Direct Modeling vs Timeline : why do you prefer DM ?

[cekuhnen]

I am a pretty timeline designer and I read here that many prefer direct modeling ideation and such.

What are your reasons to select DM instead of the timeline mode. Just curius.

[yoshimitsuspeed]

Every time I try DM I get into a situation where I wish I had the timeline.
As others have stated, maybe it's just how I have grown used to working. Every time I have tried DM for anything of any complexity or time investment I have regretted it so I really only use it if say someone sends me a step and I need to make a little change before machining or printing a drawing or something. Even then I have usually started switching to timeline.

[cekuhnen]

I think that DM fit me is good for certain models task like cube block extrusion mechanical shapes but less for complex shapes.

When ideal with fillets and trim surfaces I feel DM gives me often problems.

In that case I always prefer timeline.

[yoshimitsuspeed]

I definitely agree with that.
Even for simple stuff though I think one thing that would be a huge help is if there were 3D dimensioning options.
https://www.youtube.com/watch?v=YWvKib0XgDc

I would still generally use timeline but if capability like this was possile DM would be awesome for those times when you needed to make a little change to a step file or something like that.

[yoshimitsuspeed]

Plus that capability gives makes it easy to make 3D dimensioned drawings and that is kind of cool sometimes.

[Oceanconcepts]

@jeff_strater Jeff, I have tried the Create Base Feature workflow, and while it’s great for a simple object or two created in Fusion, what I need is a “bubble” inside which I can insert and remove many components as my design is worked out. I may need to try our many different headers or connectors to arrive at a good configuration. I may need to suggest that the Chief Electrical Engineer find a smaller inductor because the chosen part is going to be difficult to fit.  I haven’t found any way to make these changes inside a parametric design using Create Base Feature, mainly because, unless I’m missing something, you can’t insert an external component into a base feature. 

The board I most recently worked on had around 50 components and subcomponents that influenced case design enough that they needed to be in the model, or that I needed to try out in the process, and I need to be able to move those around freely. Most represent fixed parts that I don’t need to edit. Even tiny discreets I will often model because I want to know I’m allowing enough room for the layout designer to place the necessary components inside the board outline I specify. Capturing parametric data for those parts and moves doesn’t seem to me to have any value. 

Convert to DM option is one I was aware of, but in my designing, it’s mostly that I want to start with direct modeling and eventually switch to parametric. In a sense it's the workflow you are describing, just in reverse order. When I start with direct, and then switch to parametric, I am able to still freely edit and move around the components inside the base feature without it adding to the timeline, so in a way I have the “bubble” I want- I just can’t add any new components to it (without altering the timeline). So I don’t switch to parametric until I have things fairly well worked out. Or I think I do, anyway. 

I probably tend to stay in the rut that’s familiar to me, and some aspects of direct modeling, like the ability to freely move sketches around that Omar mentions, are things I’ve gotten used to for quickly working out ideas. But when I’ve tried to entirely do a design in a parametric workflow it’s the considerations above that have been the real roadblocks. I think that parametric workflows are great when you are working with data that could benefit down the road from storing that information- as when you have a shape you will want to tweak and have downstream features update. This more likely applies early in a design process if you are starting with a big picture, and drill down to the details. I need to start with the detailed components, and work out to the big picture, as we have some fairly harsh and specific requirements. 

After Phil’s class at AU on large model management I may start building boards and inserting them in to my designs as linked files, editing them externally- that may be another way to go. 

@cekuhnen Claas, I would be interested in talking about a real world example, and maybe defining a problem for students. I think the ECAD to MACAD interface is one area where tools are likely to evolve fairly quickly. I can’t share any proprietary information from our licensee, but if you want to message me outside the thread (let me know if you don’t still have my email) I’ll think about how to set up a design situation with parts and design parameters. I would love to see what ideas they might come up with. 

[jeff_strater]

Thanks, @Oceanconcepts, for this post.  I understand and share the desire for that "bubble" of direct modeling in an otherwise parametric design.  Yes, you are correct - Base Feature today is limited in that you cannot create components in this mode.  There was a good reason for this, I remember having the conversation, but of course not the reasons why...  I suspect it has to do with the way components are managed in the timeline in a parametric design.  It would raise all kinds of questions about whether to allow components created in the Base Feature to be restructured into a component created outside of the Base Feature, etc.

