Utah, AI & 3D Printing - Oh My...

[vr...@optonline.net]

One might say that this email is long overdue - as it does indeed have to do with my travels across the USA as I moved from SoCal to Ohio. But, there was a good reason for the delay...

 

So, as I lifted up stakes & Hit the Road to go Back across the USA again - my 1^st stop was Utah. OF course, a Great reason to stop there was my good buddy Joe - the 3D Printing Professor. It was my 2^nd time to visit him in Utah at the Maker Space in the Library where he worked - to oversee the Maker space. He's ALWAYS a Fun guy to hang out with. This was a shot of the 2 of us hanging out in the Maker Space:

 

So, whilst we were hanging out - and talking all about 3D Printing stuff - I brought up this one project of mine that I keep trying to get back to. A project where I have gone down a NUMBER of paths, in regards to the design of the shape. The 1^st iterations of the design actually started back in like 2011-12, after I built my CupCake 3D Printer. 

 

At one point, I tried a different approach to the design - by simply sculpting it in something like Sculptris - to 3D Sculpt the 3D shape in the computer. I did this one version of the design, and I printed 1^st down in San Diego at SD3D. But, getting the chance to 3D Print it at Essentium - where I could print it REALLY Big - that was COOL and FUN to do. This was that print as it had just completed on the Essentium printer:

 

Anyway - as I was discussing the project with Joe the 3D Printing Professor at his Maker Space job - and he then took at quick shot at doing the design in Blender - as we were throwing ideas around. But - here's the Delema...

 

I have 3 main requirements for this design:

All branches of the tree must be hollow - so a wire can be dropped down thru a tree branch tip - and go all the way down to the root of the structure. 

It needs to be printable without supports. 

It MUST be Very Easy to do design iterations - such that maybe a single design iteration must take like 5-10min each - so that I can produce 500-1000 versions of the design, and even more.

 

So, first Joe jumped on a PC and booted up Blender - and did a little work to see if he could come up with an easy way to do it. But, there were issues with that attempt. Then suddenly he jumped up, wiped clean this whiteboard and started drawing and discussing his design algorithm - which, once I had completely wrapped my mind around what he was trying to do - I realized it was BRILLIANT! This was his whiteboard sketching:

 

I did ask Joe in another thread if he minded if I posted this - but, I didn't hear back - and figured it would be fine...

 

Now, I won't go into the entire algorithm - but, will give a brief rundown of it. In the end, the idea would be to Directly generate GCode. Briefly, one would generate circles, stacked up. But, then - the circles would split and move apart - and thus the branching. Think of it like a human baby growing, and how you have a single cell, then it splits, and each sub-cell then splits again, and so forth. 

 

I first began exploring this project by doing it in FoxPro. And, I made some progress on it. From drawing circles, to drawing 2 circle intersections, but, also make it so that when they overlap - the circles are not drawn in the intersecting area. This is a screenshot of the one program, in VFP, the drawn circle & not drawing it in the overlap area:

 

Admittedly, the progress with FoxPro coding was a bit slow. And, I began to see more complications with it.  And, I was even using AI to assist with doing it.

 

I then decided to try a different route. As in, still using Joe's algorithm concept - but, NOW - doing it via OpenSCAD.  And, do accelerate the whole project - I again turned to me new best buddy - Jean Claude von Ay-Eye.

 

I began doing things in steps. I first wanted to see if I could get Claude to just generate a squat hollow cylinder - based upon parameters like inner diameter, height, wall width - things like that. It did. Then, I figured, let's see if I can have it generate a Cone - but, via these stacked cylinders diminishing in size - as that's basically how I would grow the tree. And - it did:

 

Then, next - I wanted to work on the concept of the circle splitting apart - like in the FoxPro shot above - and I got success:

 

And, for anyone curious as to some of the OpenSCAD code being generated by AI which I am using - here's the Script for the pic above:

// Two connected hollow cylinders with all internal space empty

 

