Author |
Message |
Fred Sachs (sachs)
Member Username: sachs
Post Number: 6 Registered: 01-2004
| Posted on Sunday, March 07, 2004 - 11:53 am: | |
Significant differences in solutions between v3 and v4 !!!!! Opposite signs among other thing. There are two files attached, one that runs ini v3 and one that runs in v4. The v4 file was copied directly from v3, and the only changes I made were to add "variable name:" before each equation and to move the boundray value statements below the geometry specification. You will see that the velocity fields are entirely different. |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 122 Registered: 06-2003
| Posted on Sunday, March 07, 2004 - 03:26 pm: | |
The problem is that you have overcompensated for version 4. The syntax change from v3 to v4 is merely that the boundary condition for the first segment of the path must be inside the path, not outside. That is, the BC must follow start(). You have instead interpreted this to mean that BC specifications must follow the path segment to which they apply. This is not what the documentation says, and is not what is flagged when you run a v3 script on v4. You have made too many changes, and rotated all your BC's to the next boundary segment. The equation tag and the BC inside the START(), which are required by v4, are also supported by v3. So, except for the RANGE clause (which is not necessary in steady state), the same input can run in v3 and v4. In the attached file, I have started from your Valve1_NS-253.pde, which runs on v3, and modified it to run on v4. The results are the same in both versions. |
|