Feynman's Dial Combination Lock Box

Both a Nobel prize winner and a child can learn with this model. As a child, I've seen a safe-type combination padlock but could not guess what is inside. Years later, Feynman explained it to me in his book. It was a double pleasure: I finally knew how it worked and I knew that the Nobel prize winner also did not guess it without disassembly.

This model makes it crystal clear what that hidden secret principle is. You can try to open it with the shield on - and you will likely fail. With the shield off, the mechanism is still fully functional but you can observe it! And, well, you can also just know the password which is MRKEV, the Czech word for carrot. Incidentally, a nice collapsible carrot can be inside!

To open it, rotate the lid to the right several times (direction like fastening a screw). Stop exactly at the first letter of the password. Then turn left exactly to the second letter of the password, etc. (If you overrun a letter, start again.)

The mechanism works as it is but greasing the inner part with candle wax (or solid soap if desperate) makes it much more enjoyable, especially for PETG (PLA glides a bit better but you can grease it as well). More importantly, greasing will make it more reliable in the long run as the parts will be less stressed.

Ii is designed for PETG but can also be printed in PLA. Take care when putting the rings in place, especially with the more fragile PLA. Try to stress the rings as little as possible when putting them in place. You need three rings (or less). Not four, the last rail serves different purpose, leave it empty. Some PLAs may be totally unsuitable for the rings and for the shield (e.g. my Prusa Galaxy Black is sort of OK while the Prusa Silver is too fragile). There is an alternative shield2.stl which should be more durable in PLA. In general, prefer PETG.

I printed it with 0.3mm layers. Smaller layers might be better for rings - or even just different layers may reduce friction.

I made this model mainly to explain the secret mechanism, to share with you my fascination by the simplicity of the mechanism which stores the password as you enter it. There are many ways to make it safer as a lock box which is really hard to open. I accented didactic features much more than security. You may want to make the mechanism invisible, to use a different password, to make it more robust to withstand a brute force etc. I'll later put some optional replacement parts to a different model to keep this one simple.

From the security perspective, this mechanism is very interesting (even in my didactic version). Many locks can be compromised because the blocking part can be moved while some lock-opening pressure is applied to it. It is a common flaw which allows lock picking. This mechanism (at least when rendered in high precision) is very different - it decouples setting the blocking elements from testing whether the position is right (my simplified mechanism nearly compromises this property but note the bigger dent on the lid which does not allow the smaller dents to touch the rings too early). Another example of such decoupled action were some old US coin operated phone boxes with a very special lock which first remembered sizes of dents on the key and only then tested the positions for being all OK.

Enjoy! And please publish your makes, it always helps me to make my models yet better when I see how they work in the wild!