Synchronous Modelling Guidance

[carlharr]

Hi,

While looking at synchronous modelling, I've found some great uses for it, such as:

- editing features in an un-parametric (imported) CAD model
- creating machining operations in a machined casting (i.e. linked in casting, then machine it)
- edit complex parts (several hundreds of features) where normal editing will not work
- editing parts where design intent does not allow you to make the changes you need

All of these are methods for editing existing parts, and record the changes made as features.

I've also found that synchronous commands can be used as normal parametric modelling features.
Typical examples would be replace face to change a shape, or rotate face to change its angle.

BUT

I've also seen a lot of models almost built entirely using layered synchronous commands, and these are awful.
They're hard to understand, difficult to edit and just plain messy.

My question - does anyone know of a freely available training guide to good and bad use of synchronous modelling that I could distribute, to save me the trouble of re-inventing the wheel?

Thanks in advance, Carl

http://www.jcb.com

NX 8.5 with TC 8.3

 

[cowski]

I've also seen a lot of models almost built entirely using layered synchronous commands, and these are awful.
They're hard to understand, difficult to edit and just plain messy.

Are you referring to a history based native NX model that uses synchronous commands almost exlusively, or a history-free native NX model?

www.nxjournaling.com

 

[KhimaniMohiki]

I have seen instances where external resource (contracting houses) have modified a model almost entirely using synchronous features, with every iteration of the design in 1 part file with synchronous features stacked on top of each other. The results being relatively small (30x30x30mm) plastic injected switch parts containing up to 3000 features.

I think if your fundamentally changing the fit, form and function of a part during its development then you should do it properly ie rollback and remodel or just start again.
If your doing minor tweaks to add draft and other 'hot fixes' to a part which you haven't been involved in designing but not changing its fundamental fit, form or function then its acceptable to have a number of synchronous features AT THE END of the history tree.
There are a few features in synchronous modelling which are just a quicker method of performing a string of standard features ie replace face effect could be achieved through a combination of offset face, untrim face and trim body. The same goes for delete face, move face, make offset. I think these features are perfectly acceptable within the modelling history when used in moderation.

Khimani Mohiki
Design Engineer - Aston Martin
NX8.5

 

[carlharr]

Hi cowski, the first one "history based, using synchronous commands almost exclusively".

Hi Khimani, you've hit the nail on the head with what I'm seeing. Synchronous definitely has it's place and is very useful, but not when used to create a model from scratch. Think I will end up putting something together, so focussing purely on the "bad use":
- too many features causing longer updates than necessary
- complete loss of design intent
- other engineers won't understand the design
- a model that's impossible to edit without adding more (synchronous) features
- failures due to consumed features / features depending on themselves - saw that one this week

With models like these, you'd be just as well to extract a body and delete all previous features when finished, for what use they are.

Thank you for the input!

http://www.jcb.com

NX 8.5 with TC 8.3

 

[Xwheelguy]

I wouldn't allow it in History-based mode unless for a "hotfix", as Khimani pointed out, or for revising imported geometry such as a STEP model. I could also see it being great for styling models for the start of a new project/product. I'd never recommend it for 100% full time use, start to finish of a production-use solid model (unless everyone was positive the model would NEVER change again).

Tim Flater
NX Designer
NX 8.0.3.4
Win7 Pro x64 SP1
Intel Xeon 2.53 GHz 6GB RAM
NVIDIA Quadro 4000 2GB

 

[KhimaniMohiki]

I think it depends heavily on what your designing as well, I found a lot more use for it doing quick turn-around injection moulding concepts than I did doing sheetmetal fabrications where there was no time saving advantage versus the standard tools.
Just as it wouldn't be good practise to put an extrude on an extrude on an extrude, stacking of synch features should be avoided, I think thats just about as far as guidelines can go without getting barraged with questions from engineers about their correct usage.

Khimani Mohiki
Design Engineer - Aston Martin
NX8.5

 

[cowski]

I don't think this problem is unique to the synchronous commands; before they came along, I've seen similarly bad models using the traditional commands: extrude/unite to fill in holes, offset face on top of an offset face, on top of another offset face, elaborate (and error prone) sketches used to create faces on the model that later get trimmed away completely...

Point being: extensive use of synchronous commands is not inherently bad practice, a bad modeler will create bad models no matter what tools he has. Limiting a powerful set of tools to one or two "approved" use cases may be counterproductive in the long run.

www.nxjournaling.com

 

[carlharr]

True, and I wouldn't want to limit their use because they are powerful and useful tools.

But I would like to provide some guidance on how not to use them. Agreed a bad modeller can produce bad models with any commands, I just have a feeling that layered synchronous makes it "extra easy" to produce something that no one else could use.

The feedback here has been very useful.

http://www.jcb.com

NX 8.5 with TC 8.3