Hi Jeremy



I am somewhat disappointed with Tim on this (and some other) problem(s). When shark LT was released, I was expecting this G1/G2 match surface to be fixed, and Tim did state he was currently working on the fix , but as that is almost 2 years ago and we have seen the release of Shark V7 with the same buggy tool(s) I doubt we will see a fix anytime soon.


- Steve

Hi Steve,

This lack of a fix is a strange one for sure. It seems that with G1/G2 'it does not do what it says on the tin'.
After 2 years it can't be time that's the issue here, so i guess it must be a resources issue at Punch!. If I had paid out $1295 for Shark, or more for Shark FX I would be considering asking for my money back, but would prefer to be shown that it was user error and eat some humble pie instead.
As it is, I have paid $99 for ViaCAD and am very happy with what I have got for my money.
Anyway I didn't join the forum just to gripe...and I hope to become a happy Shark customer one day soon and appreciate your posts.
I hope that all you Shark guys the attention from Punch that you need on this.
For now, I'll watch from the sidelines with interest on this one, as I have still got some re-learning to do on ViaCAD having spent the last 15 yrs on Unigraphics.

I don't have my laptop out yet (gotta wait til lunch), but, has anyone tried to see whether the Match Surface tool will match to an original mesh any surface that is derived from (object converted) from said mesh?

I ask because when i convert meshes to surfaces, i think they are not 100% matching up. If i take original curves used in one app that generated the basis of the mesh, and draw surfaces (ruled or otherwise) then color them and turn on both layers, i can see laryes from one or both dipping in and out of the other, betraying that the hand-generated surfaces is not matching. If the Match Surface tool can actually TOUCH the mesh, and use it as the basis for matching the converted surface copy, that would be awesome as an interim to VC/VCP/Shark someday automatically creating one.

As Nabed said - still waiting for a response / update / comment - ????

No, it probably fell on Tim's head, while he was plugging it in! ;):D

C'mon Tim, between playing "Wheel of Fortune Super Deluxe" and printing new business cards You could answer to this thread to give some insight to a serious problem that has been unresolved for ages.

Bump!

still waiting for a reply....

Maybe Encore have emptied out the Punch 'too hard basket'.

Hi All
My apologies for not having addressed this issue. Our G1/G2 edge matching is based on some underlying ACIS technology that constrains adjacent edges. We've made some changes in the next v7 service pack beta that improves this. Please continue to private email me examples that fail such that we can add them to our test cases.

QT movie below made with Shark FX v7 build 943.

http://www.csi-concepts.com/Demo/videos/MatchSurface.mov


Tim

I've been having a look at these older threads and this one pops out like a sore thumb. Lets be honest here. Surface matching in Shark (as it was in Cobalt) is plain bad. If you want a single reason why Shark is not my production modeller this is it.

I read about people here using SolidWorks and moaning about surfacing, well folks, let me tell you the surfacing in SolidWorks is eons ahead of Shark. There are just too many holes on the modelling armour in Shark when it comes to surfacing at a G2 level. And I find this frustrating, as this situation has existed since the late 90s - it has not improved.

Is it an ACIS issue? Maybe. ACIS has some limitations (or rather, limitations as they appear in Shark and other ACIS based systems). Try creating a surface patching a group of adjacent surfaces at G2 level. Try thickening more that one surface. Lots of annoying stuff.

I use VX and SolidWorks in addition to Shark and Cobalt. VX and SolidWorks both handle surface adjacency G2 matching no problem. Both can shell multiple surfaces. Both can generate new surfaces driven by the G2 conditions along multiple edges.

Yet, both Shark and Cobalt make creating curve networks a doddle. They make editing surfaces easy. The interface is considerably more direct and fluid.

Frankly, modelling in Shark and other ACIS based systems seems to have hit a standstill. Models are not that robust. A lot of new features just don't work well in the real world outside of demo ware. This is not a rant BTW. I bought Shark primarily for its CATIA v5 handling - and it does that very well and it was money well spent. But I would like to use it more for production level modelling. At the moment there are too many things that get in the way to do this (not to mention drafting!).

When you have to model a medical speculum using a split line a tool maker has supplied, Shark just doesn't cut it. sorry.

The original Concepts Unlimited was sold as a precision concept modeller. To my mind that means a focus on G2 modelling. We are not there yet.

Guys, now that Spatial have opened up the ACIS and CATIA kernels for licensing maybe you can look at adding some CATIA technology? For those that still doubt me, and think something like SolidWorks is inferior, I'll say two things.

Boundary surface

Fill Surface

Google them. The fill surface is like the cover surface with tangency and surface rebuild in Shark....but it works. The boundary surface has no equivalent in Shark.