I said I would do it, and here it is--in it's own thread no less! I have created a workable model for the junior barrel, and am posting it here for anyone who wants it. The files come with no guarantees whatsover, but in theory the design should work. Tolerances have been set in such a way that all the moving parts should fit nicely, while the end caps have an interference fit for a nice snug finish that doesn't require glue.

As well, the hole for the fastener has been designed around the idea of using a single M3 fastener (screw+2 nuts+washers) with the length as appropriate to the spring chosen.

So, that's about it. The pictures below illustrate how the overall design fits together, the .igs files are there if you want to look at them, and the .stl files are there to 3-D print (you'll need to rename the files with a .rar extention prior to unrolling ***INSTRUCTIONS FURTHER DOWN THE THREAD***). I will not, however, be posting the original part files for reasons that I am not willing to discuss.

Enjoy!

*And if somebody just happens to make a couple of these, I think it would be more than fair to offer me one at a very reasonable price--this design didn't create itself!

Pictures here--ran out of room in the original post.

Cheers!

Thanks! I can't wail to have it printed.

Can you tell us the dimensions please?

Oops! That would be a useful bit of information I suppose...

Height = OD =57mm. That way this thing "matches" a standard 3x3x3 cube in some sense.

*sigh*
Well, I guess now is as good a time as any to post my latest puzzle. Check the new threads in this forum.

the.drizzle, thank you very much for open sourcing this design. Mabe Scott will do the same?

Anyhow, this is just perfect for the purpose I've been thinking about. I'm much more interested in the "why" of the design. These two designs are different in approaching the same goal. Which is better? Or, what are the strong/weak sides of one or another? Why it's been done this or that way and not another? And so on.

It would be great if you guys could elaborate on that. Then we'll learn much more than just be looking at the files. Thanks!

Drizzle, what program do you use? Also, you mentioned changing the filetype and whatnot, can you briefly explain how to do that?

Great puzzle the.drizzle! Interesting design. Now someone needs to 3D Print it.

Sorry about the dodgy filetype, but I discovered yesterday that the board has a 2mb attachment limit, hence breaking up the file. So, to unroll the lot, rename all the files with a .txt extention with a .rar extention, for example

junior_stl.part1.txt --> junior_stl.part1.rar
etc...

At this point, I'm not 100% sure how to help you if you're on windows, as I only know how to use linux. However, I'm told that what you use is something like winrar to "unrar" the file junior_stl.part1.rar, and that will pull in the remaining 5 other rar files and produce a single .zip file that can then be unpacked in the usual manner. I tested this just now on my system and it works fine:



Also Aleksey, I should make clear that I have not open-sourced this design, because I have made the explicit statement that I will not be giving out the actual source files. So, it would be more accurate to say that I have made this a free design.

As for the logic of why I did it this way vs. another one, I wish I could say that I had some sort of insightful reason, but the honest truth is that it just seemed obvious to me to do it like this, and I just sort of put down what I saw in my head when I first saw the puzzle.

Cheers!

Yes, I realized that later. It's a shame though. I'll try to explain why. When learning something, it's best to learn from actual ready designs of puzzles that are created by someone who knows what he's doing. Also, when you're talking about tolerances and "a nice snug", I never know how exactly that's done, actually, to incorporate that piece of knowledge into something else. It would be just great to see an actual design open sourced once... One that can be played with to see how this or that feature is actually done.

Thanks for posting the design in any case. I really appreciate that.

As for the "why", I'd still love to see the comments from you and Scott what are the better or worse spots of each design. I am sure you guys do have some comments.

Well that I can do. Clearly, my design is superior on every level! (So there is no confusion, that's me being sarcastic)

But seriously, my basic logic is when making up a puzzle that is going to be 3-D printed, one might as well go with spherical geometry for the mechanism as opposed to cylindrical--in theory you will always get better movement out of it. I suspect the reason its used so little in mass produced puzzles is the complexity of the tooling. With that idea in mind, this puzzle is simply three spheres intersecting a circular cylinder, and the mechanism makes itself--no thinking involved!

