Six months ago I posted about fixing up and finishing some Gigaminx parts that were printed with a messed up 3D printer. I have finally used them to build a puzzle, and thought I would share some new ideas I tried.

I started from a single set of masters, and damaged ones at that. With a lot of work, I fixed them up, but I still only had one of each piece.

On a Gigaminx you need 60 copies each of three parts of the smaller, 30 edges 20 corners and 12 centers. Far to many to cast one at a time. Most people print multiple of the smaller parts and make molds containing many. A good setup is:
Four copies of the small parts (15 casts), two of the edges/corners (15 and 10 casts required) and one of the centers. So 15 casts later you should have a Gigaminx.

Starting from a single piece each, I didn't want a ton of wasted silicone rubber and a bunch of mold halves to flip over and confuse, so I decided to try pouring a single mold from the same piece, but multiple times.

I started by making three molds: The three small pieces in a row, the edge and corner together, and the center alone.

Next I wanted to make molds of the small pieces and centers next to and attached to the one I just poured. Here is how I did it:

1.) Take the bottom half of the mold and build a low box twice its size, so you have room right next to it.
2.) Spray your mold release at this point, you won't be able to later.
3.) Pile up a bunch of stuff to hold the mold top at a level where the edges just touch, but it won't let silicone drip through. I used cut up old business cards and Elemental Fluorine pieces
4.) Build the box up higher and pour a second top half of the mold, right next to the first. Since you don't have mold release on the second half, the new rubber will bond to it.

5.) Wait for it to cure, and you have a solid double top

6.) Flip over. You now have a solid double base, and a single top. Remove the top (and the pieces) and spray mold release again.


7.) Replace the pieces (careful not to touch the rubber sprayed with mold release) and pour again to make the second half of the top.

8.) Now you have a single block mold of twice the masters you actually have.

For the three-small-piece mold I took slightly different steps as I didn't want to end up with a double mold and no easy way to make a third and fourth. So after pouring the double top, I poured a single top again, and used that as the basis for another doubled mold.

In the end, I had exactly the molds I wanted but only did the clay work once on a single set of masters:


What to do now? If only I had... this...:


Next it was time to cast:

The pieces came out pretty well:


One note: When you make your alignment pegs, make them asymmetric so it is really obvious when you have them backwards. I didn't, and a few times I mashed down on the mold and thought I had waited too long on the first half. It happens instead I had the mold top backwards, so I was mashing together a mis-fit. The parts came out half center edge, half center corner. Frankenpiece!:


With *many* hours of cleanup the puzzle started to come together. This is what I had when I went to my first day of IPP:


You can see that as much time as I spent on the masters, I could have spent more. Next time...

By California Cube Evening I was about 2/3 done.
Random picture: My messy, messy desk:

I finished a day later, and it looked nice:

One and a half hours of stickering later it looked better, but you can see those in the New Puzzles thread here.

In the end the puzzle moves OK, but I didn't round the edges nearly enough, so it catches far too much. Fixable, but it would take a lot of work. My instinct is to round the masters and pour molds again, and cast again, but that is a lot of work. Maybe later.

Moral of the story?
1.) With a bit of time you can build multi-piece molds from single masters. This may make PolyJet a reasonable option where with multiple masters it may have been too expensive. It also means you can spend that much more time on the single master getting it perfect as you don't have to replicate the finishing so many times.

2.) Only clay once. Use half of your mold repeatedly as the new clay base to pour other second halves. If half of your mold tears, don't start from scratch, just pour another second half.

3.) Yes, I could have just junked those pieces and paid a bit more for multiple copies of masters. But I was in a unique situation, with a set of pieces that fell into my hands. I wanted to do what I could with them.

Enjoy,

Dave

Great job Dave! Can we expect to see a video?

I'm hungover so I had to read it a few times to figure out what you did! Very clever!

BTW - Are you hanging out in your bathroom pouring pieces? Do you stand up while you're pouring or use the built-in seat?

Perhaps on a future project. The problem is that when I am working with OOMOO or resin things get messy and I don't want to be touching my camera.
How did you ever finish your Petaminx if you had to keep stopping for bathroom breaks?

Dave

Nice work, very creative solution. Just out of curiosity, how would the time/cost compare for doing it this way rather than just printing more masters, in terms of costs for printing, oomoo, and time to polish all the pieces?

Using wax paper to cover a work surface...thats a great idea!

That is hard to evaluate. It depends on how you value your time, how much of it you have and what amount of work you intend to put in to each piece. Here are a few factors:

1.) This method favors someone who spends a lot of time preparing a master. If you have to do this on multiple copies of pieces, that is a big time investment multiplier.

2.) This method favors someone who doesn't have a lot of puzzle building time, but plenty of time in between short periods of work. This is me. I get an hour or so here a few days of the week so waiting for a mold to cure is not lost time for me, by the next time I can work it will be ready.
For someone like Drewseph waiting to pour four or more molds sequentially is probably a big waste of a week, he is better off pouring a few multi-piece molds and get casting.

3.) Obviously printing is cheaper, but OOMOO a bit more expensive since the molds produced are not as optimal as if you have all the pieces packed in. But the difference is very small. These molds probably required less than a $30 trial size of OOMOO, so even if twice as efficient (a generous estimate) you would only have saved perhaps $15 on silicone. Printing multiple parts can easily eclipse that, even for small FDM parts. For something like PolyJet it can be much more of a savings.

Dave

P.S. Garrett: It's parchment paper, used for baking, which is paper infused with silicone. Resin pops right off, but OMMOO sticks a tiny (scrapes off easily with a fingernail). I imagine wax paper might get messy because the wax would start to flake off, but you might try it.

We didn't stop. Frank and I used a bucket that was under his table!
I agree 100%. I learned the hard way! It's worth the time to get your masters looking as good as possible!

Here is a decent picture from my iPhone (who would have figured good quality from a phone?) of the internals of my Gigaminx.

Consider that the masters looked as rough as the core when I started...


Click here for the large size to see detail.

Dave

That looks really good you cleaned it up well

(that A LOT better than my camera... maybe i shouldnt have put it in water)

HOLY [You know, I would be censored anyway. Why do I exert the effort to do it myself?] , Dave!!! That's amazing! It looks production quality. How long did it take for you to clean the masters up, by the way?

-Jon

See the bottom half of this post.

Dave

Not necessary, but at a certain point when you are trying to time a bunch of first half pours, turning over a bunch of mold halves instead of just one or two can be a bother. Also with a bunch of "identical" mold halves you might get confused as to which matches which. I suppose they should all fit the other, but I think it perhaps better to keep them in sync with what was poured over them.
Well yes, using one half of the mold as the "clay" for multiple pours. Useful, but I wanted to simulate a mold from multiple copies, not just have copies of a mold.
Unfortunately as much time as a spent on my masters you can still see the printing artifacts. Part of the problem is that it is very hard to see these tiny depressions in the white and black of the piece. The other is that milliput is a bit softer than the ABS master, so when you sand it eats the milliput easier, making those parts a tiny bit deeper.

The result is not perfect, but far better than most other methods I think. If I had the patience (and I might, for my next project) I would sand lighter to try to compensate for this, and perhaps paint my masters with a thin coat of primer then sand from there. The primer should fill those tiny gaps and, once sanded down, be a better surface

Dave

I ran out of parchment paper and tried wax paper. It works, but is too soft and flexible, every time I move a mold it grabs the paper with it. Also, after two pours it started tearing.

So it works, but not nearly as well as parchment paper, which lasted dozens of pours before I had even a tiny tear.

Dave