I received so much help last time (barley twist) that I thought I would ask about the best practices for tapered fluting in SU. There's probably at least a half dozen ways of doing this but I know some of you more experienced folks can save me some headaches. This image is a fluted newel that we produce that i'm hoping to model in SU.

@jim57 said:

@archturn said:

Really fantastic! You don't see a table like that everyday

And thank God I don't have to make one like that every day, either!

Jim

But a beautiful piece to hang your hat on.

@jim57 said:

Here are some barley twists I made for the base of an altar for a church in Texas. They were machined on a CNC. The model started in Sketchup but I had to go to Rhino to tweak the legs.

Really fantastic! You don't see a table like that everyday

@jim57 said:

Sorry, thought the photos posted. The model is a bit large at 2.2 MB.

[attachment=1:29meov3i]<!-- ia1 -->Sacred Heart Altar1.jpeg<!-- ia1 -->[/attachment:29meov3i]
[attachment=0:29meov3i]<!-- ia0 -->Sacred Heart Altar2.jpeg<!-- ia0 -->[/attachment:29meov3i]

I'm wondering how the wood was laminated - face to face? What kind of wood?

@jim57 said:

Sorry, thought the photos posted. The model is a bit large at 2.2 MB.

[attachment=1:1gqmkiwh]<!-- ia1 -->Sacred Heart Altar1.jpeg<!-- ia1 -->[/attachment:1gqmkiwh]
[attachment=0:1gqmkiwh]<!-- ia0 -->Sacred Heart Altar2.jpeg<!-- ia0 -->[/attachment:1gqmkiwh]

Reminiscent of the columns in the Vatican. Pretty fantastic!

@jim57 said:

Here are some barley twists I made for the base of an altar for a church in Texas. They were machined on a CNC. The model started in Sketchup but I had to go to Rhino to tweak the legs.

About 25 hours each for cutting, matching and gluing, plus 30 hours milling and 30 hours of sanding. The plates at top & bottom were hand turned. Both the spiral and the worm diameter taper. The gullet follows along a different taper.

I have more photos of the set at http://www.SpringwoodStudios.com It's the first time I've done a whole church in SU. Thanks to all on the forum who helped with the tricky bits

Wow
Pretty incredible!!

@gilles said:

It is the interest with components.
In my file the newel with the twist is about 690Kb, duplicate X100 it is only 1178KB...
Imagine 690X100.

Pulling my last bit of hair out - The uploaded file is 2.4 MB. How can I reduce the the size of the file? It is already composed of components. How did you take my large file last time and get it to less than 1 MB
N112.skp

@dave r said:

That's not overly useful by itself but I expect it'll result in a fairly large file size by itself. Is that for the entire newel post sans twist? You could actually make the newel post as a half or quarter, make that a component and copy rotate it as I showed you the other night.

That is the newel without the twist.
I could make the newel as you say but that doesn't account for how gilles could take my existing file and reduce it to such a small size. I'm missing something.

Any thing interesting here: This is my swept newel no twist yet

@dave r said:

How many segments did your imported profile turn out to have? The profile I created last night with the Bezier tool had 22 (including the two little short segments I added at the ends. And the total number of entities after one turn of the screw was 2686.

Dave
How can I determine that?
Also I made another barley twist profile, ran the screw plugin and rotated like you did (7 1/2d for 24 seg.). I could not grab the edge of the component for copying and moving. Why did it work for you?

Re: file size of my barley twist stair newel
I'm still perplexed by the size of my file. I started over with a new SU file and made my newel without the barley twist. After imported my profile from CAD, did a sweep, created a component of the newel and my file size is 3 MB. What gives?
BTW I have purged layers from the CAD file import

@dave r said:

That's pretty much it but use a component instead of a group.
Still not sure why there is such a difference in file size

@pbacot said:

This thread is hot! Another point I would make is with screw.rb one can then make a component of spiral section that would stack directly beneath the other (without turning) and so you could increase, vary the height (within a module) easily.

I assume that can be done with Gilles' version.

If I understand you right this is one method I used - I made a 2 1/2" profile (the profile of my barley twist router cutter) then used the screw plugin to make a 5" high one revolution spiral. I grouped it, then copied and stacked it on top the other spiral. I then selected those to array the rest of the way.

@dave r said:

@archturn said:

@dave r said:

I'd be happy to show you. I can't do it at the moment, though. Maybe another live demo with you or I'll make a video.

Dave
That'd be great!!

Maybe a live demo this weekend?

Sounds good

@dave r said:

I'd be happy to show you. I can't do it at the moment, though. Maybe another live demo with you or I'll make a video.

Dave
That'd be great!!

@dave r said:

Gilles version of your model illustrates very clearly the reduction in file size that I mentioned earlier.

And even 100 of them is only 820 Kb. That's a couple of hours of production in your shop, right?

Wish I could make them that fast. I missing something - would like to know your work flow on the spiral

@gilles said:

Here is your model with some corrections and the use of components,note it is 650Kb instead of 1950.

Looks better than mine. So you redid the spiral?

@dave r said:

My preference would be to make a component from the git go and make the copies. Make Unique is a trivial thing when it comes to that. For a model such as yours, starting with components makes it easier to see how the whole thing is developing as you edit. If something isn't going the way you expect, the earlier you figure that out, the easier it is to back up and fix it.

i started to write more about using components instead of groups but I won't get on my soapbox about that now. I'll just say I have never once in more than 9 years of using SketchUp found a situation where a group made more sense than a component.

Got it, will do
Thanks

@dave r said:

No. But when you use groups instead of components and need to edit them, you have to touch every copy of them to make the changes. If you use components you only need to edit to modify all copies. If you look at the twist I posted you'll see it is made of three identical parts. Each is a component so modifications to one will propagate to the others.

You could split the barley twist into smaller identical units to make construction and editing easier and faster. It'll also result in a lower file size.

For a two start twist, my workflow was to create a one turn twist, group, hide the edges that would join the copies, copy and rotate 180 degrees, copy those two twist and array upwards. It seems since I have hidden the adjoining lines at first this shouldn't be a problem. (But in fact, I see that the finish model shows even rendered where those hidden lines are.)
I understand the concept of using components and the advantage in terms of editing one propagates to all. My thinking was to edit the group first then copy and avoid "make unique" for the bottom and top since the top and bottom have to be uniquely edited? Is my thinking right here? Am I missing something else?

@dave r said:

The unneeded edges I was referring to can be seen if you turn on Hidden Geometry (View menu)

Also more missing faces at the bottom.
[attachment=1:zqm4jyv1]<!-- ia1 -->unneeded.png<!-- ia1 -->[/attachment:zqm4jyv1]
[attachment=0:zqm4jyv1]<!-- ia0 -->unneeded2.png<!-- ia0 -->[/attachment:zqm4jyv1]

If you used components instead of groups you would reduce much of the work needed to draw this. That applies to other things, too.

Are you saying that using groups instead of components is the reason for the unneeded edges?