Inspiration

I've started falling behind in my blog on this project, but I did promise to post my inspiration for this development:

So, I’ve decided to design an Orrery. I’ve been thinking about it for about a month now and I find myself lying in bed awake at night and mumbling orbital characteristics to myself throughout the day. I decided that writing this blog about my experience developing one might help keep my thoughts clear, keep my sanity, and help anyone else who decides to try this. Also, I’ve got to hand it to my wife, who puts up with my endless nonsense in thinking about these things.

First of all, while you likely don’t even know what an Orrery is, you probably have seen one. In simple terms, it’s a mechanical model of our solar system. They are typically constructed of brass, but in my research the last four weeks, I’ve seen them also constructed out of wood, stainless steel, aluminum and other materials. Ancient civilizations such as the Greeks knew how to construct them as early as 150 BC, evidenced by the Antikythera mechanism, discovered in a shipwreck off the coast of Greece in about 1900. Orreries are designed to compute the planetary orbits of our planets around the sun, in a heliocentric model. I’m not going to go into the whole history of Orreries; you can read about them on Wikipedia.

I have always been interested in technologies that as modern day people we seem to have forgotten how to develop. There is evidence worldwide that many ancient civilizations knew how to construct mechanical marvels to handle complex calculations, and an Orrery is one such device. With modern day computers, we seem to have lost the technology to build devices like this, and after investigating it for two weeks, it’s not as simple as it seems. There are not any plans readily available on the internet, and I’ve only found about one company that sells an orrery kit. It appears that maybe plans are available in print form, but again, finding books written in the 1700s may be challenging. One guy built a custom one out of wood, which is very nice, btw. Also, another guy has a really beautiful vertically mounted mechanical clock device, that has similar characteristics to an orrery, but was not quite what I was looking for (but is beautiful nonetheless). In any event, it appears that my orrery will be harder to construct than I first imagined. I just found this great website today that has gearing ratios for the planets that I’m sure will come in handy (more on this later).

Because we are in the modern day of computers, I thought I would design my orrery first on a computer, to get all the sizes and gearpieces accurate. If I make any construction mistakes, they are easily correctable in the virtual world. Also, as I lack any brass machining skills or equipment, if I am unable to ever physically build the orrery, at least I can watch it animated on my PC. Plus, maybe someone will come along with machining skills that can help me out. I happen to have several 3D modeling softwares at home, and the one I decided on is Autodesk 3DS Max. I picked it because not only can you model things in 3D, but also you can animate objects and output video files of the animations. However, I have an older version – 3DS Max 7, and it lacks some of the features of the current version. Also, I have not taken any 3DS Max training or courses on how to use it, and so I have to stumble through the exceedingly complicated user interface. It has tons of features, but I’d be surprised if I know even 5% of them.

Off on somewhat of a tangent here – what initially inspired me to do this was the virtual orrery that appears in the Second Life world Nemo, which btw is absolutely fantastic. Here is the direct link, if you happen to have Second Life installed. I first joined SL because of some online friends who said it was a great gathering place. But, in general, I got rapidly bored with seeing virtual people just standing around in various virtual locations not doing anything. I only recommend Nemo because it is just visually inspiring and highlights many of the types of technologies (in Steampunk fashion) that I’m attempting to capture in developing a device like an orrery. I salute Nemo’s designer, Sextan Shepherd, for this masterful virtual world. BTW, my next favorite device in his world is the Tesla Clock… absolutely stunning!

So, let me synopsize my progress to date now that I’ve started this blog:

I have started the modeling in 3DS Max, and realized quite a few things. Knowing what you don’t know can be pretty daunting to say the least!

First, the actual planet orbits, distances and size are far too great to model to scale. For example, if I place the first four planets (Mercury, Venus, Earth and Mars) within an inch of the Sun in my orrery, the furthest out planet, Neptune, would be nearly 20 inches out (at actual scale). If I also include the dwarf planets Ceres, Pluto, Haumea, Makemake, and Eris, this might go out as far as 64 inches in radius! This would make the planets hard to see and also gearing would be significantly compressed on nearer planets. In addition, for solar system body size, the Sun has nearly 109 diameters that of Earth, while the largest planet, Jupiter, is over 11 times the size that of Earth. I started my model to scale, and realized that it would make for a rather unwieldy and maybe not very nice to look at orrery. Earth looks like a mere dot against the background of the Sun.

