Acis Over Parasolid?

[tlehnhaeuser]

Has anyone else noticed lately that ACIS seems to be able to perform more advanced modeling than PARASOLID.

 

I've been "taught" to use Parasolid vs Acis prinarily since its better at blending which I like. But, lately I find myself having to switch to ACIS to get blends to model.

 

Can anyone shed light on this? Did Spatial bump it a notch to take over of the kernel market?

 

Thanks

Tom

 

[Cary OConnor]

Well...I'm not sure why you need to switch to get a blend to work....The Kernel Collaboration should do that for you automatically. Can you send me an example where this is not the case?

Thanks,

Cary

[Alex Khenkin]

Cary, switching kernels to make blends work is something we all do when having trouble with blends - in my 4+ IC years I've never had IC doing it on its own.

-Alex.

[cborer]

I do also not remember of a change Kernel automatically.

But to switch Kernel helps sometimes.

Still there seems to be no rule which is Kernel better.

[mmccall]

YES... I switch kernels quite often.. mainly with blends..

 

Just don't do anything silly... like take the switching option away !!!

[tlehnhaeuser]

Cary,

I use kernel colloboration OFF most of the time since I want to be sure when I provide data to no-ironcad users it flawless translation.

 

Carlo, as for rules, the general rule "was" that Acis was very good at modeling tangencies (picture two cylinders kissing each other) and that parasolid was better at complex blending operations.

 

But like I mentioned, this "rule" seems to have been shattered.

 

Tom

[Cary OConnor]

Ok.....I think there was a post about this in the past, but let me explain Kernel Collaboration again....

 

Kernel collaboration happens at the boundary representation level of a shape (basically the solid shape). Kernel collaboration is an option that needs to be enable (it is enabled by default) under the Tools/Options/General. Kernel Collaboration is only accessed in certain commands like Blends, Chamfers, Boolean, and Shelling.

 

How it works- Example: If you attempt to blend an edge and the blend fails in the current modeling kernel, the boundary representation is sent to the other kernel to attempt the blend. If it succeeds, it will send the final boundary representation back to the kernel you were in. So to the user, you never can really tell if it happens or not unless you turn kernel collaboration off and regenerate the model to see if fails. The only noticeable thing is that the model takes longer to regenerate. I have seen the Kernel collaboration work in many cases from customers when they didn't even know it was happening.

 

Even if a model uses kernel collaboration, when you export to a solid format, the model will be translated with no issues due to the kernel collaboration since the export only cares about the final boundary representation model.

 

So this is why I would ask to see some examples were it doesn't work. There are differences in the kernel, but we can't really state which is better. Some operations are faster in one than the other and some items are not supported in the one or the other.

 

Cary

 

[dhovatter]

I never knew about switching to get blends to work. I need to try that.

 

I use ACIS mainly because it make my files smaller when saved.

 

[tlehnhaeuser]

Thanks Cary for the clarification.

I can't remember if exporting was actually a problem, I always naturally assumed it would be a problem for CAD applications running one kernel would not be able to read data modeled with two.

 

I'm glad to hear thats not the case, I'm running as we speak to put my KC back ON!

 

Thanks!!!!!

 

Tom