Rotational Casting Machine

Here are some basic instructions needed to build a rotational casting machine. Below is a list of the tools and supplies I used for the project. You don't necessarily need all the tools, but I found them to be incredibly useful. If you build this let me know what you did with the belts and how it worked. Best of luck!

1. Chop Saw

2. Drill press

3. 1 1/8” bit

4. Power drill

a. Phillips driver

b. Small drill bit – size of screw shafts

c. Big drill big – slightly bigger than the head of the screws

5. Table saw (for trimming down the sides of the 2X4s to make them square)

6. Ruler up to 38”

7. 3D printer

a. filament – 1.75mm PLA

8. Sandpaper – 120 grit

9. Screws

a. ¾”

b. 1”

c. 2”

d. M30 – 1” and coinciding nuts for the corner pullieys

10. 2mm/6mm steel core belt – 5 meters (a little extra is nice to have on hand)

11. 2 part plastic epoxy

12. Barings – 4 for the corner pulleys

13. ¾” PVC pipe about 3’

14. Wood glue

Step 1: Print Parts With 3D Printer

The first step in this whole process is to print out the pieces you need. Use 3 shells and at least a 50% infill for strength. It doesn’t matter what color you print in, but pick one you like. You can find a listing of the parts and stl files here: https://www.thingiverse.com/thing:2723443

a. Rotational pulley – print 2

b. Stationary pulley – print 2

c. Corner pulley – print 4

d. Short end-cap – print 2

e. End-cap – Print 2

f. Washers – print at least 8, but you may need more

g. Bushing – print 4

Step 2: Wood!

The second step is to cut all of the wood to the lengths that you need. I used 2x4s that I cut down slightly to square off the edges. Remember to measure twice and cut once. I made this mistake and had to recut some pieces wasting wood.

a. 25.5” – cut 4 (inner axis)

b. 30.5” – cut 4 (outer axis)

c. 31.75” – cut 2 (frame bottom left/right)

d. 14.25 – cut 4 (frame bottom left/right)

e. 24” – cut 2 (frame upright)

f. 38” – cut 2 (frame front/back support)

g. 35” – cut 1 (frame center)

h. Bracers that are 20.25” on the long side with a 45* angle – cut 4

Step 3: Lets Make Holes

Using the drill press and the 1 1/8” bit drill holes at the mid points of the six pieces of wood that will be spinning. 2 of the inner axis pieces and all four of the outer axis pieces.

Step 4: Cut You PVC

Cut the PVC pipe so that you have four lengths that are about 6.5” each. You will have some pipe left over, which is ok.

Step 5: Alignment Check

Lay out the inner axis so that you can see how the pieces are going to fit together. You want it to be wider than it is tall.

Step 6: Pilot Holes, Sink Holes, and Screws

Using the small drill bit (the one that is the same size as the shaft of your screws) drill pilot holes where you are going to place your screws at the corners.

a. Make sure your box is perfectly square before you drill the pilot holes all the way down. This will make sure that your pilot holes are exactly where they need to be.

Using the larger drill bit (the one that is slightly larger than the head of your screws) drill your counter-sink holes down about 1/4” using the pilot holes as placement locators.

Repeat this process with the outer axis, making sure that it is the same layout as the inner axis.

Using your 2” screws (16) screw both your inner and outer axis boxes together.

At this point you can set your inner and outer axis's aside so you can work on your frame.

Step 7: The Frame

My recommendation is that before you do any drilling or screwing you set up your frame so you have a general idea of exactly how it goes together. It's like a weird puzzle, but you'll figure it out. This also allows you to see if any adjustments need to be made to any of the lengths of wood.

Step 8: More Holes

You'll want to drill two more spin holes, using the 1 1/8" drill bit, in the side supports of the frame. I did it so that the top of the hole was 2" down from the top of the side supports.

Step 9: Wood Glue Is Useful

Using the 31.75" and the 14.25" left/right boards set them up and apply glue to the contact points.

