Cancel
Showing results for
Did you mean:

# surfacing issues

Genius
Below is a partial model extracted from a big one.

Issue 1: Both curves are G2 with the faces. But Thru Curve Mesh can't achieve G2 with the yellow face.

Issue 2: If G1 is used to create Thru Curve Mesh, the highlighted part of the surface is problematic.

Your suggestions on fixing would be appreciated! Either the surfacing technique, or the neighbouring faces.

16 REPLIES 16

# Re: surfacing issues

Siemens Phenom

Body 2 needs to be simplified. Far too many patches.

Body 1 should have math changed to be same degree as Body2 in section direction

The side splines should be simplified and also moved to ends of surface.

Not knowing the full part I know easier said than done. I would also look at splitting body 2, the undercutting curvature on the opposite side to the problem area you highlight is making it difficult.

Created example with internal controlling curves.

In Preferences Modeling set Update to Dynamic Continuous. Then right mouse on the bridge curves and Edit Parameter. You will see the surface update as you edit the curve.

Steve V

# Re: surfacing issues

Genius

Hi @StevenVickers   Thanks for your very valuable suggestions! Let me make more efforts to improve it.

# Re: surfacing issues

Siemens Phenom

Hi Surfactant,

Without the design intent or the background features you will most likely get many answers. Difficult to say witch one would be the better one with the missing information. However with the information that is in there this is my conclusion without the surrounding information.

Issue 1: Checking how the curve and edge of the surfaces are aligned shows that there is a slight difference between the edge and curve. This is most likely the issue for you not being able to set your G2 constraint to the yellow face. The given information indicates that you have built surfaces not necessarily to exactly match the given curves and in addition you most likely have used an approximated freeform surface with no constraint to the curve – can’t say as I don’t know the input.

Issue 2: This is based on the complexity of the input geometry and whatever characteristics they are carrying. The “kink” you see is most likely due to the shape and parameterization/complexity of the yellow face. Any new adjacent face to the yellow face and with a G2 constraint to it will reflect whatever characteristic it is carrying quite far into the new adjecant surface.

Proposal:

With that said this is what I would do with the sparse input given:

• Trim and fix the curves so that they are starting/stopping with a C2 constraint into the given boundary faces (This is Key).
• Use Sweep to create the middle area
• Trim the Swept Surface back in both ends depending on how much you would like the system to optimize the transition.
• Use Studio Surfaces to fill the gaps with C2 continuity

Finally I want to double on what Steve states regarding the complexity of the input geometry.

I hope this might be of help to you

Best Regards

# Re: surfacing issues

Genius

Hi @Sandman   Thanks for your long reply and proprosal of solutions!

I follow your Sweep-Trim-Studiu Surface workflow. It works! The kink is gone when G1 is applied. But G2 is impossible to achive before fixing the curves.

Issue 1: Checking how the curve and edge of the surfaces are aligned shows that there is a slight difference between the edge and curve. This is most likely the issue for you not being able to set your G2 constraint to the yellow face. The given information indicates that you have built surfaces not necessarily to exactly match the given curves and in addition you most likely have used an approximated freeform surface with no constraint to the curve – can’t say as I don’t know the input.

• Trim and fix the curves so that they are starting/stopping with a C2 constraint into the given boundary faces (This is Key).

In my model the yellow face is just created from the curves with Thru Curve Mesh. So when you said "there is a slight difference between the edge and curve", I feel frustrated and can do nothing to fix it without revamping my whole model.

Issue 2: This is based on the complexity of the input geometry and whatever characteristics they are carrying. The “kink” you see is most likely due to the shape and parameterization/complexity of the yellow face. Any new adjacent face to the yellow face and with a G2 constraint to it will reflect whatever characteristic it is carrying quite far into the new adjecant surface.

Yes, the yellow face is very complex with dense jungle of poles. But it's the result of Thru Curve Mesh, and I have done my best to refine the starting curves. As above, I can nothing to refine it without revamping my whole model.

