Thanks for the explanation. It's clearly the data component, then. With some vector series, I'll get the problem on specific time steps but not on others. In each case, the positions (irregular) and connections (regular cubes) are the same, just new data. There are no degenerate elements. But now I can at least try to isolate the situations that lead to the error.
Please respond to [email protected]
Sent by: [EMAIL PROTECTED]
To: [email protected]
cc:
Subject: Re: [opendx-users] problems with Streakline
It means that the streakline thinks it has stepped into a tetra, but when
it tries to find the exit point from the tetra, it can't. (If your data is
cubes, it is decomposed into tetras under the covers). It considers each
face in turn, and determines the cosine of the angle between the face
normal and the vector. If the cosine is <= 0, then the vector diverges
from the face, and it isn't considered further. If not, it scales the
cosine by the distance between the point and the face to get the necessary
step along the vector to reach the face in the vector direction, and then
chooses the face requiring the smallest step. Problem is that all the
cosines are negative (or zero), so the vector diverges from *all* the
faces. Which means that there's a problem in the data, maybe a degenerate
element or something. Or more likely, floating point error is just making
it look that way.
Greg
- Lloyd A
Treinish/Watson/[EMAIL PROTECTED] To: [email protected]
Sent by: cc:
[EMAIL PROTECTED] Subject: [opendx-users] problems with Streakline
son.ibm.com
05/30/2002 02:17 PM
Please respond to
opendx-users
"Internal error: can't find intersection".
Other time steps or starting points with the same vector series work ok.
With different data sets the problem may or may not appear. Before I start
looking at the code, does anyone know the origin or meaning of this error?
Thanks.
