Update June 2015: I’m revisiting this project to finally get it to a version where it can be manufactured at home. Check greekgears.com for future updates.
The Antikythera Mechanism is the world’s first known mechanical computer, used by ancient Greeks to predict the path of planets in the sky, the dates of eclipses, lunar phases, and several religious calendars. It was made sometime around 100BC, and it was not until a thousand years later that anything rivaling its complexity was found. Before its discovery we had no idea the Greeks had technology anywhere near this sophisticated.
For a CAD class this semester, I reconstructed the Antikythera Mechanism as a 3D model in Solidworks. It’s to scale, and it works! Watch the video to see it in action, and turn on annotations for a description. Alternatively, download the CAD files for yourself.
The video goes pretty quickly, so I’ll show some pictures of the key features below.
The front face is a planetary display. Pointers track the position of the 5 planets known to the Greeks, the sun, and the moon, all as observed from Earth. These pointers track along the inner ring, which displays the zodiac (the position of each object in the sky). Because of the different speeds of these orbits, some of the planets experience apparent retrograde motion, and capturing this epicycle motion is the reason for much of the complex gearing system. The front also displays the date, which tracks along the 12-months outer ring. At the center of it all is a ball representing the moon, painted black on one side and white on the other. This rotates to show the moon’s phase on any given date – black for a new moon and white for a full moon, with everything in between. You can see this in the video pretty clearly.
The back face is a calendar display. The top half tracks the Metonic Cycle (19 years) and the Callypic Cycle (76 years), which form the basis of the Greek calendar. It also tracks the cycle of the Olympic Games (every 4 years). The bottom half tracks the Saros Cycle (223 lunar months), and the Exeligmos Cycle (669 lunar months), which can be used to predict solar and lunar eclipses! Scientific American has a great infographic on how to use the Antikythera Mechanism to predict an eclipse.
How does it all fit together inside? Well, during my research I found that there wasn’t a complete picture of how the gears were all related. There was a good picture for the bottom half, but the rest had to be cobbled together from several different places. So to help me visualize it while working on the project, I drew up a map of how everything fit together. And because I had some free time yesterday, I cleaned it up and put together this infographic to make it all clear (it’s on wikipedia now so it doesn’t get lost). It’s pretty self-explanatory, and hopefully it will be helpful if anyone in the future wants to make their own reconstruction.
I’m pretty awestruck, to be honest, that someone was able to make this over 2000 years ago. Even though it was built during the time of Rome’s peak ascendency, it is so many levels above any other technology we have from that day that it seems almost anachronistic. I had enough issues coming to understand the complexity and detail that is packed into this small box, and I already knew how it was going to work! It makes you wonder what other marvels were created that have been lost to time.
“This device is just extraordinary, the only thing of its kind. The design is beautiful, the astronomy is exactly right. The way the mechanics are designed just makes your jaw drop. Whoever has done this has done it extremely carefully … in terms of historic and scarcity value, I have to regard this mechanism as being more valuable than the Mona Lisa.”
– Professor Michael Edmunds, as quoted in The Guardian
I’ll close off with a list of the papers I used. The first Freeth paper was by far the most helpful, and I might have been able to make it based off of just that paper alone. The others were helpful for context and comparison though. And this youtube video, even though it was for a different gearing schema than the one I used, was helpful in wrapping my head around how it would go together.
Evans, James, et al. “Solar Anomaly and Planetary Displays in the Antikythera Mechanism.” Journal of the History of Astronomy 61 (2010)
Freeth, Tony, and Alexander Jones. The Cosmos in the Antikythera Mechanism. Institute for the Study of the Ancient World, 2012.
Freet, Tony, et al. “Calendars with Olympiad display and eclipse prediction on the Antikythera Mechanism.” Nature 454 (2008): 614-617
Freeth, Tony, et al. “Decoding the ancient Greek astronomical calculator known as the Antikythera Mechanism.” Nature 444 (2006): 587-591.
Wright, Michael T. “The Antikythera Mechanism: A new gearing scheme.” Bull. Sci. Instrum. Soc 85 (2005): 2-7.
Pingback: CAD files for the Antikythera Mechanism | The Shamblog
Pingback: I made a 3D model of the Antikythera Mechanism. It works! | Amateur-Science
This is so cool. Thanks very much for posting!
