Hi. I was thinking of making an 18 piece burr but am unsure of what wood to use and what methods I should use to cut notches in the wood. So far, I have considered using either use a band saw or a table saw (with a jig) to cut the notches that have the proper width and depth.

Any suggestions to getting started with burr building?

Ok... since no reply have been to this, here is my take:

Andreas Rover is a very lovely person, and he made freely
available a magnificent program for Burrs, which is called Burr tools:
http://burrtools.sourceforge.net/

This program can even calculate (automatically!) the steps
needed to solve a puzzle, also ensuring you do not create an...
impossible object.

After making the 3D model, you can print it and play with it.

Bram and Bob, who have a lot of experience in those puzzles,
both use it!

Now, if you want to cut it in wood, you will need some expensive tools
to get the required accuracy, as well as more expensive techniques
to make the puzzle "look good". This is why many designers trust the
hands of some wooden puzzle making experts, such as Brian Young
(Mr Puzzle Australia) and Bernhard Schweitzer (Germany).




Pantazis


PS. I also have a dilemma... as this is the first such case...
Should this post stay here, or should I shift it to the Non-Twisty
puzzle forum? (I know this is the building section, but it is the
building section for twisty puzzles).

Don't forget Bob's wonderful 'eight' puzzle! (I got it! I got it! I got it!) That was made using Burr tools as well.

I've downloaded this program, read all the information and really need a one on one to understand it

Not only are man tools a problem for me, but also computer tools I'm hopeless.

I vote for off topic with a name change.

And don't forget
http://www.research.ibm.com/BurrPuzzles/

Rox, Burrs are non-twisty, and this post should be placed in the non-twisty section.
The building section is only for twisty. I just pretended I was in a dilemma to be
diplomatic and ensure that there were no disagreements LOL

I believe the contents of this post are very useful for people starting up with Burrs.

But... placing it at the Off-topic forum? Name change? What else, enrol it to a witness
protection program with a completely new ID so that no one will be able to ever find it?




Pantazis

Puzzle protection program! I love it.

I meant something like Burrs-help finding soft ware for design.

By the way, I got some BurrTools file from an Australian friend (hint: Gold Coast LOL)
for a complex puzzle design, and whose moves can be viewed in a superb animation
within this program.

It just shows the huge potential of this program! It has never been so easy to create a file,
check if it is solvable, and then 3D print it!




Pantazis


PS. maarten, wonderful link and very helpful too!

Burrtools is a great program - you can find it here:
http://burrtools.sourceforge.net/

Jurg's Burr applet
http://www.research.ibm.com/BurrPuzzles/index.html
is really only for six-piece burrs, and you asked about 18-piece burrs - you can see some 18 piece designs at Ishino's site:
http://www.asahi-net.or.jp/~rh5k-isn/Puzzle/18PieceBurr/

The current record is 152 moves to get the first piece out:
http://www.asahi-net.or.jp/~rh5k-isn/Puzzle/18PieceBurr/BurrlySane/Extreme/

For woodworking tips, take a look at Lee Krasnow's Instructables:
http://www.instructables.com/id/Precision-Puzzlemaking-Primer----Volume-1/

And Stewart Coffin's tips:
http://www.johnrausch.com/PuzzlingWorld/chap23.htm

And lastly, you might be interested in my page about burrs in general:
http://robspuzzlepage.com/interlocking.htm

Have fun, and mind your fingers!

P.S. - You might be interested in a SawStop:
http://www.sawstop.com/
http://www.youtube.com/watch?v=esnQwVZOrUU

LOL!!!!!!

Maybe it is super safe, but I will never use *my* sausage for such testing!




Pantazis


PS. Thanks Rob, sublime links as usual!
PS2. I have moved it to the appropriate forum.

