Parallel Edge Joint?

[therealsamchaney]

Is it at all possible to create a parallel edge joint? In other words, I want to take an edge of a component and force it to be parallel to an edge or axis on another component but I do not want them to be colinear. 

For example, look at the attached file. I have a simple D6 cubic dice which I want to be centered in the circle in the Dice Arrangement sketch which I've done using a revolute joint about Z. But then, I want to constrain the front edge of the dice (as seen in the Edge sketch) to be parallel to the X axis of the assembly which doesn't seem possible. I can't do a revolute joint about X because it would clash with the existing Z revolute joint. I need a parallel joint about X but it doesn't seem to exist.

There is the revolute joint but that only allows rotation in X and translation in X so it's too restrictive. All of the other joints seem to have similar failings and I cannot seem to find a way to achieve this without some hacky workaround.

In Solidworks, this would be easily achieve with a single parallel mate between the two edges. There are 6 DOF so there are 6! or 720 possible combinations. That's why it doesn't make sense to me to try to use just a handful of joints like Fusion does. Instead, users should be allowed to close off each degree of freedom one by one using simple 3D constraints. 

I tried Align but that gives an error if you do it after the joint, and if you do it before the joint, then it just gets cancelled out by the joint anyway.

Any help would be greatly appreciated.

[davebYYPCU]

Edit your joint, in the dialogue portion, the angle 180 is provided subtract 26.565 degrees from it.

Might help....

[TrippyLighting]

---

@therealsamchaney wrote:

That's why it doesn't make sense to me to try to use just a handful of joints like Fusion does. Instead, users should be allowed to close off each degree of freedom one by one using simple 3D constraints. 

.

---

The joint system in Fusion 360 covers at least 90% of all use cases and does to much more efficiently than the usual 3-mate paradigm in other CAD software. That is particularly true for top-down designs , which allow the use of as-built joints.

There 3 or 4 joint types missing and I know “they” are working on that.

However, I do think that in some situations the traditional mates work better. I believe it would be a difficult task to convince the Fusion 360 team to implement traditional mates in parallel.

[therealsamchaney]

Thank you but I would put that under the hacky workaround category. I know I can just measure the difference and hard code that into the joint, but that's a brittle solution. If I add in any other models which are at different angles, I'll have to do that for each of them, and if the bodies in the components ever change orientation in the future (while linked) it will break the assembly.

I'm looking for a simple and reliable joint to (automatically) keep one edge parallel to another.

[therealsamchaney]

@TrippyLighting Thanks, yeah I get that it captures the majority of common cases and streamlines things in those cases, but I feel that while that might work for an art program or something, leaving out the corner cases is not acceptable in an engineering CAD program. When designing precision assemblies like motors, being able to finely control each joint is important.

Maybe we will never convince Autodesk to implement traditional mates but I will hold out some hope. I'm hoping they will add 3D sketch constraints for the 3D sketch mode, and if they did that, they would already be most of the way there to making mates.

[davebYYPCU]

Your problem is self inflicted, and then you want Fusion to fix it.

Had you used the component origin in an orthographic manner, there wouldn't be any need for a hack.

Matter for you.

[TheCADWhisperer]

---

@davebYYPCU wrote:

Your problem is self inflicted, and then you want Fusion to fix it.

---

@TrippyLighting wrote:

The joint system in Fusion 360 covers at least 90% of all use cases and does to much more efficiently than the usual 3-mate paradigm in other CAD software.

I believe it would be a difficult task to convince the Fusion 360 team to implement traditional mates in parallel.

---

All this stuff just 100% works in Autodesk  Inventor Professional (and in SolidWorks) and did so long before Fusion 360 even existed.

They used to give this spiel about how classical history based modeling was tried for 20 yrs and doesn't work.

6 degrees of freedom was understood long before there was any CAD and logical removal of each unwanted DOF is simple in other parametric CAD softwares.

The Fusion developers should "walk across the hall" and ask their Inventor Professional teammates how it is done.  I have noticed over the years that Fusion has steadily grown to resemble Inventor Professional and SolidWorks.

I just purchased a new iPhone.  I can image Steve Jobs rolling in his grave if my new phone didn't do what my old iPhone 5 did years ago (and a bit more).

[davebYYPCU]

Well a Solidworks user, did not use the Born technique, and objects to having to measure the angle,  he may have missed the point, anyway.

If using Fusion (not compulsory), why not learn the Fusion interface?

