Can you give some explanation on what's the differences between the four options under Curvature Check?
Or where can I find some information on them? Thanks!
↧
Surface Continuity Analysis > Curvature Check
↧
Studio Surface - four input curves
To understand how Studio Surface works seems not to be an easy job.
I thought that (1) perfectly matched input curves and (2) four or less input curves would definitely yield a surface of (A) high quality and (B) low continuity deviation. Especially "four or less input curves" should lead to a "relaxed" surface. It seems that I was wrong.
If a surface is modelled across the center plane, a perfect surface can be created. This means that there is a quite good solution based on the original four input curves in math, but Studio Surface simply does not give it out.
Hope to have your comments. Thanks!
ADD: I'm sorry for giving an example of the bad practice of "creating a half and mirroring". And it's not intended to pick up again the topic taught intensively by experts in a previous thread. Thanks!
4 perfectly matched input curves
(TOP) loose tolerance > high deviation ; (BOTTOM) tight tolerance > heavy pole structures
↧
↧
Re: Update for no reason!?
I think using part modules could be a great way to limit the scope of update sequences.
In your experience, can Part Modules also be a viable solution when big NX models you run into "feature number overflow" ?
We experience whenever one model has ~1000+ features it can produce hairy and unexpected update behavior, even when all features are properly groomed with regular part cleanup and well aranged feature group composition.
Can part modules relieve such problems?
Greetings
↧
Re: Update for no reason!?
I think it would definitely help in cases where it is possible for you to isolate each individual tool into its own part module. Then the 1000's of features become more manageable.
I think there are three main purposes to use part modules:
- Organization/management of large numbers of features
- Isolate/delay model updates
- Enable true team design within very complex models by isolating each functional area into a separate 'Linked Part Module' and taking advantage of body modifying part modules. This way several users can make concurrent changes to different portions of the model at the same time.
↧
Re: Surface Continuity Analysis > Curvature Check
there is a short description in NX docs at "Face Curvature Analysis".
Edit: I have found a beautiful, thorough explanation I got from GTAC years ago when I asked the same question, it's attached
↧
↧
Re: Studio Surface - four input curves
Sorry, but I don't really know how Studio Surface works.
But here is a guess ...
It is a very general brute-force algorithm. So it applies the same construction technique regardless of what sorts of input you provide. In your example, there is a simple solution that's fairly obvious (to a human being), but the general algorithm is not able to find this solution.
A simple analogy might be helpful. Suppose we want to write an algorithm for finding the square root of a given number. In general, this requires an iterative algorithm. But if the input is "9", then the output should obviously be "3". However, unless there is some special-case logic built into the code, it will just run the general iterative algorithm, and it might well produce an answer like 2.99999997.
↧
Re: Surface Continuity Analysis > Curvature Check
Hi
↧
Re: Studio Surface - four input curves
Hi
↧
Re: Spur Gear Creation
no joy, it still hangs here
Block Styler: Found the file at F:\users\cadcenter\nx11\library\tools\application\SpurgearDialog.dlx UNDO_UG: Recyclable status of mark 24 set to False Successfully loaded dynamic module F:\Program Files\Siemens\NX 11.0\NXBIN\libstylercomp.dll Loaded module f:\program files\siemens\nx 11.0\nxbin\libstylerint.dll 7fed09c0000 15d000 d52805c2-4552948f-938105b0-f2a1ae4e-1=libstylerint___151220618164 version = 11.0.2.7 Loaded module f:\program files\siemens\nx 11.0\nxbin\libstylercomp.dll 7fed0b20000 a8000 2e15fd56-41ac53e5-7198a397-3bd2a39d-1=libstylercomp___151314710464 version = 11.0.2.7 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 POSIX_map_dos_error: Uncommon windows error code: 123 Resolve failed: NXOpen.Utilities, Version=4.0.0.0, Culture=neutral, PublicKeyToken=null Loaded: : NXOpen.Utilities, Version=4.0.0.0, Culture=neutral, PublicKeyToken=null
↧
↧
Re: Studio Surface - four input curves
wrote: To understand how Studio Surface works seems not to be an easy job.I thought that (1) perfectly matched input curves and (2) four or less input curves would definitely yield a surface of (A) high quality and (B) low continuity deviation. Especially "four or less input curves" should lead to a "relaxed" surface. It seems that I was wrong.If a surface is modelled across the center plane, a perfect surface can be created. This means that there is a quite good solution based on the original four input curves in math, but Studio Surface simply does not give it out.
actually, your surfaces seem not to have anything wrong.
They matches your tolerances, in both cases.
