Strange Revolute Joint Issue

[RogerInHawaii]

I'm seeing what looks to me like very strange behavior with a revolute joint. I recreated the behavior ins a simple test design (attached).

I create two components, each consisting of a simple square block. I create a revolute joint between the two and it revolves as expected. I then create a sketch, using the bottom of the block that revolves as the sketch plane. I do a Project of the bottom of the block onto the sketch. I then use that square sketch to do an extrude/join, extruding it to create another block's worth of body and joining that body with the existing revolve-able block. I try the Animate Model option for the existing revolute joint and it still works fine. The entire (now longer) block on the left rotates just fine.

BUT. BUT... I then select to Edit Joint Limits of the existing Joint so that the limits are very different than the initial, default values. I set the min, max, and rest values all to 45 degrees. And while the edit dialog is up it looks quite fine, in that the entire left (now longer)block is rotated as expected.

BUT BUT !! As soon as I hit OK on the Edit Joint Limits dialog it looks really weird. The original rotate-able block is rotated as expected, but the block that had formed the extension of that block is still in its original position. When I select to Animate Model it rotates but not as expected. I don't have a single, extended block, but rather one two blocks, one of which is tilted relative to the other block. Very weird.
 
Looking at the sketch that I used to create that block extension I see that it is still oriented the way it was originally, that is, oriented to where the bottom of the original block USED TO BE, not where it now IS (i.e. rotated 45 degrees). I would have expected that when I initially selected the bottom FACE of the block as the Sketch Plane for the sketch I was about to make, that it would REMEMBER that face, and even if that Face later got moved or reoriented, the sketch would also be properly oriented to it. But apparently not.

This just seems very weird. Why, after changing the revolute parameters, does it not properly reference the FACE that had been previously specified as the plane for the sketch? 

Here's the sequence that I went through.

[Beyondforce]

Dear @RogerInHawaii ,

1. You forgot R.U.L.E #1! Activate the component that you are working on.
2. Grounding component comes before Joints.
3. Joints like any other features are history-based features. You have created the Revolve Joint before the last joined Cube, and that was your biggest mistake. Because of that, when you changed the angle to 45, you changed something in the past before the last cube!

Here is the link to the video I have created for you (for some reason I don't have the Insert screencast!)
https://knowledge.autodesk.com/community/screencast/3c308153-ea60-436f-87df-6624f24f2485

Cheers / Ben

[davebYYPCU]

Sorry but your Joint Limts are not OK.

45,45,45 means nothing moves.

Try -45, 45, 0, in that order down the page.

Edit, Just watched Bens video, he did not find the limit problem, but what he says needs to be taken into account.

When I fixed your limits, the whole model behaved as expected.

Joints don't have to be at the end, making components from bodies is ok, and ground one component for best results, even though it works in this simple model, without it.

Might help....

[RogerInHawaii]

Yes, if you edit the joint limits to be -45, 45, 0 then you don't see the symptom that I'm seeing. But that's because setting the limits to those values means that the default position of the joint is straight down through the body, i.e. the top face of the lower body is coincident with the bottom face of the upper body, and that's because you see no default revolution of the upper body.

Try 15, 45, 20 as the joint limits. That results in the top body being tilted relative to the bottom by default, and THAT results in the symptom that I'm seeing.

[RogerInHawaii]

You are correct that making sure that the Revolute Joint feature resides in the timeline AFTER the full construction of the body to be rotate-able will avoid the symptom that I'm seeing. But that doesn't explain WHY the symptom is occurring.

When you create a Sketch and specify that the sketch's plane is determined by a particular face on an existing body the assumption (well, MY assumption) is that the sketch feature itself, that is the data structure that defines the sketch, contains not only the various lines that constitute the sketch itself, but also contains the identity of the basis for the sketch plane, that is the identify of the face on the body that was chosen during the initial creation of the sketch.

