Inspired by Oskar`s 17x17x17 Pagoda-style Design, I decide to introduce my version of "non standart" order for pagoda-style, 11x11x11 cube (there you can see why this order is not clearly binary)! So it was more difficult to implement such order design, but I didn`t stop, because the reason of desire to keep in my hands exact the same size cube! There you can see my first attempt to do the one.

My current designing progress is almost complete - you can see rendered images below.
Now implemented:

• total of 21 unique pieces;
• complete layers slice cuts;
• pin modifications for avoid poping;
• minimum bends and twists - for perfect, simplest shape and less friction during layers rotation;
• huge as possible fundamental actual corner and edge pieces, and core legs - for cut down more unnecessary heavy material;
• comfortable pillowing of cube;
• tolerances/clearances;
• fillets of all edges for smooth moving.

Several things have to be just improved:

• layers alignment after rotation - it will be done not by hidden layers and not by magnets;
• holes, niches - for cut down on the amount of material, and decrease total weight of cube; caps also;
• Screws and springs placement - without tapering of middle layer (I have some curious idea, but not sure).

Some interesting facts of the design:

• 602 visible elements;
• 637 separate pieces, which are cube consist of;
• 6 screws, 6 springs, 18 gaskets additional;
• minimal friction (of pagoda-style approach) of mathematical point of view improved in my version;
• the strongest clinging pieces each other - the pagoda-style main merit;

Many thanks to Oskar and his topic for give me a lot of thinking about.
Thanks to Drewseph and his proposition of modelling way tutorial also.
Tony Fisher, Leslie Le, David Calzone as well, for incredible super hi-order puzzles.

I invite all to discuss my project. One principal thing worry to me: there are two possible variants of 11x11x11 in pagoda-style design:


Which one is best for stable cube and avoid pops?
There you can acquaint of what does the pictures mean: left - is the core, right - corner piece, between - middle smaller pieces. I used 1st variant to combine the middle elements. The result you can see on the next images.

There is a gallery of my 11x11x11 Pagoda-style design:


#1 click to zoom in actual resolution 1224x1080px (260Kb)

---


#2 click to zoom in actual resolution 1920x1080px (220Kb)


#3 click to zoom in actual resolution 1920x1080px (286Kb)


#4 click to zoom in actual resolution 1920x1080px (283Kb)


#5 click to zoom in actual resolution 1920x1080px (222Kb)


#6 click to zoom in actual resolution 1920x1080px (207Kb)

---
Layers:


Middle layer.
#7 click to zoom in actual resolution 1080x1080px (161Kb)


2nd layer.
#8 click to zoom in actual resolution 1080x1080px (166Kb)


3rd layer.
#9 click to zoom in actual resolution 1080x1080px (192Kb)


4th layer.
#10 click to zoom in actual resolution 1080x1080px (173Kb)


5th layer.
#11 click to zoom in actual resolution 1080x1080px (168Kb)


outer layer.
#12 click to zoom in actual resolution 1080x1080px (207Kb)

---
As I complete my designing up to prepared for producible model - I`ll share STL-files for 3D-printing here for all.
Also I can give source files to such persons, who really can help me.

---
Further I`ll update the topic...

---
Thanks for interested of the topic!

WOW
I wonder when i can actually see an actual version of it. GREAT JOB!!!

This is a beautifully rendered design. I believe your Pagoda-style would also work perfectly for a 13-cube without the dilemma of deciding between two non-compatible layouts. Personally, I have little desire to go higher than a 7x7x7 or Gigaminx (due to tediousness of higher orders), but I love the intricacies in the way the pieces dance together with these new higher-order puzzles.

For reasons unbeknown to us as of yet, the V-Cube mech hits a roadblock at 11. Only Verdes knows why, but he seemed reasonably certain it wouldn't work at higher orders. But the advantage the Verdes design has over the Pagoda and Over-the-Top styles is simplicity. It seems like every higher-order puzzle above 7 that we have seen contains multilayer concentric spheres. Problem as I see it, these layers don't allow so much tolerance errors as the V-Mechs and other designs, which use spring tension to hold the pieces flush together in place.

I wish you, Oskar, Tony, and Leslie the best of luck on your designs

While this design is quite beautiful I have severe doubts in the pagoda mechanism, and unless this puzzle is giant, I'm not too happy with how thin some of the parts are, if you were casting this I'd say the design will fail due to snapping parts. If you are printing this, be sure to shell out the pieces to be hollow before printing.

also I'm not so sure how well the inner center pieces will hold in, however good luck if you plan to pursue this puzzle and pray that you put tolerances in the right places

the 13x13x13 will be produced by China to my knowledge. (China has already the sample of the factory made 11x11x11)

I'd love to see this produced. The pictures of the design look awesome.

Alex

wait to see the first sample.
very beautiful disign.

Congrats, Tesseract750! Appreciated.