// Variables for easy modification

inner_diameter = 83;  // Inner diameter in mm

outer_diameter = 85;  // Outer diameter in mm

height = 10;           // Height in mm

shift_distance = 47;   // Distance to shift the second cylinder in mm

extra_height = 0.2;   // Extra height for inner cylinders to ensure clean subtraction

$fn = 100;            // Set the number of facets for smooth circles

 

// Create the final shape using difference

difference() {

    // Outer shape - union of two cylinders

    union() {

        // First outer cylinder at origin

        cylinder(h = height, d = outer_diameter, center = false);

       

        // Second outer cylinder shifted to the right

        translate([shift_distance, 0, 0])

        cylinder(h = height, d = outer_diameter, center = false);

    }

   

    // Inner shape to be subtracted - union of two cylinders

    // Vertically centered and slightly taller

    translate([0, 0, -extra_height/2]) {

        union() {

            // First inner cylinder at origin

            cylinder(h = height + extra_height, d = inner_diameter, center = false);

           

            // Second inner cylinder shifted to the right

            translate([shift_distance, 0, 0])

            cylinder(h = height + extra_height, d = inner_diameter, center = false);

        }

    }

}

 

 

So, now that I had code in place that could generate a cone by stacking cylinders, and code that could indeed create what looked like a cylinder splitting and moving apart - I now wanted to combine both methods. Have start creating the tree structure by stacking the cylinders and reducing the diameter - but, along the way - start splitting. The following shows that in progress, 1^st split at 130 height, 2^nd at 360, total height of 500:

 

Anyway - still working on it. Will keep folks posted as I make more progress!

 

-K

[Kurt The 3D Printer GUY!!]

DARN - I didn't know all the pics would Disappear - or appear as Broken-Links! UGH...

Time to Rectumfy the situation!!!

One might say that this Posting is long overdue - as it does indeed have to do with my travels across the USA as I moved from SoCal to Ohio. But, there was a good reason for the delay...

 

So, as I lifted up stakes & Hit the Road to go Back across the USA again - my 1^st stop was Utah. OF course, a Great reason to stop there was my good buddy Joe - the 3D Printing Professor. It was my 2^nd time to visit him in Utah at the Maker Space in the Library where he worked - to oversee the Maker space. He's ALWAYS a Fun guy to hang out with. This was a shot of the 2 of us hanging out in the Maker Space:

 

So, whilst we were hanging out - and talking all about 3D Printing stuff - I brought up this one project of mine that I keep trying to get back to. A project where I have gone down a NUMBER of paths, in regards to the design of the shape. The 1^st iterations of the design actually started back in like 2011-12, after I built my CupCake 3D Printer. 

 

At one point, I tried a different approach to the design - by simply sculpting it in something like Sculptris - to 3D Sculpt the 3D shape in the computer. I did this one version of the design, and I printed 1^st down in San Diego at SD3D. But, getting the chance to 3D Print it at Essentium - where I could print it REALLY Big - that was COOL and FUN to do. This was that print as it had just completed on the Essentium printer:

 

Anyway - as I was discussing the project with Joe the 3D Printing Professor at his Maker Space job - and he then took at quick shot at doing the design in Blender - as we were throwing ideas around. But - here's the Delema...

 

I have 3 main requirements for this design:

1. All branches of the tree must be hollow - so a wire can be dropped down thru a tree branch tip - and go all the way down to the root of the structure. 
2. It needs to be printable without supports. 
3. It MUST be Very Easy to do design iterations - such that maybe a single design iteration must take like 5-10min each - so that I can produce 500-1000 versions of the design, and even more.

 

So, first Joe jumped on a PC and booted up Blender - and did a little work to see if he could come up with an easy way to do it. But, there were issues with that attempt. Then suddenly he jumped up, wiped clean this whiteboard and started drawing and discussing his design algorithm - which, once I had completely wrapped my mind around what he was trying to do - I realized it was BRILLIANT! This was his whiteboard sketching:

 

I did ask Joe in another thread if he minded if I posted this - but, I didn't hear back - and figured it would be fine...

