Assembly/ Part regeneration statistics

[rmcboaty]

Does Creo (we will very soon upgrade to 4.0) provide regeneration statistics for an assembly or part regenerate like Solidworks does?

the reason i am asking this is:

we have a rather complex-ish assm with regards to regeneration (lots of general patterns, equations). nothing serious, but it still takes 20-30s (or more) for a relatively simple model to regen.

i would like to access regen stats, so that i can optimize certain features, if i can (standard hole instead of sketched, which we had to use in WF, because it has no drill hole profile, etc.) to cut down the regen time.

(although the entire asm regens a lot longer then parts or subasm's, probably because it has a lot of equations)

thanks.

This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.

[TomD.inPDX · Accepted solution]

If I could just my SW guys to do work instead of tweaking a nanosecond from their hole feature...

Seriously, do we even want this?

I know when my regenerations are on steroids because the counter just looks at me feeding me lies.

Typically I also know why; merge operations in Creo are top dog.

They have to rewrite the surface geometry on one new merge feature at a time.

It is tedious!

But, for the most part, you will know where you are sucking up CPU cycles.

I've put Creo and SW back to back.

Not much difference in how you treat features in regeneration time.

Pretty much all the same techniques are available in Creo, and more.

[rmcboaty]

the funny thing is that if i save as copy and use the "delete user pro/program", the regen time drops to a second or so.

the regen is big only when i read the text file containing parameters though. subsequent one is usually down to a couple (or 10) seconds.

i still would like to know if i gain anything by going to standard holes for example (since they are patterned).

[dschenken]

I doubt that there is any difference in time between hole features and creating by protrusion. The same number of surfaces need to be evaluated, so there isn't a lot there to be tossed. This is different, for example, than identical vs general patterns. In that case, even though the geometry -can- be the same and would seem to take similar times, the Identical option skips self-intersection checks and others; it is the additional checks are what chews up regen time.

It sounds more like there is a network lag for opening the text file, plus the need for Creo to load the portion of the software to evaluate the Program.

A software execution trace ability is one that I've pushed for for a long time. The PTC position is they will produce software that is so good there isn't any need to profile it's execution, but that's never going to be the case. Creating a CAD model is interactive software development; it doesn't matter if it puts a picture on the screen at the end, it should have the same tools as other professional software development. Look at 'The Old New Thing' and you can see that experts in Microsoft often examine the assembler and machine language generated by their own tools to uncover problems. If those guys need it, so do we.

[rmcboaty]

all files are on local drive. feature patterns are mostly cutouts (holes, gear teeth...) which can't be anything but general, as they intersect the existing model. when i regened it, it said around 1000 features and it's a relatively fast model (20s).

geometry is basic, no unnecessary detail.

these asms are basically parametric driven product families that get used (resized...) constantly.

there are other areas to optimize also, though, outside of creo.

if the regen time is defined by # of surfaces that need to be processed, then it's probably a shot in the dark, yes.

[dschenken]

It's of greater interest when working with assemblies that are rolling in to the 10-20,000 feature range. Even 0.1 second per feature is pushing a half-hour of regeneration. Similar with getting memory requirements. For the individual part modeler it's not often a big deal, but working towards the upper levels, it is annoying to find some short-cuts that take extra regen time again and again and again but shaved a minute from the originator's day.

The problem is the 'simple' stuff is given to the least experienced users and they don't see the cumulative effects of fouling things up.

For example - instead of projecting a datum curve to represent the cross-recess on a screw head, they''ll make multiple cuts for a feature that never has a dependency. Or they go to a supplier website and download a model with 40 Mb of detailed logo that's been embossed into a block. Or, my favorite, they got a model of an alternator that included the windings when all they needed was a volumetric place holder. I wonder why the views take 10 minutes each to recreate?

[TomD.inPDX]

Lightweight modeling sure makes the super-models a different story.  They pretty much hold up their graphic representation end where the regeneration issues should be limited in everyday work.  However, I've moved around some pretty good size products where the need the regeneration is quite limited.  I also have a healthy attitude to very nested systems.  Drilling down to the affected assembly keeps your workload manageable in memory.

As for large scale wholesale regenerations... yea, we've done that.  In days of old, drawings were murder on large assemblies.  Not anything a simple feature could solve.  Haven't had to deal with those for some time.

Robert, you do have Geometry patterns that behave the same (okay, only similar) to SW.  Creo will let you do more with those geometry pattern.