I must say that surfacing in NX is so difficult. As a good amateur coder, I thought that working with NX is as easy as working with a new programming language. It seems that I was very very wrong.

Anyway, thanks a lot for your help!

Highlighted

# Re: surfacing issues

Siemens Phenom

Hi Surfactant,

Looks like youre doing excellent!

I wish I could help you more, but with with the limitation in given input its not that easy. Would love to see what you are designing but its classified - right? I just try my best to make everything I can as understandable and as intuitive as possible by also giving you as much backround as I can. Things will clear i promise you. In the meantime should you find something that you think could be improved - do not hesitate to file an enhancement request with GTAC. Finally I would like question, and you may pass on it should you feel like not answer. How long have you been using NX? What did you use before?

Keep up the good work there Surfactant!

Best Regards

Fred

# Re: surfacing issues

Genius

Hi @Sandman   Thanks for your reply! I began to work with NX about 8 month ago. Before that I ddin't have experience with any design software. At first NX design seems to be only putting together some curves, surfaces and geometries. But when it comes to surfacing, the story is completely different. Thanks again!

# Re: surfacing issues

Genius
I think that this example reflects some deep issue in G2 failure in surfacing.

My original goal is G2 with the right yellow face (problematic) and G1 with the left dark grey face (no problem).

It seems that in the original example, the two splines and their relationship with the yellow face have nothing wrong. This is evidenced by the prt file attached with this post:

• A G2 continuous surface ("G0+G2_ok") can be created immediatey beside the yellow face SUCCESSFULLY.
Hope to have more comments. Thanks!

# Re: surfacing issues

Gears Phenom

Sorry, but I don't understand what you're implying, as it appears to me that you're comparing apples to oranges since they are 2 different areas defined by completely different curves. Just because random curves work in one section of a larger surface doesn't necessarily mean they are going to work in a completely different section. Each curve can add complexity to the resulting surface and can result in different interpolations. This is where understanding the modeling tolerances and how they are used come into play....if you increase your G0 tolerance, the issue goes away. On top of that, you created a G1 surface on both ends and extracted the isoparms from that and expect it to work in a G2 setting. Not really solid logic there, IMO. Go back to that first TCM surface (G1+G1), which is a larger slab and change the end near the yellow surface to G2. Same results - cannot achieve G0 continuity.

I will even throw in a different scenario for you - create a TCM that is G1 to the left grey face and G0 to the same curve where TCM G0+G2_fail ended above. Now create a 3rd surface between the 2 TCM that is G2 on both ends. Works just fine for me.

As others have implied before, the base surfaces aren't the best to start with - take it a step further and untrim the yellow surface and look at it - not really what I'd want to be seeing if I were working on this. Plus, you're not removing any of the discontinuities/bulges/wrinkles near the sides of the original splines so at that point, I have to question why even go for G2 if you're going to leave all the imperfections near the original splines....doesn't really make sense IMO.

Since this is only part of a larger model, it's quite difficult to solve these issues because I'm sure the surrounding areas will more than likely impact the look and shape of this area.

Tim
NX 11.0.2.7 MP11 Rev. A
GM TcE v11.2.3.1
GM GPDL v11-A.3.7

# Re: surfacing issues

Siemens Phenom

Hi Surfactant,

As stated earlier your issues lies with the yellow surface and the characteristics with the boundary curves - they need to be modified - or the yellow surface have to be modified to fulfill your requirements and run smoothly.

I can show you a bunch of solutions but without knowing the design intent or the specific required process I can’t really help - we would only move the issue. What you are trying here is not to recommend but might work in some areas as you can see, but you probably show the issue to another area. This goes as far as I know for all CAD software as C2 continuity is very specific about things and generally used by studio engineers in styling departments with years of experience. Cutting things down to smaller size/chunks might work locally but it does not really cure the core issue. You have been working with NX for 8 month and are already working with creating styling surfaces that most designers/styling people have years of experience to perform – regardless software.

Have you been through Class A/Styling Surfaces training? If not, who is your local Siemens contact? Maybe it is possible to arrange a class, workshop or similar?

Best Regards

Fred