Follow-me issues


#1

tower medallion.skp (1.5 MB)

Trying to re-model a medallion that was originally created in REVIT.
A profile of medium complexity traced around a path of medium complexity.

but it’s coming out WrOnG

It’s a detail on a tower-element of a hispanic-themed restaurant.

Please help if possible.


#2

I give up… After spending more than half an hour on this, I couldn’t make it work, not even with all the creative ideas I had. Several things to note:

• The Follow Me Tool has a “defect”. When an arch (or circle or any other curved object) follows a curved path, its end twists. I find this especially frustrating because the only workaround is to manually build your curve like a spider spins its web.

Here’s an example. The second result is supposed to be the correct one.

That’s one reason I think your follow me ruins the ornament.

• It is not a question of scale because I’ve tried scaling it up 10x but still, no result.

• Also, turning it into a component, copying and scaling the copy didn’t help as well.


That’s basically it, I think. Follow Me is a good tool until it comes to round curves, in my opinion. This post is very interesting to me. I wonder if other members will be able to help.


#3

That’s not a defect. The profile is projected to be perpendicular with the path when it isn’t already set up to be perpendicular. It’s easily managed, though.

In the case of the path for that frame. replace the arcs with half-circles that have been rotated so there’s a 90° corner.

Hang on. I’ll do it.


#4

Dave, I understand what you mean with the perpendicular-to-path part (that’s why I put quotes around the word). But just theoretically, do you think it’s possible to have a sort of “Smart Follow-Me”?

I used to have another workaround. For example I would draw extended lines from the arc and start the Follow-Me on the arch from afar, like in here:

In this case the profile followed the inner U line. And I just noticed that since you say they must be perpendicular, the other side therefore doesn’t align properly,

Alternatively, if you follow the outer U, the same thing happens with the inner one.
I know this is supposed to be normal behaviour but for a hardcore perfectionist like me, for example, it’s that thing that makes you pull your hair out, haha.


#5

fix the path and the profile and it works…

both are very oddly constructed…

john


#6

Oh right. I’ve forgotten to mention perhaps the most important. The profile’s edges were not on the same plane. I went ahead and constructed a brand new one with a new follow-path but still, I still got the same result.


#7

try this…

path.skp (76.6 KB)

john


#8

I’m sorry, could you upload a 2016 version as well, please? I didn’t upgrade to 2017 this time but would like to see the result.


#9

path_v8.skp (17.7 KB)

john


#10

Yes that is one of the results in the original model, but @JDZYN rejected it because the inner radius is does not match the original.


#11

I hadn’t spotted the two different paths…

so the OP will need to run two profiles and then combine…

john


#12

Mine is on the left. Is it close enough?


#13

In between dishes, diapers, and other Daddy duties… I saw you came up with this one.
Looks good. How’d you do it?


#14

I was going to try doing it at a different scale before I got swamped with client calls.
Thanks for all your efforts!


#15

@DaveR, ok wow! Just wow! :thumbsup: How did you do that? Looks fantastic.


#16

@JDZYN and @VahePogossian, thank you.

I divided it into individual elements and put them together. I made a straight section with the miter on one end and I made a quarter-circle arc section. I used 12 segments for the quarter circle not only to make a smoother shape but also to improve the junction between the curved section and the straight. It could have been done with a semi-circular arc but then I’d have to make the join on the right side, too.

Due to the geometry of joining a series of parallel lines with concentric circles, the join between the two sections can’t be a straight line. That is, if the offset between the circles matches the distances between parallel lines. You can see how a straight line doesn’t cross the intersections.

If you were a stone mason doing this for real you would either have a curved “miter” line or you would change the molding shape slightly. I expect it would be the former so the same profile pattern could used for checking the shape. It wouldn’t create a problem because most viewers wouldn’t notice since it’s a 3D dimensional shape and the other curves would mask the non-straight join.

In SketchUp you could leave the medallion frame as I’ve done it made of multiple components and group them together. If you were planning to do something such as 3D print this in one piece, you could explode all the components and make it one. Leaving it as I have drawn it will keep its file size down.
tower medallion.skp (212.3 KB)


#17

You’re wonderful! That was very informative. Thanks a lot. I’ll try to retrace your steps with some coffee and patience.


#18

@DaveR, I would like to thank you one more time. I spent my whole evening on this but I think it was worth it in the end. Following your instructions, this turned out to be something much more simple than what I was expecting. Still, needed 3 hours to grasp the idea. I hoped very much that SketchUp would be able to handle this, as it’d be a further proof that it’s no worse than the king of CAD (3DS Max). :slight_smile:

I’m not sure if I followed you 100%. I might have improvised a slight bit somewhere in the middle but I’m not sure. The multi-planed intersection was difficult and interesting at the same time. I also didn’t increase segment count for quicker results but of course, the more the smoother.

And here’s the file. I used centimeters instead of inches because I work with the Metric System and the Imperial System is unknown to me, unfortunately, but still, I tried to draw as close to your dimensions as possible. What do you think? :innocent:

Ornament 2016.skp (110.3 KB)


#19

It looks quite good. Next time you’ll be faster at it. :smiley:

They aren’t teaching cursive hand writing and they aren’t teaching the imperial measuring units. These children are growing up to be helpless! :laughing:


#20

Glad to have presented the challenge. Thanks again @DaveR for solving it
and Vahe for your efforts.