How to Make a Rolling Ball Sculpture
Updated: 1 day ago
So, you're interested in making your own rolling ball sculpture. If you're like me, you'd rather have the satisfaction of making your own rather than just buying one. That's what got me started down this path. I was given a Spacewarp toy, basically a rolling ball sculpture plastic construction set, and when I needed some new pieces for it, I ran across a site showing the creations other people had made - it was a slap-in-the-head moment... "I'll just make my own!" I said to myself. So, I did.
Use your imagination. When I first began to make these kinetic art pieces, I just grabbed what I thought would work and got started. Some were, well, terrible. They worked, the ball rolled down the tracks, but they looked like I didn't have any clue as to what I was doing. And you would have been right, I really didn't. But every piece I finished and then tested pointed me in a new direction and let me discover what did and didn't work. And what looked good and what looked wonky.
The best part of making these was just letting my imagination direct me. I would envision a shape or style and then just start building. Sometimes it worked out, and other times I had to go in a different direction. But that's the fun of this art. You sometimes just let your project tell you the next step and try not to worry too much about your plans. When people ask me how to make an RBS, I tell them repetition is the key. My first few were not good, but I kept at it, with each subsequent project better than the last. Practice, practice, practice. You wouldn't pick up a musical instrument and know how to play it well the first time; you have to keep at it to improve. Will your balls roll and not fall off? Probably. Will it work overall? Probably. But it'll get better with each thing you repeat and do over. Even now, when I am making a new element, I make a practice one, and using that knowledge, my second and third attempts are ALWAYS better. They work better and look better. Don't be discouraged; this is the fun of creating something out of nothing, and if you're new, it takes time to learn new skills. First off, make sure whatever surface you choose to build on is level. This will affect everything that is built, and you'll be disappointed when the ball doesn't perform as it should when you move it to a different surface that is level. It's always better to be able to say "well, this table isn't level" when it works poorly, instead of the other way around.
Start here with a video I made 15 years ago. It's old information that has been refined over the years but still provides good basic info on the tools I use every day. When I get some time - and ambition - I'll make a new one and post it to this page. I have a lot more cool stuff that I use in nearly every project. But this will get you pointed in the right direction.
I use quite a few different blocks to hold things up, down, in, out, and to weigh them down. I have a bunch of 1-2-3 blocks, four 2-4-6 blocks, 1/8" flat aluminum spacers, several large aluminum plates, and other random things I have found at yard sales and estate sales. It all comes in handy when trying to build something. I even use 8"x16" cinder blocks if I have to get up somewhere tall. It's all needed at one time or another. I've gotten a few of these pieces by buying leftover ends and pieces off of eBay. Great way to get random stuff. Some things are useful because they are long and others are useful because they are heavy.
Materials
First, you want to decide what materials you'll be using. To get started cheaply, I would recommend getting some brass rods from your welding supply store (or Amazon) and start soldering. Soldering is good, fast, cheap, and easy to learn. And, like me when I first started, you can do it on your kitchen table without worrying about sparks or flames. Well, you might have flames - I used a propane torch to heat up the metal for quick joints, but you can use a soldering gun if you have patience.
Of course, you can just get some cardboard and paper and think outside the box (no pun intended, haha) and tape together some sections to make channels and tracks for the marbles to roll in. There are some nice tutorials and videos on YouTube you can look up for inspiration. Tape is cheap, and your mistakes and adjustments are easy to fix. Just get some scissors, tape, and a knife and start creating. Super inexpensive.
I started out soldering copper-coated steel rods (brazing rods), and it was fun. It was quick and fulfilling to make whatever came to mind. However, after several months, they rusted at every solder joint as I had heated off the copper, and the exposed steel rusted. They looked terrible. So, I eventually took all of them and tossed them in the dumpster. Time to start over, I said to myself. That's when I started using brass rods. It turned out way better. They looked better, the nice golden color looked good, and there was no chance of rust.
I made quite a few this way. These pieces were mostly hand-powered, where you'd just put the marble at the top and it would roll down the tracks - no motorized lift or ball release mechanism. I sold them on eBay, and to my great surprise and pleasure, every one of them sold. This was back in the early 2000s. With these early successes in "commercial" sales, I decided I should get a website, and that's when I snagged rollingballsculpture.com.
I use stainless steel, and the majority of the projects I make use these size rods.
