar(i)
> >     end do
> > 
> > 
> > This causes a runtime error. This one was bothering for quite a while, but it 
> > is so tempting to avoid errors but simply ignoring compiler flags ;-)
> > 
> > when I look into:
> > 
> > !!****f* source/hydro/explicit/splot/ppm/coeff
> > 
> > I find that coeff5(nzn6+1) and coeff3(nzn6+1) are actually not computet!
> > 
> >     do i = 2, nzn6
> >        temporary = 1. / ( scrch1(i) + scrch1(i+2) )
> >        coeff3(i) = -temporary * dx(i)   * scrch1(i)   / scrch3(i+1)
> >        coeff4(i) =  temporary * dx(i+1) * scrch1(i+2) / scrch2(i+1)
> >        coeff5(i) = dx(i) - 2.e00 * (dx(i+1) * coeff3(i) + dx(i) * coeff4(i))
> >        coeff5(i) = coeff5(i) / scrch1(i+1)
> >     end do
> > 
> > and hence not initialized!
> > 
> > I will dig out my copy of Colella & Woodward tonight still, but any hint is 
> > of course appreciated.
> > 
> > Regards,
> > 
> > Markus.
> > 
> > _______________________________________________________________
> > 
> > Markus Gross AMIMechE BEng (Hons.) Mechanical Engineering
> > 
> > Heriot Watt University Edinburgh
> > School of Engineering and Physical Sciences
> > James Nasmyth Building
> > Edinburgh
> > EH14 4AS
> > UK
> > 
> > Member of  IMechE, SPIE, CSME and VDI
> > _______________________________________________________________
> > 
> > further contact:
> > 
> > Phone   : +44 (0) 131 451 4737
> > 
> > UNiX talk: talk markus at lasersim.mce.hw.ac.uk
> > 
> > _______________________________________________________________
> > 
> > "Plans are a place to begin," Grove said. "They rarely deliver
> > you to where you expect. Make your plans knowing you are going
> > to throw them away."
> > 
> > _______________________________________________________________
> > 
> > 
> > 
> > 

--