Timeline Capture / Joint vs. Move / Component Origins

Timeline Capture / Joint vs. Move / Component Origins

0422jangg
Explorer Explorer
503 Views
10 Replies
Message 1 of 11

Timeline Capture / Joint vs. Move / Component Origins

0422jangg
Explorer
Explorer

Hello, everyone.
I hope you're doing well.

I have a few questions regarding design practices in a professional setting.

I’ve been doing CAD design for about 5 years — not in a company, but as part of my graduate research. So I would consider myself an intermediate-to-advanced user of Fusion 360.


In my work, I often deal with assemblies that contain thousands of parts. As an example that reflects my typical workflow, I’ve attached reference files from the Voron 3D printer project.

I’m curious about a few things, especially from the perspective of what’s considered “professional” or standard practice:


* Timeline Capture
I usually disable timeline capture when working. While this limits parametric control (like changing dimensions through parameters), I find it helps improve performance, especially in large assemblies. Is this a common practice in professional environments? Or is it generally better to keep timeline capture on?


* Joint vs. Move
Instead of using joints, I often position components using the move command. What are your thoughts on this approach? Is using joints generally preferred or expected in industry-level design?


* Component Origins
I typically don’t start from the main design origin. I create new components and sketches at various locations as needed. As a result, each component or body has its own unique origin point, scattered throughout the space.
Is it considered best practice to align everything to the main origin? Or is this not so important in a practical setting?

Thank you very much in advance!

0 Likes
504 Views
10 Replies
Replies (10)
Message 2 of 11

Drewpan
Advisor
Advisor

Hi,

 

There are a few things that the forum commonly recommends and for good reason. Some of

it is because it is best practice, some because of work arounds of known bugs or ideosyncrasies

of fusion, an sometimes it is the preferred method for certain groups doing certain design work.

 

One of the things about fusion is that it is parametric software and has the very powerful timeline

for a user to use. Yes, the timeline will affect performance in complex designs, but in many cases

it is the timeline that can save hours, days, weeks or months of redesign if you get to a position

in your design where you are stuck, and the only way to fix it is to go back in time. For this reason

it is NOT common to disable the timeline in Professional settings. Even for relatively simple designs

it is always better to keep the timeline on and learn how to use it properly. If your performance

is degrading significantly then there may be several underlying reasons and the timeline can give

you notice that these problems exist and need to be fixed. "The Red and Yellow flags in the timeline

should keep you awake at night" is a common mantra told to beginners because it says that your

design has problems and they need to be fixed. If you are not using the timeline and have turned it

off, how do you know there are none of these problems in your design? Ultimately if you have severe

performance issues in fusion then there is a reason. Sometimes those reasons are multiple complex

assemblies imported from several different sources, but often performance is affected by bad

workflow. There are many large and complex designs that work perfectly fine in fusion with good

workflow, so maybe your performance problem can be fixed with more grunt (processor, memory and

graphics) on your desk top.

 

DO NOT USE MOVE unless you THOROUGHLY understand it in fusion. The Move command exists and is

useful WHEN USED PROPERLY, but many users do not actually have a clue when they should use move.

If you have multiple Components and or Assemblies/sub-assemblies then you should ALWAYS use a Joint,

or in SOME circumstances use Group. One of the biggest issues beginners run into is that they move

bodies and components into place with the Move command and then think that they are joined together.

They are NOT joined together without Join or Group. There are specific times a designer should use the

Group command and knowing when only comes with experience. In almost all other circumstances you

should be using Joint. One of the major issues with using the Move command is that it causes fusion to

do a full recalculation of the design for EVERY Move. If you have one or two this is trivial, but if you have

several in a row it hammers performance. If you have several moves in a complex design then performance

will slow to almost zero. If you need two things joined then Join them. If you just want to move things

around because you cannot see what you are doing (one of the most common reasons beginners use move)

then use the Visibility icons in the Browser tree. Without actual Joints, all you have is a bunch of things

floating in space in your design and one of the most useful things in fusion and modelling is checking that

the design will actually work. If that should be a Rigid joint - make the joint. If it should be a Sliding Joint

then make the joint and Test it. A design that is correctly modelled, no matter how complex, should work

if you have designed it properly and have the correct joints in place. Joints will be Expected in industry level

design unless there are known circumstances that you know about in advance.

