Coding Mistake in Revit Regarding Reference Lines and Hosted Profiles

[Anonymous]

Autodesk, I think you have a glitch in Revit 2018. Profiles hosted to References Lines are not obeying the orientation of the Reference Line. The behavior that is expected is the orientation of the planes associated with the Reference Line should dictate the orientation of any profile that is swept along the path of that Reference Line. I have a an example where the Reference Line has a dynamic angle and the profile hosted to it is not obeying the orientation. This is a mistake in the code. Please advise if this can and will be fixed in the near future. It is breaking down Revit's ability to model accurately.

[ToanDN]

Nothing seems to be wrong here.

[Anonymous]

Try hosting another Reference Line to the vertical plane of your first reference line. Then, Sweep Up that vertical reference line. In my case the profile orientation will slip every single time despite the orientation of the host reference line. If you would like, I can share my file so that you see the problem. 

[Anonymous]

See Attached File.

[Alfredo_Medina]

---

@Anonymous wrote:

Autodesk, I think you have a glitch in Revit 2018. ... I have a an example where the Reference Line has a dynamic angle and the profile hosted to it is not obeying the orientation. ...

What is a dynamic angle?

Also, the orientation of a profile depends on the direction in which the path is drawn. Left to right or right to left produce different orientations of the profile. But that is not a problem because you can flip the profile or rotate it.

[barthbradley]

I'm going to echo @Alfredo_Medina's question: What do you mean by "Dynamic Angle"? I thought you might be describing a Ref. Line angle driven by a labeled dimension, but there is isn't one. There is, however, an Angle Parameter named "Theta", but it doesn't appear to be driving anything.  What am I missing?  

[Anonymous]

Changing the radius parameter changes the angle of the reference line but the profile does not obey that change. Profile maintains the orientation to reference line only at the parameter state it was created in. I'm sorry that term was confusing. Please flex the parameters to see the error.

[barthbradley]

How? I don't see any dimension labeled with that parameter. Where's it hiding? In fact, I can delete that parameter without a warning that it's deletion with affect anything else.  In other words: it ain't associated to anything.   

...may I ask you: are you specifically referring to the sweep created with the profile named: "Profile - Generic - GFRC - Fin - Side" ? 

[Anonymous]

The parameter is called Module Radius. The value in the family as uploaded is 12'-0". Based on the rigging of the other reference lines, if this value is changed, the angle of the flanking reference lines change. Yes, the profile is the Fin-Side. 

[barthbradley]

Okay, so bear with me; I think I may be starting to get the gist of your question.  Are you wanting the Fin's sweep path to rotate? Yes, no, maybe so? 

If Yes: no can do. But, you could rotate/drive the Sweep profile to match the angle.  You would need to change the axis/sweep path though.  This would also necessitate modifying the profile family.  

[ToanDN]

You are talking about the two vertical end caps? A sure way to make them always align to the radius lines are using in-place extrusion instead of sweep, better yet, as a nested family then you can use sweep from the same profile.

[Anonymous]

I want the sweep path to obey the rules of the system. The Reference Line that defines the path is drawn on the vertical plane of the Reference Line that is hosted to the tangent of the Reference Arc. Querying the Reference Line we can see that the associated planes do indeed stay oriented relevant to it's host Reference Line. Picking this Reference Line as the path for the Sweep instantiates the Profile relative to that orientation. But the rules appear to be breaking once the driving Radius parameter is changed. All other elements of the system seem to be adhering to the rules just not the profile for some reason (eg. glitch in the application).

I don't see any other way to get it work without abandoning the family and rebuilding from another series of templates (adaptive component), or nesting several generic models with duplicate rigging systems into each other (which is dumb).

Driving the angle of the profile family itself is untenable due to the variable orientation of the profile and the path chosen along with the convoluted math needed to solve for that angle. That's why it appears that the application is itself faulty for not being able to deliver consistent geometric results. Unless you have another suggestion I'm not thinking of?

[barthbradley]

Sorry, I don't see it the way you do, @Anonymous.  I see the sweep path as being aligned to axis point.  I don't equate that axis point to a face.  The Ref Line's workplane faces are revolving around this axis.  The axis is stationary. 

...actually, I don't believe that describing the axis as "stationary" is even correct.  It's just the mathematical center and it has NO orientation.  

...Still, I think you can accomplish what you want, by associating the Profile's Angle Parameter to the Family's Angle Parameter. 

[barthbradley]

Interesting thread. I like it. You're bringing up an interesting concept, @Anonymous. 

Following...   

[ToanDN · Accepted solution]

@Anonymous

Set the radius line reference line as the workplane when create the sweep, use Sketch Path (not Pick Path) and pick line the reference line, then lock the end of the sketch to the top reference plane to constrain the height.  It flexes fine afterward.

By the way, an easy way is to use railing with profiles assigned to start/end posts and top rail/rail.

[barthbradley]

Nice @ToanDN. 

[Anonymous]

@ToanDN That definitely worked. I still maintain that there is some misconstructed logic regarding the Reference Line orientation relative to the host. Geometry hosted to a Reference Line that is itself hosted to the plane of another Reference Lines should adhere to that orientation. This is definitely a mistake in the application from what I can see barring any logic that Autodesk may be able to provide. I really appreciate your time providing a solution! Thanks so much!

[barthbradley]

I think you may be right, @Anonymous.  In testing this, I discovered "axis" does indeed have an orientation and does revolve.  It's clear as daylight if you edit and redraw the path (using the Pick Path tool) after changing the angle. When you do, the extrusion along the sweep path changes it angle.  Similar to the Update to Face for Mass hosted elements -- I manually updated the Sweep Path and the extrusion changed its orientation.  

Interesting. Maybe an Idea?