 

Now, I won't go into the entire algorithm - but, will give a brief rundown of it. In the end, the idea would be to Directly generate GCode. Briefly, one would generate circles, stacked up. But, then - the circles would split and move apart - and thus the branching. Think of it like a human baby growing, and how you have a single cell, then it splits, and each sub-cell then splits again, and so forth. 

 

I first began exploring this project by doing it in FoxPro. And, I made some progress on it. From drawing circles, to drawing 2 circle intersections, but, also make it so that when they overlap - the circles are not drawn in the intersecting area. This is a screenshot of the one program, in VFP, the drawn circle & not drawing it in the overlap area:

 

Admittedly, the progress with FoxPro coding was a bit slow. And, I began to see more complications with it.  And, I was even using AI to assist with doing it.

 

I then decided to try a different route. As in, still using Joe's algorithm concept - but, NOW - doing it via OpenSCAD.  And, to accelerate the whole project - I again turned to me new best buddy - Jean Claude von Ay-Eye.

 

I began doing things in steps. I first wanted to see if I could get Claude to just generate a squat hollow cylinder - based upon parameters like inner diameter, height, wall width - things like that. It did. Then, I figured, let's see if I can have it generate a Cone - but, via these stacked cylinders diminishing in size - as that's basically how I would grow the tree. And - it did:

 

Then, next - I wanted to work on the concept of the circle splitting apart - like in the FoxPro shot above - and I got success:

 

So, now that I had code in place that could generate a cone by stacking cylinders, and code that could indeed create what looked like a cylinder splitting and moving apart - I now wanted to combine both methods. Have start creating the tree structure by stacking the cylinders and reducing the diameter - but, along the way - start splitting. The following shows that in progress, 1^st split at 130 height, 2^nd at 360, total height of 500:

 

Anyway - still working on it. Will keep folks posted as I make more progress!

 -K

[Bill Rockhold]

Welcome to Ohio!

Bill

--
You received this message because you are subscribed to the Google Groups "3D Printing Tips and Tricks" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3d-printing-tips--...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/78851131.13b36.197237e8c7a.Webtop.15%40optimum.net.

[Kurt The 3D Printer GUY!!]

OMG - Another Ohio person - COOL! Nice to meet you Bill!

Are you at least close to Cleveland? I suspect probably not...

-K

[Bill Rockhold]

Nice to meet you. I'm in the Columbus area. 

To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/ea53af53-279e-493e-972f-b246622c7a56n%40googlegroups.com.

[Kurt The 3D Printer GUY!!]

Cool! That's only like 2.5Hrs drive south!!!

Pick a Brewery there and we shall have to meet up soon!

-K

[Kurt The 3D Printer GUY!!]

I am coming back to this project lately - and, this one I am throwing at Anthropic/Claude - Fable 5.

They are still in this pre-release phase - with, I think it's like a 50% token usage - I am sure I am NOT Explaining that right - but - Whatever...

Anyway - I got some great results FINALLY Tonight before leaving work - I then push another prompt thru - for it to do the next phase - as we are kinda implementing this in phases (me & my BFF Claude) - and I think this may FINALLY Really work. I figured - Fable 5 would be a Great option to throw a more complicated problem to the AI!

I will follow-up soon with Results!!!

-K

[david merten]

I am curious why you are trying this? Are you trying to make cool "tree" lights? Or some kind of scenery?

I had an interesting thought but while trying to type it out I realized you have a bigger problem than I first envisioned. 

            David

--

You received this message because you are subscribed to the Google Groups "3D Printing Tips and Tricks" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3d-printing-tips--...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/a9988d04-02c6-4644-bb8b-96eb8187aa31n%40googlegroups.com.

[Kurt]

Yes, the idea is a Table Lamp - structured to look like a Tree. 