 

All components need an Origin if you actually want to fabricate the design. Where those Origins are do not

matter but a design needs to be in relation to some known point to be useful. There are two methods of

design that are in common use. There are no hard and fast rules which one to use, often it is a matter of

personal preference and sometimes it is an industry standard. Some designers stick to one method and some

will only use a certain method under specific circumstances. Other designers will mix and match. Whatever

method you use there is only really one rule - be logical and consistant so that others can follow your design

at a later date.

 

In general terms, most designers start at the easiest component in their design and place it with reference

to THE Origin. This is not always the case, sometimes there are reasons to set an Origin of the part at a

different place, however this different place will ALWAYS be in relation to THE Origin and set up first. From this

starting point the designer makes a choice as to which method to use. The designer can use the Origin for ALL

of the components in the design and join them together in the correct place with Joints. Or the designer can

create new origins based on one or more components and design the new component In Place. The designer

can also mix and match these two methods. The only way to know which method to use will come with

experience. There are certain times when using one method will make things much more difficult than using

the other. There are times you can mix and match and times you should NEVER do so. The only way to know

will be with experience. My advice would be that when you start, use the method that is most familiar to

you, but also teach yourself the other method for when you need it. Find out if your industry has a preferred

method or ask people in your design team if there is a preferred method or an actual Design Standard. As a

designer it is good to be flexible and adaptable. Sometimes there are no specific rules and sometimes the rules

are written in stone, and in some cases blood (depending on the industry).

 

Hope this helps answer your questions.

 

Cheers

 

Andrew

Message 3 of 11

0422jangg
Explorer
Explorer

Hi Drewpan,
Thanks so much for your detailed and thoughtful reply. I really appreciate it.


I just wanted to check if I understood everything correctly:

1. It’s better to use the timeline, since it helps with design modifications.

2. Move can cause issues, so I should use joints instead.

3. It’s not strictly necessary to align all component origins to the main origin point.


Also, would it be okay if I ask a few follow-up questions?

For example, regarding fasteners — they are often repeated many times in a design.


1. In my case, I’ve created a library folder and uploaded various fasteners by type. I then insert them into my designs using external links.

2. Would breaking the link improve performance?

3. In another case, what’s the performance difference between "Paste" and "Paste New"? If I use one fastener multiple times, would "Paste New" (making independent components) be heavier than just using "Paste"?


Thanks in advance for your help!

0 Likes
Message 4 of 11

Drewpan
Advisor
Advisor

Hi,

 

Yes, you have the right idea with the first part of your post.

 

Absolutely it is better to use the timeline.

 

A Joint will also move the Component into the correct position and form a relationship

with the other component.

 

If you design each individual Component and then assemble then designing every part

at the Origin has advantages. In complex designs it is sometimes better to Design In Place

and create a new origin for each part based on other Components. Experience alone

will teach you when you should use either one OR mix and match.

 

My first question is what is the purpose of modelling fasteners? Is this the design that

is going to be used in marketing to create those fancy exploded instructions or is this

a practical engineering design to model the device and see if it works? Sounds simple

but make a HUGE difference.

 

If I am in the Marketing Department then yes I want all of this stuff and I want it to look

pretty because making it pretty is my job. If I am in the Design Department however, I want

to make sure the design actually works and I just plain don't need all those fasteners

that will slow down everything with unnecessary detail.

 

Marketing wants to see each screw with the washer and lock washer and the nut on the end.

They want the threads of the screw and nut to be modelled because that prints up nice in the

very expensive super-high definition colour laser printer or the offset printing contractor we

use.

 

Design however knows that the actual design will have all of these bolts, nuts, screws, washers

and lock nuts, but that we will buy them in boxes of 5000 from the suppliers. We will NOT be

actually fabricating each of these fasteners except for any custom ones and we want to keep

the costs down so will be avoiding doing that if at all possible. Design also knows that because

we will be purchasing this stuff that on the bill of materials we will just have 25 M5x0.8 x 20mm.

 

Design also knows that if you do actually insert all of these fasteners into the actual computer

model that all that will happen is that performance will be hammered keeping track of each

fastener and the threads.