This is really cool! Just curious, what were the specs on the computer you used to create the animation?
Not good! It took a long time to process, and as you can see still had glitches.
Non-the-less, Scott, this is very impressive. Thanks for posting.
Anything come from this?
There is an issue with your model. The concentric tube system is too skinny. Perhaps you confused radii with diameters? The three fixed superior planet cogs Jup1, Sat1 and Mar1 should have a 14.4mm diameter hole in them, as should the Front Sub-Plate. This is to allow 0.2mm clearance for the concentric tube system, which has an outer diameter of 14mm, Saturn being the fattest. Each tube should have a wall of 0.5mm thickness and a clearance of 0.2mm between each tube. The moon in the middle is a solid tube, 2.1mm in diameter.
The back spirals are wrong. They do not connect up together as you have them – they are quite separate, with a gap and writing between them. See:
https://commons.wikimedia.org/wiki/File:Antikythera_Mechanism_back_panel_-_Freeth_and_Jones_2012.svg
You are right on both counts (this and your other comment). I’ve fixed the back spirals in my current draft of my update. As to the tube diameters, this was just meant to be a functional recreation rather than a faithful reproduction. I think my largest crime is probably ignoring all the fasteners that are in the real mechanism.
Understood about the functional recreation, Scott. I am in the process of redrawing it, using your data as a starting point. Triangular teeth etc. Plus all the mounting brackets. Happy to throw open what I end up with, to those that can use it. It’ll be in .OBJ format though. It’s a complex thing to wrap one’s head around, for sure.
Hi Scott, I owe you an apology. I have just been reading Alexander Jones’s book “A Portable Cosmos: Revealing the Antikythera Mechanism, Scientific Wonder of the Ancient World” and with regard to the back spirals, he says the two spirals were close, if not touching or joined. They were certainly not one spiral width apart, as in many of the recent reconstructions. Jones shows them joined as one continuous spiral, so your model reflects an expert! For the purposes of stability, it would have been helpful to have had at least a small bit of metal between them.
Greek Rhodes is a possible candidate for the origin of the mechanism, but Syracuse in Italian Sicily is a more likely candidate, from the inscription evidence. Home of Archimedes.
Brilliant to see some movement on this again – really looking forward to seeing the first finished working example. It’s a complex old beast to be sure – I’m the idiot who builds the wooden versions by the way.
So many gears…
I have been looking at many sources to try to figure out how to build my own and I’ve seen you EVERYWHERE! The mechanism you built is the best one I’ve seen!
Hi Anna,
Thank you!
It’s not often that there are comments on this blog at the moment, so I was a bit surprised when yours just popped up in my notifications. I just hope that Scott is getting on well with his printable version – it would be great to finally see it working.
No, I guess that I haven’t been shy about shouting about my machine over the past few years, but I guess it’s my way of making a basically very lonely pursuit (many, many hours spent in my little shed…) into something a bit more social, and, let’s face it, we all crave a bit of online love for the stuff we’re doing every now and then!
If you want any more info please feel free to drop me an email at davidgoodchild69@gmail.com – I’m more than happy to answer any questions and give you any info that you need.
All the best,
Dave.
Any updates since Freeth new model was discussed in 2021?https://www.nature.com/articles/s41598-021-84310-w
Hey Anthony, not yet but that’s on my to-read-and-update list.
Scott
Your model and the gear chain image in Wikipedia have been my saviour.
I first started thinking about making the mechanism about two years ago and your model has been indispensable in understanding it’s workings and layout.
I’ve reworked it in SW and with the exception of some details I now have a workable model. I’m having trouble with mates in that the gear chains run ok in sub-assemblies but won’t run in the main assembly.
I have manufactured the assembly from the plate carrying the metonic and Saros pointers up to and including the intermediate plate. All works exceptionally well with all the gears meshing well but there is some excessive friction in the E5/6 K1/2 assembly which I’m trying to sort out. I have also manufactured the plate assembly containing the Sun, Venus and Mercury gears and this works well when slave assembled on the lower assembly.
It’s made almost entirely of brass and all the gears have cycloidal profiles which i have machined with a home made CNC macine using a 1mm end mill. Unfortunately this has necessitated using a module of .7 and the model will be 40% bigger than the original. Clocks, in general, use cycloidal gears and are designed with more slop between teeth. Wheels are 1mm thick and pinions 3mm.
