Using no zone, glue and magnets for precise relational positioning

[lure23]

There's been plenty of request for "snap", "alignment" and generally precise relational positioning outside of the sketch environment in Fusion 360. This was also brought up in the Aug 6th 2013 webcast. 

I'm coming from the software world, and started to think how similar stuff is handled there (mainly in web development and UI gadget layouts in general). This suggestion tries to bring that to CAD - with a hint of computer game feeling to it. It's intended to be both precise, flexible as well as intuitive to people who haven't used traditional CAD systems.

Implementing this would probably be a huge task, which I'm glad to leave in the hands of Autodesk. You may throw your criticism on this - I'm just trying to pump up creative discussion so that whatever the snap/alignment/precise positioning solution will be, it will be a good one.

That's more prefix than I normally put in my suggestions. Let's see the meat: 

Each face would have added properties to it, individually adjustable on both sides:

- "no zone" (red below) which means no other element is allowed to enter that distance from the face

- "sticky zone" (green jelly) which means any element with also a sticky area would get stuck with this one

- "magnet" feature (not depicted), pulling other magnetized faces either apart or towards this one

None, some or all of these features can be applied, for a particular edge. There are ways they can be nicely presented to the user using modern transparent, possibly animated 3d graphics.

For a face, the areas are 3-dimensional space like those produced with the push/pull tool (which I used in the very mock-up pictures).

If the box was to be moved to the right, the red "no zone" would start pushing the cylinder within it when the cylinder touches the "no zone" edge. I.e. no other object can remain in that zone.

The green jelly-like blob (the "stick zone" or "glue zone") has no effect on the cylinder now, since the cylinder is not having a stick face. If it had, however, and the box was moved to the left, the cylinder would follow, as if glued to the jello.

The stick zone can be of zero depth, I.e. there would be only one allowed distance between the stuck objects. Or it can be used as a range, as in this picture.

The magnet option (not depicted here) is required for positioning things in between two walls. Make both the left and the right wall have "magnet +" and they'll push and pull other magnetized objects (s.a. the cylinder's face if it were to made magnetized) to their center. In this case, moving the box would also move the cylinder. Moving the cylinder would either also move the box, or let the cylinder fall back to the center, unless moved completely outside of the box (which would break the magnetic relation).

Similarly, just bringing a stickable object into a stick zone will make it stuck. Moving it fully outside the stick zone (with maybe a nice gummy bridge forming in 3d between the two objects, until it breaks! - like the iPhone gummy scrolling effect) would remove the stuck relationship.

Now, would this cover the actual use cases for relative positioning?

1. Snap two faces

- make them both sticky (with depth 0 or anything you like); no "no zone"

2. Safety margins, i.e. nothing closer than 0.5mm of these edges

- make a "no zone" of 0.5mm

3. Snap in two (or three) directions, with measures

- make a "sticky" relationship in all two/three dimensions. Must have enclosing walls in all these dimensions (if not, one can use Joint Origins).

( There needs to be an option to not have the "no zone" to have this usable. )

4. Center an element

- set both surrounding faces to magnetism "on"

- set center element's face/faces to magnetism "on"

5. Place an element proportionally (i.e. 2:3 towards the right wall)

- set left face's magnetism to 2.0

- set right face's magnetism to 1.0

- set center element's face/faces magnetism "on"

6. Snap multiple objects at different distances

- make the anchor object have certain stickiness distance; vary the stuck distances of the other objects

This brings the options we'd need for a one face, in one direction to:

- distance 1: thickness of the "no zone" (if on) and/or start of the sticky area

- distance 2: thickness of the sticky area 

- no zone: on/off

- stickiness: on/off

- magnetism: on/off/numeric (on=1.0, off=0.0)

All of these only have effect within the particular component level (i.e. they don't stick to, prevent the placement or attract another component's elements).

Precise (editable) measurements can be shown alongside the visualizations, if the user so vishes.

The system can be applied also to edges (either single segments or full paths) by providing the normal direction (with faces the normal direction is always implicitly there). "No zone" then becomes a 2d sheet, as does the stick zone (or it's a scaled, offset version of the edge). The logic would be the same.

I'm keen to hear of a case / need you think couldn't be implemented with the above mentioned mechanism.

I am aware that this suggestion is far from traditional CAD world (it is, right?). But it's very close to the "physics simulation" world that people who've played Angry Bird and Amazing Alex are used to. It's intuitive.

The visualizations don't have to be on all the time. They should be shown when in 'move' mode (since that's where they matter). In addition, there can be a switch in the component level to force the visualizations to remain.

The system can be made to work with solid bodies as well as with T-spline bodies. The same logic applies, no matter how complex the surface of the element would be. The necessary offset calculations are already there in Fusion 360's code base.

What do you think of this?  

[kat.ingalls]

Hey Asko! Moved this to the IdeaStation. I like the principle behind this suggestion. Being honest, there's a lot of technicality here (stickiness, magnetism, zones), but it sounds really similar to the "snap" functionality in Adobe's CS products (Photoshop, Illustrator, InDesign). I also love the focus on moving "far from traditional CAD....It's intuitive." I say "Hear, hear!" to not being afraid to break away from tradition if it makes more sense from an intuitive perspective 🙂

[Oceanconcepts]

In one sense this is an expansion and development of the concepts in the "joints" tools to cover more cases, and allow for some latitude.  It has corollaries in graphic design, where you can attach elements to each other, and create standoff zones, these properties attached to design elements make playing with different layouts far easier. 

I can think of ways that something like this would free up the design process for complex collections of parts, where you are trying to position many elements relative to each other. It could be very liberating for complex organic shapes. Exciting idea. 

Ron

[Anonymous]

I disagree.

Frankly, there are plenty of poligonal tools that do this kind of creating. Pushing herds of polygon with magnets is available FREE in Blender.

I need CAD. The poligonal shaping tools are nice, but the whole POINT of Fusion360 is to do real CAD in an affordable format.

Building all the sculting tools found in (free) Blender would cost a great deal to the company (and in turn, us) for a simple design element. When you send these file to a mill or a production house, they want dependable numerics.

Blender falls flat at the mill (faceted, no numerics, no numbers for the machinist)... but then, it's free. Ther's no reason why you can't import Blender models... or any other expressive modeler you might have laying around (Rhino, Alias, MOI... anything that makes .obj).

Design is important, but you have to have more than just design to build products and mechanical designs. 

Blender has "Magnets", but it is useless for mold engineering... but I can derive a mold core with Fusion 360 easilly (like I did last friday). I can see when the part weighs so I can mix the right amount of material... in a snap.

Just like Adobe products, it is a real tool for building in the real world, those are the underpinnings... real parts... maybe thats important without adding a bunch of costly stuff to make designing more "fun"?

[lure23]

Hey, no need to get excited Ted/Tom? (forums makes it quite hard to see your name anywhere)

I posted this first under general discussions since I don't think it's ready as an idea. Kat moved it here. 

I'm absolutely in the same camp with you, needing precise measurements and constraints. Fusion 360 clearly lacks that snap/align/precision right now. I was simply opening discussion that it didn't necessarily need to be implemented in the "old" CAD way. Maybe I'm wrong.

What I think happened here is that since I only placed one mod picture without attaching i.e. a made-up dialog of the precise measurements of those no-zone and sticky parts, you took it like they are unprecise. Not so. Also, you keep mentioning magnets (maybe you have bad experiences of such in Blender - I don't).

Magnets I brought in only to solve the "in the middle" problem. It's already possible to place Joint Origins in the mid of two edges so maybe that's it.

It's just discussion.