In any case, I'd like to suggest "hold on". Theoretically beautiful though, concentric spheric umbrella can cause somewhat unexpected jam effect which is not comparable with planar/cylindrical concentric circular bands. That is why i don't think Oskar's first 17th design will ever work.

Below is an image showing a possible structure. Hope it helps.

Leslie

Thanks for grade my work!

I`d update the gallery... Look again on the top!

---
Now my answers:

I know why, as well)! I`ll show you some presentation what`s happen in real produced puzzle by V-mechanism if order-number is too big... Please wait until I`ll done it in not far next posts, I need some time to figure the one.
So, Stardust4ever and Leslie Le, - that is the reason why conical layouts (I don`t know about cylindrical actually) are not compatible for super high-order puzzles. I repeat: super high-order cubes! My opinion that: cylindrical structure - applicable for sizes 3x3, 4x4, 5x5; conical structure - applicable best for sizes 5x5, 6x6, 7x7, 8x8, 9x9, 10x10, 11x11; but for higher cases - applicable be best spherical structure! ...I`ll show you later, why conical/linear are worse than spherical/binary styles for super-high.
---

You`re wrong: the elements are size like V-7th elements. During designing I`d compare with my real disassembled V-7 cube always. Thicknesses are the same, and more strong in some cases really. But I`ll take into account your remark.
One more my cogitation: the pagoda/spherical layout high-order puzzles are not for speedcubing at all. Because of,.. you know... )). So speedy rotations with cut corners possibility are not necessary. In this area - different things are key cases, like: strong and stable construction, comfortable keeping in hands and suitable weight, not tiny pieces and not huge total size of cube, embarrassless rotations, etc. - do not think about in policy of speedcubing.
---

Oh, eah, about tolerances. I placed it by my estimation, and You proposed. But I`d like to hear TP members opinion also: what will be better?.. But be carefull: pagoda style is something different, so look attentively on the images of layers, i`d posted on the topic, - and tell your propositions.
---

No. Thanks a lot, but one more "no". You proposed a half conical/linear style again. The pops - one of the main reason to avoid linear style disadvantage: if parts are close to each other in "in one direction" - it is simple to withdraw. There will be a cleft between two pieces. See below, why "no":

But I`ll thinking about, thank You for third variant ).
---

Believe me - I`d care about jam problem! It is quite difficult to undestand - but I have a solution. When the design will be complete - I`ll tell what exactly I meant.
About three or four interesting improvements I have in my mind, I`d never noticed here, for flexible and smoothly layers rotation - to all have its own time to tell!

---

to be continued... )

Flexible layers?
Like the centerpiece of the floppy cube?

))) no layers flexible - but moving of them. Just little improvement of pieces construction for avoid jams during rotations (Leslie Le wrote).

My thoughts are
#1: WOW
#2: THis puzzle seems like it would have some problems, the main one being that uness the pieces are PERFECT this puzzle wont even turn half decent
#3: I want to one day make this puzzle

One more I have to say:
I believe in Oskar`s Pagoda-design. It is no choice from worse and worst, binary vs linear styles, or any else. I suppose it`s just only possible (on current moment invented) layout for orders over 11th. Reason is: solid form of puzzle, but not uncovered, fatiscent, disintegrate housing... Of cource friction between layers during rotations (just think: 602 parts for 11th, or 1538 - for 17th!) would be great. So, we are considering the problem to decrease this effect.

I am who preffer "opensource" like it`s calling in programming area. That`s why I`m here to discuss decisions about the project. On my complete, I`ll post here my results for everyone can take them. To share my current process progress - and listen to your opinions, other offers by looking on a problem from your point of view.

Yup, that is a problem...to be considered.
actually the main reason that i propose this sketch was that you put a large sphere on another sphere. my instinct told me that this could be unstable -- a slight off in coaxial precision will lead to jams unless special recovery mechanism was designed or very special surface treatment of face intact was introduced. And of course, find a way to keep coaxial precision will also make it.

As much as i appreciate open source projects, it is a fact that i'm currently, at least on this topic, reserving contributions to some future date.

Keep going with your project!

Leslie

Wow. You ARE serious. Y really should thicken the parts' legs 'cause casted parts are more fragile than mass-produced ones. I suggest to go at least 4mm for casting.

I know you believe in his mech, but unfortunately the reality is that its unstable. If you cast, all of those corners will snap. The feet of all the rest of the parts will break on assembly. I'll bet money on it. and the weight of the puzzle will ensure that turning will cause hefty tension on the rest of the fragile parts.

as you turn it, no matter how slow, the corner necks would be rubbing onto the edges of the second layer,and act like a butter knife slowly sawing away at the neck, until it snaps in half.

One thing to consider with part thicknesses is that the strength and smoothness of the plastic is important. If one were to cast a V-7 in Alumilite resin I don't think it would be nearly as strong or move as well. The plastic Verdes chose for his puzzle is an important part of its success.

So basing your design on the dimesions of the Verdes parts is probably fine if you intend to have it built with that plastic, but could prove a problem if you prototype with the methods used by most of us on the forum.

