May someone could offer a good explanation about "check manifold" 3

[cforall]

May someone could offer a good explanation about "check manifold" which appears in many dialogs within CATIA V5

TIA

 

[KevinDeSmet]

Just a guess. But I imagine it is checking whether you have non-manifold geometry. For example a cylindrical face tangent to another face imagine like a hole lying right on the edge of a part.

The tangent point is non-manifold, because there is zero thickness there. And that should be avoided. CATIA can handle it quite fine, but some other systems can't and it also can not exist in reality. There always should be, at least some, thickness there.

I just checked in the CATIA help and it says it is only available for curves... at least in the Join command. That's weird, can anybody explain?

Certified SolidWorks Professional

 

[Azrael]

I would maybe call it multi directional, not sure it makes sense, imagine that you have 3 curves that share one point, this would render that after a join operation that you don't have one resulting direction and therefor when trying to perform an offset will render problems.

 

[JohnRBaker]

Another classic example of a non-manifold condition is when you attempt to unite simple blocks like the ones shown here:

The issue is that you can't have more than 2 faces adjacent to a common edge.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

UG/NX Museum:

http://www.plmworld.org/p/cm/ld/fid=209

To an Engineer, the glass is twice as big as it needs to be.

 

[cforall]

Thanks guys for help me reach a basic understanding about this, the only thing left is to practice I believe.

BTW, this forum really gathered a lot of professionals here, that's great!

 

[jackk]

Good question and some great answers! But I'm still a little confused.

In Azrael's example; is it manifold or non-manifold when the 3 curves are joined at a common point (such as a "Y") ?

What about John's example; is the partbody manifold or non-manifold when more than 3 faces share a common edge?

 

[cforall]

Based on the descriptions above, more than three different curves sharing one end-point and more than three different surfaces sharing one edge are both non-manifold conditions I believe. and I also think it's the reason why CATIA call it "check manifold" rather than "check non-manifold".

 

[Azrael]

Jackk, in both mine and John's example after a join or unite it will render a non-manifold situation that initially is fine, it can be done and will work. The check for non-manifold is to make the user aware of the situation that can give you issues, if you offset the curve or the solid face it will give you an error but trimming/splitting is fine.... good to know I guess

 

[JohnRBaker]

... is it manifold or non-manifold when the 3 curves are joined at a common point (such as a "Y")?

What you're describing is NOT an 'edge' but rather a 'vertex' or corner. Generally speaking, a vertex can be shared by an unlimited number of faces, or to put it another way, any number of edges can be joined at a common vertex.

As for your second question, what determines a manifold condition is NOT the edge/vertex nor the face/vertex relationships but rather the edge/face relationship. So to answer thae question, ANY situation were you are attempting to create a model where an EDGE is shared by MORE THAN 2 FACES. BTW, while it's fairly common for someone attempting to create topology where four faces share a common edge, having a situation which involves only THREE faces, now that takes some real creativity ;-)

There is also another, although generally more obscure, non-manifold condition and that is where an attempt is made to join two faces into a single model where NONE of the faces share a common edge. To visualize an example of this think of a 'T' where one face 'joins' the other at a mid-point on that face. This again would be considered non-manifold if you tired to treat it as a single topological body.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

UG/NX Museum:

http://www.plmworld.org/p/cm/ld/fid=209

To an Engineer, the glass is twice as big as it needs to be.