CAD – Target Enigma

[SillyOldDufferModerator @sillyoldduffer]

I’ve long hankered to own a WW2 Enigma cipher machine, but they are extremely rare and expensive.

Building a replica isn’t completely impossible, though as far as I know there are no detailed plans.  A moderately complicated device, and I don’t have the time. I have few books explaining how Enigma works electrically and mechanically, but few good pictures showing what’s inside.   So this will be a free-lance exploration with lots of guesswork and design decisions. I won’t be doing the huge amount of research Cherry Hill put into her models before making them.

I think I can reverse engineer something like an Enigma in 3D-CAD and maybe 3D print most of it in plastic, suitably reinforced with metal.

My preferred CAD tool is SolidEdge Community Edition.  It supports create part in place, whereby new parts are added to existing ones.  After modelling a wheel, the axle is developed on position from it, and then the chassis from the axle.

Typical photo off the web.

First design decision was where to start, and I began with a mistake!  Started with a rotor wheel, roughly half-scale, using next photo as inspiration:

and this drawing:

The web shows several Enigma variations existed, so lots of guesswork ahead, and my effort may not be like any of them.  This is my first attempt, about 5 hours work yesterday. A 26 toothed ratchet wheel missing, and I’m not sure how rotors 2 and 3 are turned over.

Began by modelling the Aluminium thumb wheel which is the main part of the rotor.   It determines the size of all the other components, all of which are modelled as an Assembly from this component.  When I eventually got to the brass electrical contact pins, of which 208 are needed, the head is only ⌀1.8mm with an M1 threaded shaft.  A watchmaking size, and my WM280 doesn’t make producing tiny things easy!  Can be done, but this shrieks ‘production problem’, so I’m tempted to double the scale.   With luck SE has a tool that scales all the parts in an assembly with one click.  Bet it doesn’t, I’m expecting to find that all the parts have to be scaled individually and re-assembled!   Another mistake, the alphabet ring is too wide, and I’ve put it’s spring lock on the wrong side.  Also need to change some internal pillars to make more room inside the wiring chamber.

3D-CAD helps find design blunders without cutting metal.   Physically making a rotor only to find it’s wrong must hurt.

A few negatives about SE.  Adding A to Z lettering to the curved alphabet wheel is long-winded and all eagle-eyed inspectors will have noticed the B is a mirror image.  Flipping it was too difficult yesterday, so another baffled learner session needed.  Would be handy if SE could change several parts in an assembly with one operation in situ, like sharpening a pencil, but, apart from the hole tool, I’ve not found a way, so more work!   SE allows complex features to be copied, but all the faces have to be selected.   How to edit a repeated complex feature later on took time to understand – not done as I expected.

Have an hour to do add the pawl gear now, maybe, more.  Will report again if I make progress.

Dave

 

 

 

 

 

[JasonBModerator @jasonb]

Can  you not wrap ABCD………….XYZ as a single strip onto the curved surface? May need doing to a smooth surface before you cut the shallow slots

[JasonBModerator @jasonb]

Or if engraved

Though it looks like the letters are flush with teh cylindrical surface and have the square background engraved to depth

[JasonB]

On JasonB Said:

Can  you not wrap ABCD………….XYZ as a single strip onto the curved surface? May need doing to a smooth surface before you cut the shallow slots

Good idea, I’ll experiment.  Yesterday I couldn’t see how to type in a vertical strip, and now think I was being dim.

Here’s the result: SE produces the vertical strip, but in the wrong plane.   Hope Nigel is reading this, because when I tried to rotate the text box, the model went wild!   On the right track, I feel, except I’m doing something wrong, maybe selecting too much.

 

I’ll come back to it later.

Thanks for the advice,

Dave

[SillyOldDufferModerator @sillyoldduffer]

Aha, the problem is selecting the text directly on screen opens SE’s text reposition tool which doesn’t do rotate, why not I wonder?  But if the sketch is selected from the pathfinder tree, the SE wheel tool opens instead, and it does rotate and more.

That done the project sketch tool still won’t select the curve though.  I bet money it’s because the curve has been segmented as Jason says.   So this has to be done in the right order.  Project letters first, then emboss/engrave them, then add the segments.  Have to change the alphabet ring because it’s too wide and will apply both changes when I get to back to it.

Many thanks,

Dave

[Nigel Graham 2Participant @nigelgraham2]

Interesting project!

Printing the rotors might even be partly prototypical if the original cores were phenolic-resin (as in’Bakelite’) castings for insulation purposes. I don’t know if they were.

Have you read David Kahn’s Seizing The Enigma, which details the history as well as the principles and operation of the machines, originally invented for secure commercial communications?

[Ian PParticipant @ianp]

If you are thinking of printing the wheels there are some filaments that when printed are look quite metallic but presumably non conductive.

As regards the multiple contacts needed, there are lost of ready made sprung and not sprung test probes like these, (lots of other sources incl Mr Ali)

https://www.peaktest.co.uk/products

The volunteers at Bletchley are very helpful and some extremely knowledgeable, A long way from your location though.

Ian P

[Author]

Posts