That SHOULD mean (shouldn't it?) that whenever that Sketch feature in the timeline is processed it should use that Face-On-Body sketch plane identity to locate that body and that face on that body in order to figure out the plane that it should actually use as the plane of the sketch. That should mean, at the very least, that if the referenced body later gets deleted or the body gets changed in such a way as to eliminate (or re-number) the referenced face on that body, the Sketch feature should get yellow-flagged with a warning indicating that it has lost its association with the necessary sketch plane.

It should also mean that if the body simply gets moved or re-oriented, as can happen in a number of ways, the effective plane used by the sketch should likewise get moved.

But that does not seem to be what's happening, at least not for re-orientation resulting from a change in the Joint Limits of a joint.

It SEEMS to me that when a sketch is being created and you specify a face-on-body as the basis for the sketch plane, it internally, invisibly, actually creates a construction plane based on that body/face and it's that construction plane that gets stored with the rest of the data for the sketch, NOT the identity of the body/face that was specified by the user as the basis for the plane to be used by the sketch. As a result, what happens when the body gets moved or re-oriented is that it has NO IMPACT on the location/orientation of the sketch. Rather, the sketch remains in the original position/orientation that existed at the point in time when the sketch was originally created. And that just seem, to me anyway, to be counter-intuitive.

From the user's standpoint he's effectively drawing ON the (extended) face of the body. If that face later gets moved or re-oriented then it's perfectly reasonable for him to expect that the sketch should still be ON that face. But it's not. It gets left back at where the face existed at the time the sketch was originally created. And that's just confusing. There are many other situations where a Feature loses track of some necessary parameter, often because some change had been made to a critical part that exists prior in the timeline. In those many situations it almost always results in a yellow/warning indication for the affected feature. In fact, a warning is almost always generated when the change-to-prior-feature will indeed have a negative impact on some subsequent feature. But that does not happen when you edit the limits of resolute joint.

Consider the situation where, instead of specifying a Face-on-Body as the source of the sketch plane, you specify an actual, existing construction plane. If you later DELETE that construction plane you'll get a warning when you go to do the delete (will affect subsequent features) and the related sketch feature will appear as yellow with the warning that it has lost reference to the sketch plane. But if you specify a Face-on-Body as the sketch plane source and later move that body there's no warning of any kind and the sketch does NOT follow along with the move of that body. And it appears to me that you don't even get a warning if you actually DELETE the referenced Face-On-Body sketch plane source.

It just seems ... weird ... to me that it works that way. I expect that the sketch feature remembers the sketch plane's "source" and references that source each and every time that the sketch feature is processed, meaning that any change to the source will have an impact on the processing of the sketch. But it doesn't do that. The sketch doesn't "track" with changes to the source and, as a result, bodies subsequently generated (e.g. via extrusion) don't end up where expected.

[Beyondforce]

You should always remember "History-Based Modeling" and what it means.
If you modeled in Direct modeling, then it wouldn't have happened.
When working in History-Based Modeling the sequence of features and what you do matters! Which is why workflow in Fusion is very important.

[RogerInHawaii]

You're still not understanding or addressing the question that I asked.

[laughingcreek]

Sketch2 is empty.  was that where you were trying to project geometry into?  extrude3 is based on sketch1, not sketch2.  It seems to be behaving as I would expect.

[davebYYPCU]

My take on the sketch plane problem.

The timeline is before the last extrude, 

Sketch 1 in the Rotator component, is visible, it was created on the same plane as sketches in the root area.

When the rotated block is in the current position, the projection into that sketch updates to only the visible face, as a purple projection.

How did this happen?, A root sketch was visible when you selected the underside of the block for the creation of this sketch, and therefore Fusion selected the front article to sketch on, not the face of the block.

This file highlights why we recommend Rule #1 and or the Eyeballs, for clutter / neatness.

To fix it, delete the 3rd extrusion.  With the tilted block, for the Rotator Sketch, redefine sketch plane, to the base of the block, delete the new sketch content, and project the profile if it was not automatic, update the last extrusion to this profile.

Now the sketch follows the joint movement.

Might help.