I suppose you wanted a lighter math than this, but in such a case this is perhaps just impossible.
If you have a go using Thru Curve Mesh you can notice the difference, Studio Surface is better.
↧
Re: snip a slab surface into bezier patches
wrote: Hi
Thank you for your reply!
My goal is to get a light edge (as highlighted in the image below). The part file in the original post is only a simplified example, in which the trimmed edge seems not to be too bad. But in most practical situations, trimmed edges are too heavy to be usable for further modelling. I just want to know whether there are better general workflows dealing with those situations.
Tim is right, please post an example in which a trimmed edge "too heavy to be usable for further modelling" causes you troubles. As this happens to you "in most practical situations" you should have several such examples.
↧
Re: Drafting Parts List - reference external spread sheet
Thank you cowski1 for the help. I think this is the only way to get it to work. As you stated, and I am finding to be very much true, pulling information from Excel is quite tricky. I think haveing it set as a part attribute will be the only way to get it to work.
Thanks again. I will post any working code here late if I get it to work.
↧
Re: Fit Curve vs. Rebuild
wrote: Many surfacing commands have the Rebuild option, which reparameterize the input curves.
well, while this matches what is stated in NX docs, it seems to me that Rebuild doesn't operate on the input curves but rather it refits the output surface, imposing the math you specify.
Changing the math of input curves (by Fit Curve) doesn't necessarily determine the output math. NX could refit the input curves on its own, so you could end up with a math different than what you wanted based on input curves.
↧
↧
Re: snip a slab surface into bezier patches
Hi
If you want to create an AFB you don't need any edge.
Yes, AFB don't need any edge. But, because AFB gives a heavy blend surface now, it seems that the only choice is to use Studio Surface to build the blend manually and then a light edge is necessary.
↧
Re: Drafting Parts List - reference external spread sheet
Below are the expressions that I created in the part when I tested it:
In my case, part numbers were in column 1 of the excel sheet and descriptions were in column 2.
From here, you would create a part attribute and link it to the "test_description" attribute. This will be inherited as a component attribute when used in an assembly and the attribute can be referenced directly in notes or parts lists.
↧
Studio Surface vs. TCM>Simple
I just find that Thru Curve Mesh > Simple gives the same result as Studio Surface.
Are they almost equivalent in creating a surface? Thanks!
↧
Check Mate - check if reference set contains one solid body
Hello everybody
I`m trying to make some test using Check Mate, which will check, if the specified reference set contains only one solid body.
I think the most job of my requirement I can do using one of the template checks, called "Check Reference Set/Object Mappings". Using this check I can investigate if specified reference set has specified type of geometry. But, how can I check, if in this reference set is only one solid body, not more?
Thank you
↧
↧
Re: Fit Curve vs. Rebuild
wrote: Hi
Thank you for your comments!
it seems to me that Rebuild doesn't operate on the input curves but rather it refits the output surface, imposing the math you specify.
This is a surpise to me. I thought that Rebuild reparameterizes the input curves first and then create the surface with the rebuilt curves.
In a previous thread, I upload an example. In that example, a Studio Surface used the Rebuild option, but gave a resulting surface not following the continuity specification. It seems that Rebuild did not try to REFIT the output surface.
it did. You imposed 5 degree 1 patch along the V, and your resulting math was exactly like that.
As for the (very light) math you wanted being inconsistent with your continuity constraints, well, that's another story, you can't have your cake and eat it too :-)
↧
Re: snip a slab surface into bezier patches
It's only necessary if you're going to do some sort of offsetting using that edge. There are other methods that can arrive at a similar solution. You may have to offset that edge and further refine the resulting curve to get what you desire or close to that original sharp edge shape - the blend tangencies do not necessarily have to match the edge shape 100% - and they probably won't since you'll have to refine the derived curve.
You could also try using the rebuilding and edge matching commands to try and finesse the blend afterwards. No, it won't be a simple workflow but that's the trade off when you're going to scrutinize each and every aspect of a surface.
↧
Re: snip a slab surface into bezier patches
If you want to create an AFB you don't need any edge.
Yes, AFB don't need any edge. But, because AFB gives a heavy blend surface now, it seems that the only choice is to use Studio Surface to build the blend manually and then a light edge is necessary.
not necessarily. You can use Fillet Bridge, as in the attached example, which has a pretty decent math. You can turn it into a real fillet by specifying a spine, the math get a little more complex then, but not that much, so I don't think the edge math have such a huge impact as you fear.
Once again, please post a complete example in which the edge math has the catastrophic impact you fear.
↧