I don't have much information on the strength of SLS compared to the Verdes ABS parts, but I can imagine it isn't as strong, and I know it is certainly not nearly as smooth.

I wonder if any puzzle of this complexity has a chance of working well if printed vs. injection molded. I know puzzles of this complexity certainly push the limits of traditional silicone rubber casting of two part epoxies.

Dave

Ok. I respect You guys, so I`ll thinking about You said.

As I understang we have two main reasons for criticism: the first one is amount of friction during layers rotation; second one - is strength of material produce the puzzle.

For a first task - I have a solution in my mind almost. You`ll see it further, I`ll describe it later. But for a second problem - I realy don`t know what to say... It is nearly impossible to increase a thickness of the parts legs - there is no space for this! 1/8 of circle around in a slice by a layer - only a placement...

If you can't thicken the parts, then you'll have to print the whole thing.

Wow good job tesseract!
it can be better than the V MECH but i'm not really sure
I like to see your 11x11 if you have make it.

I suggest a half printing. Print all of the really small parts that are to fragile to cast. Also, I have cast 2.5mm feet for my 1x2x4 in alumilte, I think a 2mm foot would be the thinest if you were to cast in smooth-on. I have cast v-7 pieces with smooth-on and they are a little brittle and thin. I consistantly get an air bubble in the "wings" that the center pieces and + centers that usually destroys "wing". However, on the edge pieces, the wings come out perfectly. I think it was just the way I made my molds. I hope this helps, Tanner Frisby.

Look at your corner, it has so many contact surfaces. As a result, the friction will cause the corner part gradually wear off and be loose and even might fall of.

You both are talking about single entity. Corner piece length is just about SQRT(3) larger then middle-center piece length. So contact surface are not larger this value.
There is the one of the main principle of Pagoda-style - to divide loading of all middle elements: on middle-center piece and on outer pieces (but not on middle-center piece only like in linear style). So that`s why outer layer looks such huge - it keeps a half of total cube weight. It gives to a Cube strong and solid properties.
The neck/arm of corner element has about 5mm width - I think this is enough to avoid fail like "butter knife" effect, as Drew wrote. So I want to ask You: why in V-mech such effect is not observing, where the similar arm is thinner and looks more fragile?..

So I`ll intend to do exact the same.

Because of the expensive injection molded ABS plastic. It is stronger than printed parts, and it is so smooth compared to the other resins and very polished. With a bit of oil (or silicone, but I think people have better results with oil on this plastic) you have almost no friction at all.
Compare this to a print (of any of the rapid prototyping processes) and the surface is terribly rough. Only the DLP Jason posted about seems even close. Even with friction sanding off the big edges over time, the material just won't ever be as smooth as the slick skin of the ABS.
SLS is a big resolution improvement over FDM, but no where near smooth.

Drewseph can confirm, but I don't think any V-mech inspired designs are as small as the actual V-Cube part dimensions. I'm guessing the mini-gigaminx is the closest though.

Dave

My absolute minimums for any puzzle follow this one rule:

1) Make a cross section of the thinnest part.
2) Make a circle perpendicular to the length of that "neck"
3) So long as you can fit a 4mm diameter circle within the walls of that neck, you're part should be safe.

That is a good rule to follow. But I have found a way of strengthening really thin stems. If you cut pieces of steel T-pins, and lay them in the mold each time you cast a piece, it makes the pieces much much stronger. Last fall I tried making a gigaminx. The final puzzle never worked, but those T-pins made the corners much stronger! I recommend using T-pins on anything smaller than what Drew described.

-π (Eitan)

There is a rough demonstration of that happens in reality with V-design for very big cube`s orders. I show an example of V-13 order (in my own opinion) why mr. Verdes decide to stop on V-11. You can see on images below, that such design has its limits on practice. The cube opens to all sides like fir-tree, or some flower (look at second image). The cause - the radial tolerances. By the way, you can observe similar effect in a video by Tony Fisher`s 11x11, especially in space marked red line on a first image.

Pagoda-design, in opposite, has very and very strong structure, without such artefacts. And so stable as a brick. That`s why it has a good perspective for cube orders over 11x11x11.


V-13 break 1.
click to zoom in actual resolution 1024x800px (128Kb)


V-13 break 2.
click to zoom in actual resolution 1024x800px (117Kb)

what program do you use?

also is that a simulation of the puzzle sitting under gravity?

I haven to disagree that such a huge gap would form, however I do think that the outer inner corners of the faces will lift on their own and slide out. very precise "bloating" of pieces wold be needed to counter this effect. and keep tensions tights

2 Drewseph:
it`s SolidWorks only )
without any simulation - manual modelling only;
of cource the gaps are exaggerated - it`s a "rough demonstration" - for preview only.
---
To avoid a indexer of misunderstandings - the last illustration are of V-mechanism design only. For Pagoda design such things never happen.

I think this design is worth a try...