But, as you point out, using a separate design as this "bubble" works very well.  Just create a direct modeling design, do all your direct modeling stuff in there, including creating components, etc.  Then, just insert this into your top-level design.  The worst impediment for this workflow today is that there is no "edit in context" - no way to modify your referenced design "in place" in the assembly.  We do recognize this need, and are working on it.

The workflow of "start in direct, then switch over to parametric" is interesting.  Today there is no good way to do this, short of starting over.  We have definitely tossed around some ideas here, including Feature Recognition.   However, we've found that in a lot of cases, FR technology today is just not sufficient.  It works great for simple shapes, but tends to fall apart for more complex shapes.  But, for the simple shapes, it's usually not that difficult to manually re-design the model, I've found.  And, doing so allows you to design in a more complex paramertization, if you want.  So, as  yet, there is no good solution to this workflow.  

These are the kinds of forum discussions that I enjoy the most.  I like discussing real-world workflows and the challenges that they reveal that we may not have thought of.  That's where the Fusion team gets to benefit from real input.

Jeff

[Oceanconcepts]

 Jeff, I’m not really experiencing this as a deficiency, just trying to respond to the original question as to why I use direct modeling. This dual approach actually works well for me. If I build an assembly in direct, I place all my electronics in a master component. When I switch to parametric, that assembly remains editable without affecting the timeline, I just can’t add new components to it. 

Maybe I haven’t been completely clear, but this is placing all my electronics and trying out ideas while in a direct environment. When I have a path that I want to lock down and get more precise with, I have my electronics in a component container(s), and can build out the case design parametrically. I sometimes create separate versions of the file to follow different paths. This is a kind of “best of both worlds” approach. I don’t need to start over, I can flip the “capture design history” switch whenever I feel that a parametric workflow will be valuable going forward.  The thing that might improve on this would be branching when it arrives. For now, I feel working out general ideas in a direct environment allows me to be more intelligent about parametrization when I reach that point. 

Editing an externally referenced design in context would of course be a big plus as well. 

[O.Tan]

Good question, because Fusion doesn't allow me to mix both modelling techniques, I guess only when heavy pattern (both sketch or model features) is involved.

If I'm allowed to use both environment, usually the base model will be done in DM (Direct-Modelling) whereas HM (History-Modelling/Timeline) will do the fillets, holes and pattern.

Fillet and hole is done in HM is just to make the DM base model much cleaner as I don't have to worry about fillet breaking or such. Hole is more like a 50-50, I do prefer that in DM I can move hole whereever I want without having to open sketch or registering "move" into the timeline, but I guess if I'm doing things like a base plate (a plate which contains 20-100++ holes), then HM will make more sense.

You're not completely wrong, but Fusion only allows it to be done 1 way (either DM to HM or HM to DM), I can't use both at the same time (DM and HM). Before Fusion 360, I use SolidEdge and SE allows me to use both DM and HM at the same time, I thought of explaining it but upon Googling they're already examples you can see. Here are some videos showcasing what I mean (take note the vid shows ST3 and ST6, it's now ST8 so it's much more advance now)

[Anonymous]

I always started with Timeline, and when the model designing is finished I will save a Paraemetric file, then switch over to DM mode. At the DM mode I can delete lots of hidden files and sketches  and make the file smaller. I then export the DM f3d to STP for mold designers.

[O.Tan]

Hmm, since you'll just export to STEP, you don't need to delete all the hidden files/sketches as STEP conversion will only take the model selected.

[Anonymous]

@O.Tan

Hi, not really, there is a distinctive difference between the Parametric File and the DM file. I do keep a few files of the same designs in DM mode, but only one or two Parametric Files for future changes. 

A product could be made with different materials. For example, it could be wood, plastic, crystal, glass, carbon fibre, metal, gold, bronze, stainless steel to name a few. Each type of material requires different ways to separate the parts for fabrication and final assembly. DM mode has ways that are not available to Parametric Mode, for example using "Move" to change the face of a model directly., or to move construction planes to any position without switching back and forth to "Create Form" or "Base Feature" mode in the Timeline envornment. Clean DM files are easier to make products matching up with different manufacturing requirements. A Parametric File remains as the original design suitable for all envoronment. 

[O.Tan]

Hmm, I probably misunderstood what your original post as what I said is that a STEP file doesn't retain your sketch drawings or any form of intelligence, it just takes your solid model and convert it into a format which is translation friendly. So they'll be no size difference if the same part is exported in TM or DM.

Of course if you converted to DM cause you need to do some changes like as you mentioned, then there's a valid reason why you did so.