Noooooo! "Easy Eight / Hard Eight" (the name is a pun on craps terminology) was NOT designed with Burr tools. (But I won't dispute that it's wonderful. ) For those who haven't seen it (most of you), it's a packing puzzle, with a square tray on one side and an (almost) circular tray on the other:



Each side has (essentially) one solution, using the same E I G H T pieces. Burr Tools could not possibly help design this puzzle, which was sort of the whole point. The pieces have curved edges, and the space they must fit in is curved (on one side), so they can go in at any angle. It's not a discrete search space. To computationally search for solutions, ensuring that there's a unique one, means searching a 15-dimensional continuous space (x, y, theta for each piece), a highly nontrivial task. That was the challenge I set myself when designing it, and from my point of view what makes it interesting. I came up with a search program that can do this, as well as solve / help design similar puzzles.

My primary inspiration was was Edi Nagata's truly outstanding packing puzzles:



These are all also two-sided packing puzzles, which can't be solved with discrete search programs like BurrTools. Edi's have the additional property that all the solutions have some sort of pleasing symmetry (a property Easy Eight / Hard Eight necessarily lacks, having all distinct pieces). I asked Edi at an IPP a while back how he designed them, and he said he did it by hand. This got me thinking about how one might go about writing a search program to tackle the problem.

Back to BurrTools, although I agree with everything that's been said about it here -- it is a wonderful tool -- I don't use it much myself. The reason is that what I bring to the table as a puzzle designer is the ability to code up large, complicated, custom searches like I did for the Eight puzzle. It's the same with Subway Shuffle / Athena -- there are hundreds of hours of computer search in there to find the most interesting challenges, and there are a lot of clever tricks to make the search efficient.

Actually there is one design I'm working on that I would like to use BurrTools for, but it won't quite do what I want: coordinate motion disassembly in octahedral / tetrahedral geometry. Well, the public version won't; Andreas' private version will, and he kindly tested my initial design, and found a bug in it.

Just to add to this. The Easy Eight / Hard Eight is by far one of my favorite packing type puzzles. Very well built and satisfying to figure out.

-d

Oooo. So many puzzles to build (when I have the time ).
Thanks for the links.



Haha. I remember solving this puzzle at the puzzle meet up in the tech shop. I really liked how it was made and was interesting to see how much it could pack together.

See man tools and computer things. I just don't understand them. Sorry for the mix up Bob. I remember you talking about burr tools when I was trying to solve it. We might have been discussing the cast rattle or some other puzzle at the same time.

But I still say the Easy Eight/Hard Eight is a great puzzle.

Let me add my voice to those who have praised Bob's Easy Eight / Hard Eight design. I went out of my way to obtain one, and really enjoyed it.

Bill Cutler has also designed a program to solve Nagata-style tray packings. He demo-ed it at the Dartmouth Puzzle Day - Bob, didn't we talk there? He said he uses a "throw and shake" method - but he hadn't perfected it - though it was cool seeing it solve Nagata's arrows in real time after only a few iterations.

And here is another video - the SawStop inventor stands behind his product - he actually stuck his finger into the spinning blade!
http://www.youtube.com/watch?v=E3mzhvMgrLE

Back on-topic - there is an interesting thread on 18 piece burrs over at the Puzzle World forums, wherein Guillaume Largounez discusses his attempts to make burr pieces using various materials:
http://forum.johnrausch.com/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=1&t=000298

Thanks everyone, and sorry to jump on you for misremembering, Rox. (And thanks again for being a customer!) Yes, I think we were talking about Bram using BurrTools for the Cast Rattle (which is amazingly cool).


Yep. Bill's approach to the problem is completely different from mine, which is essentially a highly optimized, pruned, discretized exhaustive search, with local-minimum finding for near misses (that last part is the same as Bill's). I remain convinced there's a better way still, that translates the problem into an equivalent discrete, exact search of a much smaller space. I've got some notes on where to go with that kind of approach, but no code yet.

Back on topic myself -- there are two chapters of Stewart Coffin's "The Puzzling World of Polyhedral Dissections" that cover Burr puzzles, essential material for any enthusiast. They're available online, starting here. (Rob had linked to the woodworking section of this book.) BTW Bill Cutler (mentioned above) also figures prominently in the history of Burr puzzles.