Well, if you choose your starting points that way you may get a *lot* of
streamlines - one for each vertex in your velocity field. There are
various tools to select starting points intelligently.
Just to clarify, the Streamline module receives a vector field that is in
essence F(x,y,z) = (direction scaled by velocity). To track a streamline,
it takes the current point, advances it a small distance (determined by the
velocity and the size of the element in which the current point resides)
in the direction interpolated from the vector field, and creates a new
current point. It also computed the curl along the steps of the
streamline. Output is a sequence of points along the streamline, a "time"
component that indicates the elapsed time along the streamline, a "data"
component containing the tangent to the streamline, and a "binormals"
component, whatever that is.
The domain of the vector field is not assumed to wrap - a streamline that
exits the vector field stops.
Greg
"Frederick R. Phelan Jr."
<[EMAIL PROTECTED]> To:
[email protected]
Sent by: cc:
[EMAIL PROTECTED] Subject: Re: [opendx-users]
Streamline Module
son.ibm.com
07/08/2002 11:07 AM
Please respond to
opendx-users
Hey,
Thanks for information, I appreciate it. In the days between the original
post and this, a co-worker who is much more a dx guru than I, managed
figured out much of what you detail for me. The main problem I was
experiencing was due to the fact that I have time series data, and I needed
to tell it to grab only one series at a time. Once we got beyond that,
everything else fell into place. We pass the velocity field to streamline
in the first input. For the starting positions, we "Mark" the "positions"
of the data stream coming out of the import box and use that. Seems
reasonable.
Only problem is that the streamlines do not all connect back on one
another, which is not quite what I am used to. In quiet regions of the flow
field, some of them look like tracer particle paths that just stop. I am
not sure if anything can be done about this. Maybe you could comment ... I
looked on the streamline manual page and it seems that the algorithm it
uses to calculate the streamlines is to calculate particle paths in space,
based on the velocity field at a point, rather than computing the curl, and
calculating the streamline contours from the vorticity.
I'll try your suggestions in regard to Ribbon and Tube ...
Thanks!
Fred Phelan
At 10:12 AM 7/8/02 -0400, you wrote:
>
>Very little. It must be a valid velocity field - eg. have 3-D positions,
>elements (cubes or tets), and vectors, or 2-D positions, elements
>(triangles or quads) and vectors. If so, you can simply send it in to the
>first input of Streamline. Without any other specifications, the
>streamline will start at the centerpoint of the vector field and proceed
>from there. Pass its result into Image. For more info, you could first
>pass the result of streamline through AutoColor and it'll get the
default
>color map indicating the magnitude of the velocity at each point along the
>streamline. Collecting the output of streamline with maybe ShowBox of
the
>vector field before Image might be better. Pass the Streamline through
>Tube or, if the curl of the vector field is interesting, Ribbon would
also
>be good. You can specify a non-default starting point several ways, such
>as a vector interactor or a Probe. Past this, the options become endless.
>
>Greg
>
>
>
>
>
> "Frederick R. Phelan Jr."
>
> <[EMAIL PROTECTED]> To:
>[email protected]
> Sent by: cc:
>
> [EMAIL PROTECTED] Subject:
>[opendx-users] Streamline Module
> son.ibm.com
>
>
>
>
>
> 07/03/2002 09:31 AM
>
> Please respond to
>
> opendx-users
>
>
>
>
>
>
>
>
>
>Simple question . . . if I have some 3-D velocity data, what kind of
>manipulation do I need to do to send it through the streamline module and
>view the streamlines? Assume simple grid.
>
>Thanks.
>Frederick R. Phelan Jr., Ph.D.
>Multiphase Materials Group
>Polymers Division
>National Institute of Standards and Technology
>NIST, Bldg. 224/Rm. B108
>100 Bureau Dr., STOP 8543
>Gaithersburg, MD 20899-8543
>301.975.6761 (VOX)
>301.975.4932 (FAX)
>[EMAIL PROTECTED]
>http://www.nist.gov/frederick_phelan
>
>
>
>
Date
Frederick R. Phelan Jr., Ph.D.
Multiphase Materials Group
Polymers Division
National Institute of Standards and Technology
NIST, Bldg. 224/Rm. B108
100 Bureau Dr., STOP 8543
Gaithersburg, MD 20899-8543
301.975.6761 (VOX)
301.975.4932 (FAX)
[EMAIL PROTECTED]
http://www.nist.gov/frederick_phelan