I got some nice tests. Initially - I told Claude/Fable 5 - to just create 2 circles, 1 above the other - the smaller is above - and shifted - and then the idea is - connected the vertices together to make the surface. This was that basic test:

Then, I was like, do a bunch of them, gradually getting smaller as one went up in Z and also shift them - so we got a curve kind of branch:

Then - try the concept of a branch splitting - these were 2 more test - for those splits:

So - things were going good - he figured he would now do the whole thing - since we had the basics as functional and looking good. But - DANG - it sucked:

But - just now - I pushed back. Although sadly, it caused me hit my Usage Limit!!! Except - what caused it was that he had generated the code, run it, got a new tree generated - but, then he was doing the Analysis of the tree structure - to confirm it was good - and THAT is what caused me to hit the usage limit. 

Still - it DID create a new full tree design - and BOOM - check it out:

Needles to say - it's NOT a Finished design. Far from it. But, I now have algorithmic code that is Viable and has gotten me much closer to a potential tree design. And - yeah - all branches are hollow - trunk is hollow - and for this tree - I could drop wires down each tree limb tip and they WOULD feed all the way down to the Trunk!!!

Yippee!

-K

[Jody Harris]

Reminds me of my Magnetic Tree (Beer/Bottle Cap/Paperclip Tree):

https://www.printables.com/model/845541-magnetic-tree-beerbottle-cappaperclip-tree

-j

Think carefully

To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/eaedbc66-bcf5-4834-a10e-b10655af29ebn%40googlegroups.com.

[Kurt The 3D Printer GUY!!]

OMG Jody - your reply is Hysterical - as I did go to the link. So - you are Mr. Havoc??? Sounds like IRL you are a Trouble Maker!!!

And - I guess you like Chewing tobacco. I only know One person who uses the stuff - and he's a VERY Long time friend (best friend in 4th grade - his family/Father/Uncle/Brothers all worked with my Dad who did Arch work for their various offices and lumber production facilities - and, ironically - my buddy is also my sister's Ex-husband! - I know - a weird combo of life!) - and to this day he still does the chewing tobacco. 

Yeah - your Tree design does look like mine. But, of course - yours is a ONE Off design. For mine - I need an Algorithm - that can help me produce Hundreds to a Thousand similar tree type designs. Cause - YOU Know - that is what 3D Printing is ALL about!

How did you make your tree? IT does look Sculpted!!!

Yeah - I got a chuckle out of the Printable page and all the things stuck to the tree!

FYI - the VERY Early model that I printed - back in 2019 - a buddy of mine suggested it would be a Great way for women to hang jewelry - necklaces and such. Even my boss at SD3D - was considering pushing it for that - and making Custom ones for various Jewelry stores! Heck - maybe I should return to that concept as a potential viable money maker! Cause - you know - we ALL GOTTA MAKE MONEY!!!!!!!!!!!

-K

[Jody Harris]

DEFINITELY  one-off! I drew it in OnShape

https://cad.onshape.com/documents/682ec38c1f2245966205389f/w/8d2d75ea731e681466796637/e/a5fa9f735224307a8d2e65aa?renderMode=0&uiState=6a5175daace52b2aefd40f2f

I fought with it for HOURS. So many user defined planes! Just for that little tree, there are 14 user defined planes to get the shape. 

Never was happy with it. Finally I just posted it in hopes that it would stop nagging at my brain.... Apparently that worked! I've barely thought about it since.

The dip can lids belong to a coworker. Several of my coworkers dip and she was collecting them at the time.

-j

Think carefully

To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/28501499-b5ed-4ad1-a5ed-61db8573cfa1n%40googlegroups.com.

[Kurt The 3D Printer GUY!!]

Funny Stuff there Jody! Heck - the main tree part - you could model that in like 15 min with Sculptris! It's STUPID Easy to use!!!

I actually taught the program to young kids in the past - young Teens - they learned in like 15 min - and within 30 min from Start - one young gal had created a kind of bunny! So - it should be Easy Peasy for you!!!!!!!!

IF you want - I will send you the program - it's kinda small EXE wise - and does not even really require a kind of Install. 