As for the tolerances, its not a quick and easy answer--you can take whole engineering courses on tolerancing! However, the basic idea is to look at which pieces will be in contact and where, and figure out what spots are in contact because they are controlling motion, vs. areas that are simply rubbing together. You then want to add / subtract some material from the parts that are just rubbing, so that in a perfect world they are actually no longer making contact. The tricky part, though, is than in doing this, the "shrunk" region may be the spot that is the determining region during another movement of the puzzle! Add to that the fact that the tolerances add up with piece counts / size, and it can turn into a bit of a pain.

For this puzzle, though, there are so few bits that it was no big deal

I'll take a stab at this. I think the.drizzles designs are fantastic though, let me just get that out in the open.

My Design
Pros:
More tolerance (possibly). My logic here is that are mechanisms are essentially identical except that my vertical turning has an extra tolerance because of the spring. Springs flex more then plastic.

Easier molding. My pieces look like they would be easier to mold with a rubber.

Cons:
More expensive to print. My design has two different center caps, and the pieces I'm printing aren't "hollow". Meaning more plastic, and more $ required to make it.

the.drizzle's Design
Pros:
Cheaper to print. Hollow pieces, fewer pieces.

Lighter puzzle. Because a lot of this puzzle is hollow, it should weigh a fair amount less. Mine on the other hand is going to be a little heavier.

Cons:
Less tolerance (possibly). My logic here, is his vertical turning. Each piece turning vertically is held in some part by the core. Which is fine, but the core of this puzzle doesn't flex at all. Meaning the only part of the puzzle that will be somewhat tolerant will be the top and bottom pieces, which is great for horizontal turning, but not so much for vertical turning.

Harder molding. The core and top centers look tricky to mold / cast.



Once again, I think the.drizzle's designs look great and I cant wait to see both our puzzles in action.

Thanks, Scott. I noticed the tolerance issue as well. the.drizzle's design is like Eastsheen cubes. Close tolerances and no real need for spring adjustments. While yours is like a Rubik's cube. It seems to me the the.drizzle's design is more difficult to make unless you have access to means of making the parts with such precision. It seems like it can be done with machining parts, but less so with printing. I don't really know.

just wondering what program did you use do design these pieces?

That is the idea exactly--well-fitting parts + spherical geometry=no need for springs with large amounts of play. Take a look at a skewb mechanism; sure, there are some springs, but they don't really do anything in terms of part alignment.

As for it not being possible to make this item precisely enough, I don't think that should be a problem. Of course, I could be wrong too. I guess it depends on the resolution of the printer you're planning to use. However, even if the printed parts were to be a bit "rough", I can't imagine that the problem couldn't be fixed with a bit of sandpaper and effort. Until one gets built, though, it's really just a matter of academics

As for the design software, I used SolidWorks for this particular puzzle.

I don`t really understand how you open the .STL files... can anyone help me please

The directions I gave above are pretty clear--what part are you having problems with?

I change every .txt for .rar , then I try to open them with WinRAR and it doesn`t work ... I have windows XP

Based on the design by 'The. Drizzle' for the Junior Barrel, I have made two samples of this puzzle.
The pieces were made by the SLS rapid prototyping technique and were tumbled to improve the surface smoothness.
Next steps are to colour the pieces (black) + to arrange the stickers.

The design is excellent and the pieces move great.
Below are some pictures.

Geert

Impressive!

wow !! Geert.. fantastic job !!!

one for me please...

Has anyone contacted Uwe about these? I think he would be glad to hear his puzzle has been materialized!

It appears to look exactly like it did on my screen when I modelled it--3D printing is SO cool!

I guess I should create a sticker template to go with it too; the puzzle is pretty trivial without them

Cheers!

Yeah, some thread bumpage--apologies in advance, but I think it's OK here.

I thought I would mention that Geert was kind enough to print me a few copies of this puzzle. It took only a few minutes of work to get the motion to a level that I would rate as very good / excellent, and now that I have sealed all the nylon off it should never swell again. That, and the stickers adhere quite well to the clearcoat sealant, unlike the nylon, which is immune to stickers it would seem...

Anyway, one junior barrel. The cat is not impressed, but what does she know.

Out of curiosity, what do you use as sealant?


After you and Drewseph I think everyone with a cat should take a cat picture of their new puzzles.

fixed