What change does the sewing operation do to a B surface?

[WilsonBlr]

What change does the sewing operation do to a B surface?
I am using UF_MODL_create_sew to sew two surfaces. But when I query the target surface for knots and control points using UF_MODL_ask_bsurf before and after the sewing operation I see no difference. So I am not sure if I have the right results.

Does sewing alters the geometry of the surface itself Or creates a new surface with the altered geometry apart from the initial surface ? Or it is the only the representation that is changed ?
Is sewing saved in the part as a separate feature or something ?

I am trying to understand the format in which a part is saved in NX file and how are all the features saved? Please point me in the right direction.

Thanks in advance for the help.
Regards
Wilson

 

[JohnRBaker]

Technically NO. The previous sheet bodies are NOT altered in any way. Granted, it may LOOK like the edge of one of the adjoining sheet bodies has been moved to align with the other sheet, but that's just a sort of illusion. In reality , as long as the two edges were within the modeling tolerance distance, NX simply assumes that what are now two surface patches share the same edge, which will be one of the original edges, but it does this without actually changing the other sheet body. This is accomplished using something called 'Tolerant Modeling' where NX, for all intents and purposes, assumes that the adjoining edges of sew sheet bodies are represented as a single edge even if they were not perfectly matched, but did no deviated more than the modeling tolerance distance.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

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

 

[WilsonBlr]

Thanks John,

How then is it possible to get a surface which has a C0 or C1 continuity with the adjoining surfaces ??

 

[WilsonBlr]

 

[WilsonBlr]

I am trying to create a set of 'through curves' surfaces using UF_MODL_create_sew. I have two parallel surfaces and a third adjoining surface which should have a C0 or C1 continuity with the parallel surfaces.

I am creating all the three faces using a thru_curves method which accepts set of curves as input and returns a face tag of the surface created. Initially I create the opposite parallel surfaces (SURF 1 and SURF 2). Then I am creating the curved surface (SURF 3). For the 3 rd surface I pass the first two surfaces as object identifiers of neighboring surface for continuity. but I do not get a C0 continuity.

// Snippet : Function to create thru_curves surface from set of curves. Returns a face tag. 
/// Creating surface from set of curves 
UF_CALL( UF_MODL_create_thru_curves ( &tSection, &tSpine, &nPatch, &nAlignment, dValue, &nVdegree,
		&nVstatus, &nBodyType, UF_NULLSIGN, dTol, tCFaces, nCFlag, &tSurface ));

UF_CALL (UF_MODL_ask_body_feats (tSurface, &tFaces));	// Ask surface feature faces
UF_CALL (UF_MODL_ask_list_item (tFaces, 0, &tFeature));	// Ask through curves 

tag_t tFace =	NULL_TAG;		                // Feature tag
UF_CALL (UF_MODL_ask_feat_faces(tFeature, &tFaces));	// Ask surface feature faces
UF_CALL (UF_MODL_ask_list_item (tFaces, 0, &tFace));	// Ask through curves 
return tFace;

I try the sew operation

UF_CALL(UF_MODL_create_sew( option , num_targets , targets , num_tools , tools ,tolerance ,type_body ,&disjoint_list , &sew_id ));

But as I said I get the same control points and and knot values before and after sewing. What needs to be done to get a surface with an actual C0 continuity rather than just a representation of it?


Regards,
Wilson

 

[JohnRBaker]

That's NOT done by sewing. You'd have to use another routine which actually would change one or both sheet bodies. A simply sew body will NOT, by itself, create tangent or curvature continuos seams.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

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

 

[Xwheelguy]

The underlying curves will have to achieve the desired continuity within the allowed tolerances for both distance and angle tolerances (and curvature in some cases) in order for the surfaces to do the same. It's pretty much a garbage in (for the curves), garbage out (for the surfaces) scenario.

Tim Flater
NX Designer
NX 8.0.3.4
Win7 Pro x64 SP1
Intel Xeon 2.53 GHz 6GB RAM
NVIDIA Quadro 4000 2GB