1/16" for fill rod when welding
3/32" for some fill (big gaps) when welding and the rod inside tubing for pivot points
1/8" This is my mainstay size. All of my tracks and much of my supports are made from this size
3/16" for stronger supports
1/4" for minor framing and the pivots in my aluminum machined pivot pieces
3/8" for framing on smaller RBSs, like my pyramid frames and smaller boxes
Ball Size
This is when you want to decide what ball you'll be using. As a general rule of thumb, the larger the ball you use, the easier it will be to keep on the two rails. But! The bigger the ball, the more room you'll need when passing one track under another. So, it becomes a tradeoff between ease and density. When I do very large RBS's, I like to use a bit larger ball, maybe a 1.5" or so. With all the extra room, the need for space to pass one track under another isn't much of an issue. The smaller you go, the more pronounced this effect will be.
For most of what I make, I like the 1" glass ball. It's large enough to easily see and light enough to not create too much inertia when going around corners, through loops and spirals, and making quick direction changes. Plus, marbles are colorful. I would avoid solid steel balls because of their weight. However, it is a challenge to get perfectly round marbles that are all exactly the same size, something that is easy to achieve with steel balls. You can use plastic balls; they sell them in all sizes, but their very light weight makes them harder to work with when you have any elements that use the weight of the ball to function. Tipping arms become touchy, and the less the ball weighs, the easier it will pop off the tracks. Not something that will happen with steel balls.
Framing or Support
No matter what you decide to use for your materials, you'll want to first make your frame or support system. This will help you hold things up and give your project some shape and size to start.
Now, most of what I'll be sharing further on will be welding-based, but you can certainly use this info to solder and take some of my tips and processes to incorporate into any material you choose. When I used the brass rods for my projects, I used copper plumbing tubing for the frames. The brass would readily solder to the copper, and I could make squares and pyramids and whatever came to mind for the frames. The plumbing connectors (elbows and sleeves) worked great for making shapes and connecting the copper tubing pieces together.
Your frame will give your piece its initial shape and scope. And most importantly, it will give it rigidity that just tracks connected together won't. I mean, you can do it that way, but you'll be left with a LOT of connecting support rods and pieces, and it won't really look that good - they'll look crowded and messy.
I have made rolling ball sculptures without a frame, but it's hard to keep things up without something to attach to, and it will look more like a heap of stuff rather than an organized linear section of paths. If it's cardboard or paper, just build on a wall. That flat surface can be your support, and it will hold things in place as you build your paths.
For any metal you want to use, you'll likely have to make a frame yourself.
I like to make a box. You can have the tracks stay inside the box or wrap around the outside of the box. Depending on the size of the box you make, you'll likely need a few upright pieces inside it. My general rule of thumb is to have my supports 12-18" apart. Close enough to keep the tracks from vibrating once supported, but not so close that they start being in the way when you're routing your paths. If they're too close together, they can just get in the way. Too far, and you'll be left with very long support pieces that won't really look the best.
Do You Need a Motorized Lift?
Once your frame is completed, you'll want to decide if your project will have a motorized lift, a finger-operated marble release, or if you will just put marbles at the top and they roll down as you place them. This can be a big decision. To answer this question right off: No, you don't need a motorized lift. If you go with no release mechanism at all, you'll constantly be putting marbles at the top to see them roll. After a while of doing this, you will find that it's tiring, and if you're showing the piece off to people, they will be the ones enjoying it. You'll be concentrating on putting the marbles on and not really get to see it go. At least, that's what I discovered when I started out building. Everyone else gets to watch while you do all the work. Still, it's perfectly acceptable if that's the look and feel you want from your piece - and it's the easiest way to see if your tracks and other elements work as you want without going through all the trouble of crafting a lift or release mechanism only to find out things in your project don't really work the way you intended.
The obvious middle ground is a release mechanism that you can press to release one or more marbles into the tracks. When I do this kind of project, usually for a smaller desktop, I build the lift and then build the "marble queue" at the top that holds the marbles in line. Then I make the release using the frame as supports for this, and in this way, I know where the marbles will be when held back. I like to have a ring at the top. When the ring is down, the marbles rest against the top of it. When the lever is pressed, the ring is raised, the marbles pass through it, and the bottom of the ring prevents it from being raised too far. Lower the ring, and the marbles stop. You can release one marble or all of them. This is fun as you can have a whole row of balls ready to release and invite others to press the lever, getting them involved with the action.
This is a bit tricky as you'll need two pivot points on this mechanism: one to allow the handle to be pressed down and the other at the end supporting the ring at the top, which will also need some movement. This should all be done before you continue with your build. You'll have all the room you need to build it, instead of trying to dodge tracks and elements that you may inadvertently put in the wrong place.
