Problem with surface loft 1

[spiatt]

I'm trying to render a 3D model of a rat spinal cord using Solidworks 2010, and am having problems doing surface lofts. I've used AutoTrace on each cross section, then fit a spline, and split each one into equal number of segments. For some reason, I have no problems lofting between some sections, but some of them I can't get to go, despite trying different segment configurations, guide curves, simplifying the spline, etc. I've attached my part. I've gotten the 'outer border' to work just fine, but am hung up on 'grey', only being able to loft between some of these sections. Please help!

 

[Eltron]

There was no part attached. It is a link to some sort of Word doc that I can't open.

Dan

http://www.eltronresearch.com

Dan's Blog

 

[spiatt]

Sorry -- google docs doesn't seem to like .sldprt files. This seems to work though.

Thanks for the quick response.

 

[CorBlimeyLimey]

36MB ???

No thank you.

 

[ctopher]

Maybe a mouse problem with the file?

Too big to download for my virus scan to handle.

Chris
SolidWorks 10 SP4.0
ctopher's home
SolidWorks Legion

 

[CorBlimeyLimey]

Either suppress all features, or rollback the Feature Manager, then ZIP and use the [/b]...or upload your file to ENGINEERING.com[/b] link below.

 

[TheTick]

Most likely, you are putting too much faith in the integrity of the curves resulting from fit spline of AutoTrace curves. A quickie review will likely reveal some wild curvature excursions. Also check number of defining points in troublesome splines.

 

[spiatt]

Ok, I took all the extraneous stuff out, leaving only the feature I'm having problems with (only 13MB now). Thanks a lot everyone.

 

[CorBlimeyLimey]

Still cannot download from Google Docs.

 

[ctopher]

I tried the model. I also can not use a surface loft between the sketches.
Will play with it some more later...

Chris
SolidWorks 10 SP4.0
ctopher's home
SolidWorks Legion

 

[rstupplebeen]

I created a boundary surface that went through all of the profiles. I was careful to have the connectors on similar landmarks in each profile. I suggest creating a couple of 3d splines to help guide the interpolation between profiles.

Please see the attached image and the google document link. The file is rolled back and zipped to save space. I hope this helps.

http://docs.google.com/leaf?id=0ByB...lhMTgtNTY2OGI5ZWRmNzEy&hl=en&authkey=CJDE3MgP

Rob Stupplebeen

 

[spiatt]

Thanks a lot, the boundary surface method seems to be more promising than lofting. However, trying to recreate it on my own, I keep getting an error that it produces self-intersecting geometry. Is there a good way to fix this? I really appreciate your taking the time to help.

 

[rstupplebeen]

I would suggest creating some 3D splines to help guide the surface. Basically you have sections that have sharp radii that are changing significantly between slices. In my model I chose a particularly abrupt transition that seemed to persist in most sections and used that as the control point on each. Please see the attached image showing the location of the control points. Could you download the SLDPRT? Creating this complex shape with only 1 feature may not be the most appropriate method. It may be helpful to break each slice into multiple parts and/or only use a couple slices at a time. In the end you just need to try a bunch of different methods and then post back your final solution if possible.

Out of curiosity what are you doing with this model?

I hope this helps.

Rob Stupplebeen

 

[spiatt]

Ok, that works out pretty well for me. Basically, I'm taking a series of 2D figures (like the one below), and turning them into a 3D model (my progress thus far also below). The research group I'm with is working on spinal cord stimulation, specifically running FEA of electrode arrays on solid models of spinal cords -- the hope is to be able to compare these to my model to be able to easily visualize which functional regions of the spinal cord are stimulated by certain arrangement of electrodes, and thus what reaction would be produced. Now that I've gotten past the bigger and more complicated layers, the rest of the details should hopefully go more smoothly. I really appreciate all your suggestions.

one of the 2D layers:

my progress thus far:

 

[rstupplebeen]

In that first image there is a cleavage at the bottom. I would make sure that you model that region with at least 2 different surfaces. Sounds like a cool simulation. Maybe I'll see you over at the FEA or Abaqus forums in a bit.

Rob Stupplebeen