AW: AW: AW: AW: AW: [Paraview] Combine vect

[Sören Freudiger]

Thank you very much!
That's the point!
After changing the component names from x1-vel to x1vel and switch the
calcation direction from iHat*x1vel to x1vel*ihat everything works fine...


--
Dipl.-Ing. Soeren Freudiger
Institut fuer ComputerAnwendungen im Bauingenieurwesen
TU Braunschweig, Pockelsstr. 3, D-38106 Braunschweig
Tel.: +49 531/391-7595, Mobil: +49 176/210 17 444
email: freud at cab.bau.tu-bs.de
http://www.cab.bau.tu-bs.de/
-----Ursprüngliche Nachricht-----
Von: David Thompson [mailto:dcthomp at speakeasy.net] 
Gesendet: Freitag, 14. Mai 2004 07:13
An: Amy Henderson
Cc: Sören Freudiger; ParaView mailing list
Betreff: Re: AW: AW: AW: AW: [Paraview] Combine vect

On Thu, 2004-05-13 at 13:16, Amy Henderson wrote:
> To my knowledge, there is not a problem with your file format or the 
> magnitude of the velocity components.
> 
> Earlier in this thread, you said that the expression you were 
> evaluating with the calculator was iHat * x1-vel  + jHat * x3-vel + 
> kHat * x3-vel. Was this a typo? It looks to me like you should not be 
> multiplying the same scalar array by both jHat and kHat. If this 
> really is the expression you input to the calculator, it would explain =

> why your streamlines are not doing what you expect them to.
Actually, I noticed a similar problem on the CVS trunk yesterday. To
replicate it, load multicomb_0.vtu, run the "Extract Surface" filter, and
then create an Array Calculator. If you enter
	Result = iHat*Density+jHat*Momentum_0
you'll get an incorrect scalar field. However, if you enter
	Result = Density*iHat+Momentum_0*jHat
you'll get the correct result. For some reason, multiplication by the unit
vectors isn't commutative. I've tried this on two different machines with
recent CVS checkouts and got the same results.

		David