Instead, I plan to model all my planet orbital distances and planet/moon equatorial radii based on a Base 10 logarithmic scale. That will make the smaller planets and distances seem larger, and further out planets closer in. Also, I can put a logarithmic scale marks on the planet arms to show actual distance from the Sun. Also, I plan to use a logarithm multiplier so that the planets don’t compress so much that the gearing overlaps each other.

Also, I became concerned with the “look” of the planets in my model. Those of you who have 3DS Max may know that it comes with some Planet textures already for putting on spheres of various sizes (in the tutorials). However, I noticed that the textures have seams and aren’t very nice for 360 degree viewing (or maybe I just don’t understand UVW mapping very well). In any case, I found this website that has some great planetary textures for downloading, including some bump and specular maps for terrain. It even had texture/transparency maps for some of the planets rings, like Saturn and Uranus for example! These look very nice when rendered.  I'll post some pics tonight.

I had some challenges figuring out how to map textures to the ring planets Jupiter, Saturn, Uranus, and Neptune. I started with Saturn first and it kept messing up the texture orientation on the model of the ring system. Also, I suppose I’ll have to make some maps for Jupiter and Neptune, based on images from Voyager on Wikipedia. Again, the rings for all four planets also have inner and outer radii that need accurate distance modeling (on a logarithmic scale). Also, I’m assuming at this point that the rings orbit the planets around their hemispheric equator, but I’m not entirely sure. Further, I need to figure out what to do with the “transparency maps” of the rings.

My intent is to convey both orbital revolution around the sun and planetary rotation on its own axis in my orrery. Each planet has its own axial tilt which will have to be modeled, and some even spin in the opposite direction (retrograde), like Venus.

Planetary satellites also become a problem. For more moons, more gearing is required. Jupiter has 63 moons! That would be nearly impossible to develop, and so I’m going to try to limit the maximum number of moons per planet to maybe five of the largest ones. My intent is to pass out two spin speeds on each orrery planet arm (via rotating cylinders), the planet orbit speed and the planet rotation speed. Moon orbital speeds will have to be derived with additional gearing on the end of the arms. Also, I probably won’t model the axial tilt or rotation speeds of satellites, as this would be entirely too much work, except maybe for the Earth’s Moon itself.

Some other things I realized in my first attempt at modeling, was that the orbits of planets are not circular. They are only “roughly” circular, and really follow elliptical paths around the sun. I was using the Semi-Major Axis distance as my distance from the Sun, which could be used if my model was entirely circular, but could be wildly inaccurate for highly eccentric orbits, like that of Mercury. Hence, I had to compile a table of Aphelion and Perihelion distances and consider how to model them. Now, the challenge I am faced with, is that elliptical orbits could be much harder to model mechanical gearing for. Perhaps, it is possible I could use gearing somehow to create a mechanically driven Trammel of Archimedes for my ellipses, but I’m not sure yet.

Additionally, the planetary orbits are not all exactly on the ecliptic plane, each offset slightly by a number of degrees (Mercury, for example is 7 degrees off-axis). Pluto, if I choose to model the dwarf planets, has over 17 degrees of inclination. If I want to accurately represent that, my planet orbits all have to be slightly inclined with respect to the Earth’s rotational orbit (Earth inclination = 0 degrees), and also, they have to be pointed in a particular direction on the orrery, using the angles of the other two axes: the Longitude of the Ascending Node and the Argument of Perihelion.

Also, I just realized today that the planet speed around a given orbit is not a constant velocity. That could present some serious mechanical challenges to say the least. You can see now why using orrery mechanics to model planetary motion are starting to make my head hurt!

Ultimately the goal would be to actually build my orrery, but obviously tools, gears, equipment, and of course cost could prevent that. I figure I’ll just start with the computer generated model and see how it goes. As I’m just sort of outlining my project goals in this post, I’ll get into further progress later.