[therealsamchaney]

---

@davebYYPCU wrote:

Your problem is self inflicted, and then you want Fusion to fix it.

Had you used the component origin in an orthographic manner, there wouldn't be any need for a hack.

Matter for you.

---

Sure, but all problems in CAD are "self-inflicted". Just because my design comes up against a use case that the Fusion dev team didn't account for doesn't mean I'm doing something wrong, within reason at least. I'm not asking them to allow for shapes with negative volume or anything crazy, just to allow for basic mates that most other CAD programs have. I'm not the only person asking for this.

And, of course there is a reason I didn't make the model orthogonal to the origin planes to begin with. I didn't stand the dice with its vertex pointing straight up by accident. The file you're seeing is a dummy version. I've removed all of the valuable IP that I can. The dice component is oriented that way because I also designed parametric support material for SLA 3D printing which connects to the edges of the dice where they won't affect the faces. This is the best way to orient the dice for printing, so the component was designed with that in mind. I'm using that dice component in this assembly but now I need it to be orthogonal. Simple mates would allow me to achieve that easily, and in a way that is flexible with parametric design. I don't want to break the link because I want the dice component to be a living file that can be updated and this assembly should update with it. 

[therealsamchaney]

---

@davebYYPCU wrote:

Well a Solidworks user, did not use the Born technique, and objects to having to measure the angle,  he may have missed the point, anyway.

If using Fusion (not compulsory), why not learn the Fusion interface?

---

Well, the thing is, Fusion is in a way compulsory because it's the only really capable CAD program that individual creators and startups like myself can afford. SolidWorks recently started a $99/year option which was tempting but it's capped at $2,000 profit per year which doesn't work. If the cap was higher like the Fusion startup license is, I honestly probably would have gone for it and jumped ship even though I've sunk so many hours into Fusion. Believe me, I love Fusion 90% of the time, but that other 10% can be pretty painful and frustrating. You might call it a love-hate relationship.

And, I do know the interface. I understand the joint system and like I said above, I get the appeal for the use cases that it covers (which as @TrippyLighting said, is most of them), but there are still some valid cases that they don't cover, and there is nothing in Fusion to fill the gaps. 

Heck, if Autodesk offered a monthly paid version of Inventor Pro that was close in cost to other subscription programs like the Adobe suite for instance, maybe something like $20 - $50 per month, I would do it in a heartbeat.

There is a big gap in the market for individual engineers / industrial designers and currently Fusion 360 is the closest thing to filling that role, and it's almost perfect but has problems that can't be ignored. That's why it's so frustrating when I run into things like this where I'm unable to do something that would be trivial in most other CAD software, and when I ask about it, people just say "don't do the thing you want to do". It's like when you go to the doctor and tell her that it hurts when you move your arm and she says just don't move your arm. I don't understand why so many people jump to dismiss the desires of users. If you are fine with the joints system that's great, but I'm not and I think it's totally reasonable for me to request it be improved. Why do you feel the need to try to shut me down?

[TrippyLighting]

---

@therealsamchaney wrote:

I don't understand why so many people jump to dismiss the desires of users. If you are fine with the joints system that's great, but I'm not and I think it's totally reasonable for me to request it be improved. 

---

I find your request perfectly reasonable!

I've come across a number of situations, where a simple mate would have been very helpful, but instead I needed to use the one or the other clumsy workaround. 

BUT ...

Amongst all the other established and more mature players Fusion 360 is very young and still very much a Work In Progress. Just because a specific functionality does not exist today, does not mean there aren't efforts underway to add it.

The frequent comparisons with SolidWorks and Inventor are not only often unfair, but also somewhat single sided.

A wile back I created my own modeling challenge and compared 4 CAD systems. Fusion 360, ZW3D, Inventor and SolidWorks.

Granted, the model was very specific and used a very specific tool set. I was able to create the model without problems but only semi parametric in Fusion 360.  Parameter changes were stable but required manual rework due to lacking pattern tools. In ZW3D it was a much easier workflow, and fully parametric but I also had to find a workaround for a bug. It was indeed very tricky to create the model in Inventor but ultimately I persevered and found one workflow, but the model was somewhat fragile in that some parameter changes worked and others not. In SolidWorks I was not able to even complete the model (I've worked with Solid Works for 15 years!), and I was also baffled how many fields in the creation of that specific model did not accept user defined parameters, something I take completely for granted in Fusion 360. 

In essence, all  tools have their limitations if one just looks closely enough!