Follow

Streamlines stopping prematurely

The streamlines will stop when one of the following conditions is met:

  • the velocity field that generates the streamline is zero;
  • the total time integration is larger than T_max;
  • the number of segments that determine the streamline is larger than the max. value.

If you observe that your streamlines stop when you would not expect them to stop, check the followings: 

a) "Emit" options in the Streamline menu. Check the "total time limit" specification, and try increasing it by 2x or 10x. 
b) Parent part(s). Check that all of the fluid domains are selected as the parent to the streamline.
c) Are the streamlines stopping at the boundaries between fluid domains? If yes, then some air gaps between the fluid domains could be causing the stop.
d) Color the streamlines by velocity, and see what the velocity looks like towards the end of the streamline. Does the velocity drop to a very low value, basically causing the streamlines to stop?
e) Take clip planes through the domain near where some of the streamlines stop to see if there are some type of discontinuity in the velocity field.
f) Each particle trace has a maximum number of segments. The maximum number of segments in each massless trace is by default 5000 for surface-restricted traces, and 6000 for all other massless trace types. This default can be changed by entering a command line entry (File>Command) into the command dialog, using the following commands:

- for Surface restricted:
test: max_skin_segments <value>
- for all other trace types:
test: max_ptrace_segments <value>

g) Check the composition of elements in the domain : hex, pyramid, polyhedrals. It has been reported in EnSight 9.X problems with stopping streamlines in correspondence to polyhedral elements. Set the environmental variable ENSIGHT_NFACED_METHOD = 0, and try the streamlines again.

Was this article helpful?
3 out of 3 found this helpful
Have more questions? Submit a request

Comments

  • Avatar
    Prashanth Rao

    Thanks a lot! I faced the exact same problem with a Fluent dataset and step g) worked PERFECTLY. Setting ENSIGHT_NFACED_METHOD = 0 seems to be the solution when streamlines don't plot in the fluid domain of interest. Setting it to zero uses centroidal values so I guess this allows the streamlines to be integrated over the cells rather than the nodes of the wall boundaries.