This goes for a motorized lift as well. Once you have the frame built, you'll want to build the lift. You'll want everything open and accessible.
Type of Motorized lifts
There are many ways to get the marbles back to the top without any effort more than just plugging it in. I like to use a chain lift and a helical lift, and in the past, I have used a rotating arm lift too.
Below you can see a 45-minute video I made 15 years ago on how to make a helical lift. It's a bit dated as I've refined my process since then, but I think you'll find the majority of the information useful. I used to sell this video, but here I offer it to you at no charge. If you like it, please leave a comment on this page at the bottom, and if you have any questions, you can just email me directly through this website.
As a general rule, try to keep your motor RPM low. Your RPM will dictate how many balls are released per minute. 4 RPM = 1 ball every 15 seconds, and 10 RPM will release a ball every 6 seconds. Too many balls rolling at the same time can become chaotic, and more faults can happen. It is also hard to follow the action when there is so much going on at once.
The other type of lift is a chain lift. I like this one for tall sculptures as there is zero friction, and you'll have to stock far fewer marbles to fill up the lift. I use "ladder chain" on mine as I can weld directly to the chain, but you can use bicycle chain as well. They sell links that have tabs on them that make welding to them very easy. The other advantage of the chain lift is that you can group the pickups together and leave spacing between those groups, creating a more random effect. With the helical lift, your ball release timing will always be the same unless you do not fill the lift completely. The last type of lift I've used is a rotating arm lift.
This one has a long bar rotating in a circle like hands on a clock. At each end of the arm is a ball pickup. The arm scoops up a ball, and when it reaches the top, it releases the ball onto your tracks. This is a fun lift and interesting to watch, but it creates headaches when using it because you cannot route any paths through the arc of the arm, forcing you to go around the ends of the arm's travel. For this reason, I haven't used this type in years - they take up too much space. And if the arm becomes snagged or stopped for some reason, it will strip the gears on the motors, something that generally doesn't easily happen with the helical or chain lift.
What's next...
Now that you have your frame and lift or ball release mechanism installed, you can move on to the next step I take in creating an RBS - the exit from the lift with track splitters and the entrance track leading back to the lift you've created. Now, if you've opted for a non-lift type of sculpture, you won't need to worry about the collection at the bottom, but if you do have a lift, the next two steps are quite important.
Next, I add the track splitters leading away from the lift. I try to space them apart so that there is plenty of room to lead the tracks out from the splitters without being in the way of the other tracks I plan to build. Once they are added, I stop and move my attention to the bottom where the balls enter the lift.
You might say this is getting ahead of myself, but in my experience, this step is crucial to making sure what you are going to make will lead back to the lift without any issues. What issue, you might ask? Well, it's happened to me a few times - I've made a track that ends lower than the entrance to the lift. It's very difficult to make the ball go uphill, so you can either cut out some of the track and move it up, which may or may not be possible. So I learned that if you create the entrance first, you will never run into this problem. It might be just a track that leads out from the lift entrance. That's fine, as long as you know how low you can go. But I like to use a flat steel plate to collect the tracks and merge them back into one. It's usually quieter than having the ball drop from track to track, and it creates a bit of a different look where the ball can roll around on the plate instead of being led around on rails. If your plate is large enough, you can direct the ball to go uphill and then return, creating something fun and interesting, more than just one direction.
Now, once that is complete you are ready to start leading the ball from the top to the bottom and you know exactly where you need to go instead of guessing. It gives you a target to aim for.
Now for the fun stuff - actually building your rolling ball sculpture.
Track Clamps
This is now where you actually have to start making tracks and connecting those tracks to elements and leading the ball around and around your designated space. Depending on what ball size you choose, you'll need to work out the rail spacing. I've found that a good rule of thumb is to make the rails 4/5ths of the ball size. And if still unsure, wider than you think is always better. Too narrow and the ball can come out of the rails more easily than if the rails are a bit wider.