Almost all the gear assemblies are bonded together using either super glue or Looctite. In some instances I have used solder.
I haven’t yet worked out how to mark the front and rear plates with the Greek characters, possibly engraving or printing, so any advice would be welcome
I think it will take me a couple more years to finish the mechanism as every step takes infinitely longer than planned.
Thanks again for your model.odel. I’m having trouble with mates in that the gear chains run ok in sub-assemblies but won’t run in the main assembly.
I have manufactured the assembly from the plate carrying the metonic and Saros pointers up to and including the intermediate plate. All works exceptionally well with all the gears meshing well but there is some excessive friction in the E5/6 K1/2 assembly which I’m trying to sort out. I have also manufactured the plate assembly containing the Sun, Venus and Mercury gears and this works well when slave assembled on the lower assembly.
It’s made almost entirely of brass and all the gears have cycloidal profiles which i have machined with a home made CNC macine using a 1mm end mill. Unfortunately this has necessitated using a module of .7 and the model will be 40% bigger than the original. Clocks, in general, use cycloidal gears and are designed with more slop between teeth. Wheels are 1mm thick and pinions 3mm.
Almost all the gear assemblies are bonded together using either super glue or Looctite. In some instances I have used solder.
I haven’t yet worked out how to mark the front and rear plates with the Greek characters, possibly engraving or printing, so any advice would be welcome
I think it will take me a couple more years to finish the mechanism as every step takes infinitely longer than planned.
Thanks again for your model.
Bob H
What an awesome project. I contacted Tony Freeth a couple of years ago asking if I could have a copy of his model – but alas the answer was a very polite no. Incredible that you have done this – very exciting indeed !
I’m not sure how far Scott has got with his machine – it’s been a long haul and I think that the poor fella is busy! Many years ago I also tried to contact Tony Freeth and got the same response, so I just built my own… out of wood…!
I’m now literally just finishing cutting cutting all the gears for machines 10, 11 and 12 which I’m building as commission pieces. Yes, I’ve built a lot of these now.
If you’re interested in having a look – I can’t take on any more work at the moment, so I’m honestly not looking for a sale – but please do check out my Instagram account where I go by the name of antikythera_man. If you wade right back through my posts you’ll find pics of finished versions of my machines from a few years back, but I’ve more recently been diverted by building lots of complex automata, so don’t be put off by all the wooden magicians, but I’m now back to Antikythera again!
Anyhow, if you’d like any info, you can reach me on davidgoodchild69@gmail.com
Cheers!
I congratulate you on your excellent and painstaking work in recreating the Antikythera calculator.
I am writing a book about it and I will include the link to this video if you like.
Congratulations!
Fernandez
I’ve recently manufactured an Antikythera mechanism almost entirely in brass. I produced a working Solidworks model of the unit as I used a CNC router to machine most of the parts. It’s 40% bigger than the original as I was using a 1mm end mill to machine the gears and daren’t go any smaller.
It’s all working but I haven’t set the heavenly bodies correctly as I’m uncertain how to do it.
I’d send you some photos but don’t know how to.
If I can help in any way let me know.
Bob
Hi Bob,
Could I possibly have a look at your finished machine? I’ve built 12 fully working Antikytheras over the past 8 years, all with planetariums (all of mine are made from wood), but I’d love to see your version, if I may. Mine are also 40% bigger than the original!!
I’d love to build one out of brass eventually, but I’ve only just recently started moving my workshop in that direction by investing in tools that can handle more than just wood!
My email is davidgoodchild69@googlemail.com
They’re are very few people out there like us who are willing to have a go at building one of these incredible machines, so I’m always interested to see what others have done with the design!
All the best,
Dave Goodchild.
UK
Dave
We communicated some while ago and I will forward you some images of the mechanism and some from the CAD model. I spent most of my hobby life working with wood and migrated towards metal when I started making the clocks. I have a Myford lathe and a home made CNC router.
Bob
Hi,
Great work, thanks! I followed another approach, modelled the device in Excel (see http://www.astronomy-morsels.ch/morsels). First the gear train, and then the back/front dial. On the front dial, there are both a Date Pointer and a Solar Pointer. Do they represent respectively the mean and true solar time?
Greetings from Switzerland,
Anton