Yeah - VERY Funny to hear the woman collecting the Dip cans. Didn't know it was referred to as Dipping. Was SHE also Dipping???

In the end - I will TRY to keep folks posted on this project...

I really gotta do posts of Several other projects I am scrambling on. One is a Repair via a 3D Print - and it's a rather FUN REPAIR! An Item I am about to USE - going across the USA - in about a week with it...

-K

[Kurt A 3d]

Kurt-1

If I were to look into trees I would try to wrap my head around this paper.  There are some cool historical formulas that have to do with fractals and the actual nature of trees in nature.

https://lsa.umich.edu/content/dam/math-assets/math-document/reu-documents/ugradreu/2020/Gao,Jingyi.pdf

Kurt-A

[Kurt The 3D Printer GUY!!]

OMG Bro!!!!!!!!!!!!!

Funny you should say that...

Years ago when I was in College - as a CS Major - and back then - there was NO Such thing as a CG Major or Minor. So - I Made it up - and just did it! And, part of that - I did CG to do a Fractal Tree design - but, purely 2D! The Algorithm was SIMPLE!!! And - the basic tree design I produced - for an assignment - was GREAT! I could NOT Believe that fractals and Simple math could do it - but, Yes It Did!

I WILL indeed take a look at that paper - and potentially use it in my Algorithm!!!!!!!

Thanks Mr. Doppleganger! 

-K

[Kurt The 3D Printer GUY!!]

So - I'm looking at the document - and I see this one section - and those section of images look VERY Much like the tree version I did:

The algorithm I did way back then - I remember it was Recursion based. Which is a very powerful programming technique. 

I may feed this Doc to Claude and see if he can implement some of the concepts for my tree design. Thanks again for pointing me to this doc!

-K

On Friday, July 10, 2026 at 7:44:54 PM UTC-4 Kurt A 3d wrote:

[Joseph Larson]

Kinda cool to see how far you've been going with this strange little idea has gone.

[Kurt The 3D Printer GUY!!]

Thanks SO MUCH Joe!!!!

Yeah - I gotta admit - the last round, with the AI - getting some python coding that actually is Workable and is creating something TRULY Along the lines of what I seek - I was SUPER EXCITED when it worked!

Of course, Even I have to admit - I still have a ways to go! Even if I create something with a tree structure that's Printable - with proper hollow pathways for the wires - I Still have other ideas to make it look good AND be Functional. Like - have the AI create textures on the tree - like bark. And, the idea of the LED's on the tree limb tips. I need to produce a another structure that is like the holder for each LED - that has the split for the 2 wires - properly holds the LED - and lastly - like a bulb around the LED. Might almost look like a kind of leaf type structure - main holder - then the bulb covering - might almost be like a fruit shape. 

AT that point - ALL my efforts have been in regard to getting the main structure first. 

I will indeed try to keep my progress posted here!

-K

[david merten]

With 4 wires you can use rgb leds and have so many colors. 

The possibilitys are endless for the lights. From leaves with the leds embedded in them to flowers, like in the little shop of horrors. I can imagine all sorts of ideas...

No please my head is already full of ideas......lol

          David

To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/99a0c249-faf5-40cd-8bf4-a6c5849ac978n%40googlegroups.com.

[Kurt Gluck]

Or you could just thread some fiber optics and have a couple of leds in the base 

Kurt

On Jul 15, 2026, at 8:03 PM, david merten <davidme...@gmail.com> wrote:



You received this message because you are subscribed to a topic in the Google Groups "3D Printing Tips and Tricks" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/3d-printing-tips--tricks/X9RbESuGHCY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to 3d-printing-tips--...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/CAPZzpqvwe0xZp_zT_7ybdpD6Z60X2puTaCHs3WgZLbqmXQCTfw%40mail.gmail.com.

[Kurt The 3D Printer GUY!!]

People have said fiber optics before - You're Def not the First Mr. Doppelganger 

But - AFAIK - the tips of those optics fiber will NOT give off enough light for it to function as a TRUE Table lamp. Right??

-K