What you'll need to get started are track clamps. If you're soldering or welding, you want to keep the rails in place as you add track spacers. The heat can distort the track and move it while you're trying to get it attached, creating significant headaches. I make clamps out of 3/4" x 3/4" solid aluminum square stock. I put two grooves in the top of them at the right distance apart (about .080 or 3/4" for 1" marbles) and make them .080 deep. The SS rod I use is .125 (1/8") and this leaves a bit of the rod above the top of the surface so it can be clamped down. I drill a hole between these two grooves and insert a "T" bar through the hole. The top of the T bar is 1/4" x 1/2" flat bar (mild steel is just fine for this) and has a threaded nut as the upright. The threads stick out the bottom of the aluminum and I use a wing nut to tighten the T bar down onto the rods in the grooves. This keeps the steel from moving and warping while I weld and keeps the track spacing the same throughout the run. Also, when you are making tracks, eventually you'll get to the end of the rod and need to "butt" weld another rod to that one to continue the track onward. If you keep that butt joint inside the groove and use the aluminum as a backer when you weld them together, it keeps the rods straight and true and prevents the weld from dropping or drooping. It really makes the joining of the two rods pretty easy. I make my clamps with a milling machine. I can get 1/8" end mills to make the groove in the top and dial in the depth. You can do the same with a Skil saw. It's not elegant, but it is effective. Clamp down your aluminum square stock (before cutting them into shorter blocks), set the depth on the saw, and use a guide to keep the saw straight. Any normal carbide-tipped blade will work. Most blades are not 1/8" wide, so you might have to slightly adjust your saw guide over after your first cut just a touch to get more width and make a second pass. Again, don't worry if it's not perfectly 1/8" wide. All you want is for the rails to be kept firmly clamped so they cannot move. That is what is important.
You will want at least three clamps. I consider this the minimum. But more is definitely better. I have 14 clamps and use them all sometimes, especially when I am making a loop d' loop or if I have a long run on a larger project.
Using those track clamps
Now that you have track clamps, you are ready to move on to adding tracks and elements into your project. There are many different ways to make track splitters, and I'm not going to try to give you one "correct" way to build them. I would go to YouTube and watch my videos, David Morrell's videos, and the many other artists out there to see how they have accomplished this task. Look at what's been done, how it was done, and adapt those methods to the tools and materials you have available. That's what I did when I started out, and it served me well. If you're a tinkerer, you'll find a way.
So, I use my clamps in several ways. Obviously, I use them to hold the tracks firmly in place when I add track spacers. I clamp the two rails together and, using alligator clips (or better - test lead clip), I weld one side of the spacer on. Then I remove the alligator clip and, using my needle-nose pliers, I hold the other side of the spacer firmly against the rail. Having a good mechanical contact is key here - if the metal is not touching, sometimes the spacer end will ball up and not flow onto the rail. If this happens to me, I twist off the spacer and put on a new one. The balled-up end is now shorter than you want, and the metal does not weld the same. It's better to start out with a new one.
To make my spacers, I have a hole in my workbench. I feed my rods through the bottom of the bench and into a pipe. I have many different diameters, but you'll want small ones for these, maybe 1.25" or so. Once the wire is through the hole and into the pipe, I twist the pipe to make a coil. I stretch out the coil so I can get my bolt cutters (12" or 14") between the twists and cut lots of curved spacers this way. It's fast and easy, and they all look the same. If your pipe is too large in diameter, the spacer will not hang down far enough, and the ball can hit them.
Now that you have a spacer on, you'll want to continue bending your tracks. I leave my clamp "downstream" from the spacer and keep it clamped as I form the next curve. This is also key, as the metal will bend easier where you have made a weld, and this keeps the rods from bending at that spot. As you form your tracks, keep adding track clamps as you go to keep the rods spaced correctly. Once you have gone as far as you want or need, weld spacers on at each clamp location. Make sure your clamp is on the downside of the last spacer you welded on and keep going. This will keep your spacing correct as you go and prevent that welded spot from bending as you do so.
Adding the large elements first
My process from the top is to add the largest elements first. This would be my loop-de-loops if I am adding them. They take up a huge amount of room and often cross the entire sculpture. If there are any tracks made at all... they'll likely be in the way. Remember, you have your track leading into the lift completed so you know how far down you can go. Next, I would add a tall skinny spiral, tall vertical zig-zag, or large spinners. Anything that uses up a lot of vertical space quickly goes in before I even start to make the final, long track(s).
I sometimes connect these tracks to the end path at this time, but not always. It's up to you. Just don't forget to visualize how these paths will meet up at the bottom and leave room for all your paths to connect. I try to build from the inside out, keeping things more to the inside as I start and then moving out from the center as I go. It can be tricky to get your welding torch inside if you build too many things on the outside of your space. And it's harder to get your hands in there to bend tracks if stuff is in the way.
Now to add the longer tracks
At this point, you should have your frame, lift, track splitters, fast vertical tracks, and the end tracks built. This should leave you with your "long" track. It's what I always leave until last. I leave it to the last because the long track will weave in and out of all the things you've got built and will fill in the spaces between them.