 

Design simply wants the model to work so the simplest way to do this is just use a rigid joint

between those two components. EVERYONE knows that that joint will actually have those M5

bolts, but they are irrelevant for testing the model.

 

If Marketing want the design then insert all those fasteners because they actually need them

for their fancy posters, technical manuals and such. They should be the LAST thing you put

into the design. If you must put them in the design then at least do not model the Threads

unless you want performance to be in the outhouse. Having the links is better if you need the

fasteners. Breaking the links will speed up performance but the actual work of designing and

testing you should not use them at all.

 

Are you going to change anything after you Paste it? If you use Paste then if you change one

then you change them all. If you use Paste New then it will not affect the others. In terms of

performance, you are effectively breaking the link between components when you Paste New

and so there is less tracking under the hood of fusion.

 

Cheers

 

Andrew

Message 5 of 11

0422jangg
Explorer
Explorer

Thanks again for your help.

In my case, I often work on both large and very small designs.
For example, I sometimes have to fit electronics, mounts, and other hardware components into a space as small as 15 x 15 x 9 mm.

Because of this, I feel the need to include fasteners in the design — not just for assembly purposes, but also to check size constraints.


However, I’ve started to realize that my designs may include too much unnecessary detail, such as modeled threads on fasteners or components like T-slot nuts that don’t really affect the assembly due to enough clearance.


To improve performance, I’m thinking of removing threads from fasteners and excluding certain parts like T-slot nuts when they’re not essential to the design.

Also, I plan to keep the external library links, rather than breaking them.

Thanks a lot!

 

0 Likes
Message 6 of 11

Drewpan
Advisor
Advisor

Hi,

 

As a designer you should design to the brief and include what is necessary. I fully agree

that you need to check sizes and clearances. There is no point fabricating 100,000 of the

products only to find there is insufficient clearance for a vital fastener. You should always

also think about how you are going to fabricate your design. Do you need to actually

model a knurl when all the machinist is going to do is put the knurling tool into the lathe.

How is the machinist or CNC machine going to mount the blank to cut the part? You

would be surprised at the number of designers that design a lovely round part that needs

the machinist to use a 4 jaw chuck and dial indicator to line it all up, when if they had

thought about that they easily could have simply put two flats on the part and everything

lines up nicely without all of the extra work.

 

If performance is a real issue then a faster processor and more GPU and memory will always

help. That said, only put in your design what you need. Are you actually going to fabricate

that standard OTS fastener? Then don't put it in. If you do have to put it in then there is an

option in fusion to model the threads or not - just a simple check box, makes the world of

difference. Don't stress too much about performance. Good workflow will keep the design

clean and some users in the forum use very large and complex designs on a desktop without

too many performance issues.

 

Cheers

 

Andrew

Message 7 of 11

TrippyLighting
Consultant
Consultant

@0422jangg wrote:


..I usually disable timeline capture when working...


...Instead of using joints, I often position components using the move command. ...

 


In professional settings - I've worked with CAD, designing manufacturing automation machinery for over 30 years - no one* disables timelines (Fusion), or feature trees (SolidWorks, Autodesk Inventor and others ). Designs are broken into smaller external (linked) assemblies, that are assembled / linked into the overall design.

That allows you to isolate the timeline to these sub-assemblies and make design changes fast without having recomputes take too long.

 

If geometry from the main design or between different external subassemblies needs to be referenced, that's what Assembly Contexts are for. Use references thoughtfully and sparingly! 

 

If no geometry is created in the main assembly and it only consists of linked external assemblies and components, you could consider not having a timeline just for the overall assembly.

 

The reason you can get away with moving and aligning components is because you are not utilizing the timeline. Once you are using a timeline, you should not use move/position capture to position and orient components. You should use assembly joints to do that.

 


@0422jangg wrote:


* Component Origins
I typically don’t start from the main design origin. I create new components and sketches at various locations as needed. As a result, each component or body has its own unique origin point, scattered throughout the space.
Is it considered best practice to align everything to the main origin? Or is this not so important in a practical setting?

Thank you very much in advance!


If you design top-down (design parts in place in their final location) , most, if not all components will likely share the same as the original origin. 
Once you start designing parts bottom-up and move or joint them into position, the origin of the component moves with the component. 

As long as you are aware that origins are important, I don't see a problem!

