curious question to the community - reasons for parametric or direkt modeling

[Anonymous]

1) why do you use Fusion 360 in parametric modeling mode?

or

2) why do you use Fusion 360 in direct modeling mode?

Thank you.

(oops, sorry, typo in post title)

[PhilProcarioJr]

@Anonymous

I only use parametric modeling mode for parts that have generic use, like brackets, aluminum extrusions, C-channel, I-beams so on. The main reason I only use it for these parts is because I do a lot of large assembly work and Fusion just can't handle large assemblies with a timeline.

Direct modeling allows for larger assemblies (although not very large). I do a lot with T-Splines and will never work with T-Splines in a timeline. That is a disaster waiting to happen and also a lot of missing functionality that is essential to really take advantage of T-Splines. I also find DM much less restrictive and a lot more capable then Fusions parametric modeling. One of the biggest reasons is some errors in the timeline can not currently be fixed.

My train project would have been completely impossible with the timeline on and even in DM it came to a hault.

Train Project

[cekuhnen]

@Anonymous

It really comes down to how you want to work and maybe what and the scale of the project you work on.

First my background is in surface modeling as an industrial designer.

I mainly exclusively use parametric because of the ability to adjust design quickly and because I have quite some years of modeling experience it is quite easy for me to predict how I have to structure the timeline to prevent hick-ups later.

Since I never execute blue prints the parametric route is always the best start.

However with model detail it can get cumbersome to manage.

Sometimes particularly for detailing I then bake the timeline and switch to DM to do adjustments or bake the model sofar and continue with the detailing as parametric.

DM is nice and fast but the problem is that for exploring or adjusting designs it might have some modeling downsides resulting into more manual labor than parametric but is much much easier to use than parametric.

DM is also not DM programs implement it differently based on their focus.

[TrippyLighting]

I think first we need to clarify that parametric and direct modeling is not necessarily mutually exclusive. This has to be evaluated for each CAD software. In Spaceclaim, for example direct modeling can also be parametric and if one takes a closer look, Many of the direct modeling features in Fusion 360 seem to be direct copies off the functionality in Spaceclaim sans the parametric nature.

When mesh modeling that can also be mixed. In Blender for example the actual mesh is direct modeled, but the modifier stack is parametric. Very powerful!

The strong separation between direct modeling and timeline modeling only applies to Fusion 360.

For designs with relatively few features and a relatively small number of components I prefer timeline modeling.

I would use direct modeling more but the problem is that all too often when I try to do simple things - such as lengthening a slot -  they don't work as one would imagine. 

Also, it is pretty clear from comments by AD folk here on the forum that direct modeling has not seen any development and bugs that pertain to DM rarely get fixed.

I don't really want to beat that dead horse, but I most certainly won't bet on it!

[ToddHarris7556]

Our workflow (cabinet work, mechanical interactive exhibits & whatnot) does tend to favor parametric modeling, simply for speed of revision. Using a skeletal modeling approach and careful modeling, just tweaking a couple of control parameters can very quickly change an exhibit module the way we need. Caveat - most of our assemblies have maybe low DOZENS of parts. As assembly size increases, the need for consistent modeling practices increases exponentially, and there are simply practical limits, as @PhilProcarioJr suggests. We might have 5 different designers jump in an work on any given model - if someone went in and got sloppy, then you most definitely do not want to be the person to follow behind and try to straighten it out. We're still working on best practices for managing our more complex assemblies that get up into the 100+ component range. Anything that starts looking like a complex machine (5k+ parts) we use Inventor.  

We find DM to be super-useful for things that are organic and/or standalone or clean interface with other components. We're still learning to appreciate and exploit the power of it fully. Ideally, as @TrippyLighting I'm sure there are hybrid approaches that make use of both.

[ToddHarris7556]

I just watched THIS video that talks about using direct modeling for repair of imported geometry.

This isn't something we do a lot, but I can see it sure would be useful.

[TrippyLighting]

Ahhhh... that is a interesting aspect of Fusion 360 I forgot to mention.

I almost NEVER turn the timeline on in imported models. The ability to de-feateiurre and repair imported geometry without cluttering up a timeline with unneeded operations is nice.

Unless changes have to be made to that imported geometry that have to be parametric there is no compelling reason to turn on the timeline.

DM Models use only about a third of the data a timeline based model uses.

[Anonymous]

@ToddHarris7556

Just for information: the video you mentioned is part of a very interesting one hour AU-presentation:

Brad Tallis CP124298: Direct Modeling in Fusion 360

Brad Tallis also shows the top-down approach to create a new part starting at 39:00. I was impressed. This was one of the reasons for me to start this thread: how well does it work in real life?

[cekuhnen]

@Anonymous @ToddHarris7556

I think the DM vs parametric you also overcomplicate.

the mainly and on,y difference between DM and parametric is that parametric is more interactive while DM is just manual. 

The modeling tools you use are the same. Essentially with parametric you can automate tasks like adjusting a surface by adjusting a sketch that creates the surface. In DM this is all undo and redo or replace face if possible.

the biggest difference is also speed. DM models always perform faster.

Most clients I know use Rhino for a reason which is more a DM like surface modeler. Yes manual work is more present but there is no overhead of parametric. One furniture company particularly does no even need parametric since everything is easy to manage via DM.

i think the push for must be parametric comes more from interesting opinions that it is superior and thus a must have for manufacturing.

i always would suggest based on the complexity and need to decide which route you use and how you mix em.

as I mentioned my designs are simple (body count) enough that I only use parametric.

in my architecture projects this however is not feasible so we use DM.

Fusion360 is by the way a terrific SketchUp killer for architecture modeling. 😉

[ToddHarris7556]

@cekuhnen I may not completely follow your argument, but I didn't mean to make it sounds complicated. I don't think it is. I don't disagree with many of your points. I respectfully suggest:

Direct modeling is faster and lighter weight for a specific reason: it simply allows the user to 'Directly' edit the geometry. 

Parametric modeling, by contrast, provides a set of 'global variables' that are then mapped to the geometry. (At the API level this is offered as ParameterList object, or more correctly, UserParameters and ModelParameters) It's the extra overhead of maintaining these variable mappings that results in the performance hit. While Parametric modeling does offer great power, it also (unsurprisingly, really) means that it's possible to do things that break those relationships. i.e. timeline errors. 

---

The modeling tools you use are the same. Essentially with parametric you can automate tasks like adjusting a surface by adjusting a sketch that creates the surface. In DM this is all undo and redo or replace face if possible. 

Well, yes, and no. When in parametric mode, parameters are being created on the fly, which are then mapped to geometry. This is transparent to the user, but every entry into a dialog box is then used to create entries in the Parameter table, which are linked to features. In DM mode, the entries are just taken straight from the dialog and used by the modeling engine to create geometry - no stopping along the way to create, or manage, parameter entries.  

I'd submit that one simple example of parametric power would be a parameter like 'BoltSize' that controls dozens (or hundreds) of bolt holes through a model. One simple parameter edit, and it's done. 

I didn't mean to suggest parametric was 'better' - it is one of two very powerful approaches available, each with their own strengths.