This is where you will need to bend corners, add dips, make straight sections, and lead the ball into tipping arms and other fun elements.
For corners, I always bend the inside of the curve first and then bend the outside to match. Working from top to bottom lets you roll the ball down what you've created and see what the speed of the ball is. This lets you know how much you'll want to bank the corners. This takes a bit of practice, but if the ball is moving, you really can't add too much banking. Too little banking and the ball will come off and fly out. If this happens to you after you've finished your path, you'll need to add guardrail. I try not to use any guardrail on my paths, but sometimes it is required, such as when the speed of the marble is not constant and might come into a curve at different speeds, or if I have a very tight hairpin corner. Guardrail is a must in those cases. Keep in mind that while the corner worked beautifully when you were building, after your piece has been running for a while, the ball builds up dirt and so does the track. This creates traction and will, over time, cause the ball to come off. Be prepared to add more bank or add guardrail.
A common mistake I see new builders of RBS make is trying to get the track down to the bottom quickly. This will lead to headaches and make your piece look "rushed." Take your time and keep the slope of the tracks as minimal as possible. It's fine to let the ball go faster if you have a dip in the track or somewhere you want extra speed. If you're into measuring everything as you go, the minimum I allow my tracks to fall is about 1/8" every foot. Less, and the ball can come to a stop once the tracks get dirty. You can often just step back and look to see how much the tracks are sloped. If it looks like too much, it probably is.
As I build the tracks, I put the ball down them often to see how it rolls. When I finish adding the spacers to a new section, I hold up the new track section with some chain or thin cable and roll the ball down. Once I am satisfied with how it looks and performs, I add supporting rods to keep it in place. I completely support it so it cannot move before moving on to the next section. This keeps what I've built exactly as I want it, and I know as I move on that the speed of the ball won't really ever change as that section is secured. I often use curved supports so if I need to adjust, it's hard to tell I've done so, and if it's curved, you've left yourself some extra room to make that adjustment - you can move something in or out... and the curved supports look better in my opinion. For things that need to be very solid, I use the "V" method and use two supports for extra rigidity - usually straight rods for this purpose. You'll have to use your judgment on what looks the best.
Some people make track sections and then add them to their sculptures. I like to bend my curves in place as I go. If there is a spot where there is little room to do that, I will bend the tracks first and then put them in place, but for the most part, I bend them in place. This makes the corners exactly where I want them without having to make adjustments afterward. And if you have spacers already welded, sometimes you cannot make the change you need and will have to start over. I like to mark out on a piece of cardboard the space I'm working with to keep the tracks inside the allowable room I have. Much better than guessing... or constantly lift the piece of track into place to make sure it fits.
Additional Elements
Now that you are laying tracks, you will want to add some variations and other elements to your project. I have the mainstays in my projects, such as tipping arms, ball collectors, flat spirals, wavy tracks, horizontal zigzags, serpentine tracks, flat plates, gravity wells, and more. You don't have to add more elements. Some of my best works have been just tracks where the ball makes a fun, fast, or lazy path to the bottom uninterrupted by anything until the end. But adding these things can be a great way to break up the track and reverse the ball path. Maybe you are in a spot where there is little horizontal space to keep going, and you want to lower the ball down a bit without making a steep slope. This is where adding elements can save you and look great while doing it.
Check out the video below on how I make my spirals. I don't use the "combs" anymore but instead use my track clamps. Still, this should be useful for the most part. Check it out.
Pivot points
You'll at some point need to have pieces pivot or move. This can be accomplished a few different ways. I have machined aluminum pieces that have a bronze Oilite bearing inserted into them. Very custom, and I use a lathe and milling machine to build these. I also have a jig to drill the holes in the side of the round aluminum, which would be very difficult to do without it. I really like these because I can make micro adjustments to its placement before securing it in place, handy when its positioning needs to be juuust right.
You can use tubing. It's easy to get off Amazon, comes in a lot of sizes, and is stainless steel. When making a tipping arm, build the arm completely and then find the point at which it tips with a ball and then returns. Mark that spot and then weld on the tubing. You'll probably have to drill through the tubing once it's welded to remove any material inside. The downside is that you cannot move it once welded, so you need to be sure it's in the right place. If it's off... you can add some weight to the end that needs it.
The final way I have done pivots is with some washers or nuts and insert a rod into the opening. Effective, but it wouldn't look super good - but it will work and is great for a prototype and you probably already have those in your shop.