 

* when I say "no one" that refers to the 90%+ market penetration SolidWorks has in machine design.


EESignature

0 Likes
Message 8 of 11

0422jangg
Explorer
Explorer

Hello TrippyLighting,


Thank you again for your detailed explanation. Could you please confirm if I understood your points correctly?


1. The reason I haven't had many issues using only the Move command is probably because I’ve been working with Timeline disabled. However, in a professional environment, it’s standard to enable the timeline and use Joints to properly connect components.

2. When using top-down design, it’s acceptable for origins to be distributed across components, and this doesn’t cause significant problems.

3. It’s best to divide the design into sub-assemblies stored in separate files and use external references in the main assembly to reduce computation in the top-level file.


I’d also love to ask a few follow-up questions... 🙂


1. In Fusion, is there any preference or difference (in terms of performance or workflow) between using "Rigid Joint", "As-Built Joint", or regular "Joint"?

2. Based on the Drewpan's feedback, I’ve started modifying my design with the timeline enabled.
In some cases, "Capture Position" has been helpful for adjusting parts. Do you see any downsides to using it occasionally?

3. Let me give you a quick example from my component level tree in my project:


Main_assembly.f3d (Level 1, Top level)
├ Gantry (Level 2)
│ ├ X Axis  (Level 3)
│ │ ├ XY Joint - Left (Level 4)
│ │ │ ├ Upper Unit
│ │ │ ├ Lower Unit
│ │ │ └ Mounting Fasteners
│ │ ├ XY Joint - Right
│ │ ├ Toolhead_Assembly
│ │ └ LM Rail
│ ├ Y Axis
│ └ Z Axis


In this kind of hierarchy, at which level would it be best to create individual .f3d files?
Should I break things up into many .f3d files at the lowest levels and then build assemblies step-by-step through file references?
Also, should I enable the timeline inside each of these .f3d files?
Or would it be sufficient to keep the timeline enabled only in the main assembly, and leave it disabled in the sub-assemblies?


Thanks so much again for your guidance.

0 Likes
Message 9 of 11

TrippyLighting
Consultant
Consultant

@Drewpan wrote:

Having the links is better if you need the

fasteners. Breaking the links will speed up performance but the actual work of designing and

testing you should not use them at all.

 


Breaking the links will N.O.T improve performance, it will do exactly the opposite!!!

 

There is a reason that distributed designs - designs with linked components and linked component groups (assemblies) - are faster!


EESignature

0 Likes
Message 10 of 11

TrippyLighting
Consultant
Consultant

I can confirm #1, 2 3.

 

 

 

New questions:

 

1. I don't think there is any significant performance difference between joint types.

 

2. Enabling the timeline on a design that started with the timeline and was developed without a timeline makes little sense. That only makes sense if you start designing this from scratch. If you are still in process of designing new external components and assemblies, those could have a timeline. I would try to avoid the capture position command as it captures the position of every component in the design that has moved since the last time position capture was used.
That is particularly performance degrading in designs with many components!

 

3. I would need to have access to your design and see it with my own eyes to provide details on how it could be broken into subassemblies.

 

Other stuff that you asked:

 

1. Copy/Paste and component patterns create instances. Those  a pointers to a data structure. You can copy/paste and pattern many instances, but the component data structure only need to be kept in memory once. Copy/Paste-new creates a duplicate but independent data structure occupying more memory.

For purchased standard parts  that don't need t be modified individually - screws, bolts, washers, nuts etc. you should always use copy/paste, not copy/paste-new!

 


@0422jangg wrote:


Or would it be sufficient to keep the timeline enabled only in the main assembly, and leave it disabled in the sub-assemblies?

 


No, if your main assembly is, say 80 or 90% done and does not have a timeline, do NOT enable it.

Subassemblies that are done and don't have a timeline don't need a timeline. Minor edits can still be done using direct modeling. That is fine!

 

If you create new subassemblies, consider designing those with a timeline enabled.

 

You might have picked up habits that are fine using a direct modeling approach, but that are hurtful in a timeline based design.

 

If you have questions come to the forum early and ask! 


EESignature

0 Likes
Message 11 of 11

0422jangg
Explorer
Explorer

Thank you for your answers. Your answers were very helpful. Thank a lot!

0 Likes