Step 10: More Pilot Holes, Sink Holes, and Screws

Drill 2-3 pilot holes in the boards at the points where they will attach to each other, keeping in mind where you need to place your side supports.

Then drill your sink holes where the pilot holes are about 1/4" down. Just like before.

Step 11: Clamping and Screwing

Clamp your frame together and make sure it is squared. Then you can use your 2" screws to attach your pieces together. Use a little extra wood glue if needed to make sure that you have a nice solid frame.

Step 12: Wait

Let the glue set and dry for 24 hours before continuing.

Step 13: Putting It Together

Using the 3D printed parts you should be able to piece the machine together. You'll need the 1" screws to attach the printed parts to the wood pieces.

This is the point that you'll want to use the M3 screws and nuts to put the bearings in the corner pulleys.

Step 14: Belts

The belts are the tricky part. I am still working on perfecting it. In the minimal instructions given on the original thingiverse post it mentioned using staples to put the belt together. this didn't really work. The belt was either too slack, or the staples didn't hold. I am still working out the kinks in the belt issue.

I did solve some of the slippage issues by using epoxy to glue a length of belt to each of the rotational pulleys. This made it so that they had teeth for the belt to hold onto.

If the belt it too loose there will be lots of slippage and your machine won't do what you need it to. The best bet to get a nice tight belt is to wait to put the second corner pulley on each diagonally opposite corner. That way you can use the pulley to increase the tension to what you need it to be.

Make sure the teeth of the belts face in so that they connect to the teeth on the pulleys you added.

Step 15: What Crank?

Due to how construction works there was no way to attach a crank to spin the machine, even though in the original thingiverse post says there should be one.

It's not a problem though. It really is easy enough to just manually move the outer axis. If everything went together right it should automatically cause the inner axis to rotate at the same time.

Step 16:

Comments

[DrRadium]

I'm interested in how you attach molds. I am using a cassette made from plywood but it is clumsy and takes forever to change molds.

[erjack1]

You strap the mold to the two support beams in the middle. I used ratcheting straps because they go on easy and come off quickly.

[DrRadium]

I put support beams on mine and have been using mold straps with much easier changes and placement than the original sandwitch. Thanks!

[DrRadium]

I'll have to try that. I wanted something to keep the mold's center of mass right in the center of the frame so spining the caster would require equal force at all times, so I went with the sandwich system to avoid being limited to the width of the frame . But you've got me wondering if it would be easier to go with support beams and ratcheting or mold straps. If the balance matters as much as I was fearing, I could add weights on the beam side opposite the mold to shift the center of mass. As is I have to add some weights anyway because the molds are never symmetric anyway.

[PS118]

I really like this. Nice blend of 3d printing and traditional construction.

Keep us updated when you get the belt problem resolved!

[DrRadium]

I think I solved the problem by changing to a bevel gear system and chain/sprocket. See my comments from today on step 14. It uses bevel gears that are printed in Taulman bridge nylon. The bevel gears have held up for just shy of a year without any problems. They've been a big improvement over using T2.5 belts, which were in a big improvement over G2 belts. Only issue is that it does require making a stationary sprocket. I did this by welding but it should be possible by soldering with a simple propane torch using stay-bride silver solder.

[DrRadium]

I made a rotational caster using the same thingiverse source that you used. My solution to getting the belts to work was dumping the rubber GT2 belts and replacing them with T2.5 steel reinforced urethane timing belts. The urethane is thermoplastic so it can be welded with heat to circularize the belt. Belts were joined by scraping off the urethane until I could just see steel wires over area of at least 5 ribs. On one end I scraped from front to back, the other back to front. Used heat gun from soldering station to start melting the belt urethane, pressed together hard between wood blocks or with vice grips. Excess urethane was trimmed to leave uniform edges. These have not broken though they do lengthen as the ribs at the junction stretch apart. To help keep the belt useful as long as possible, I made tensioners using nails and some small copper tubes.
Here is my make https://www.thingiverse.com/make:327099

EDIT: even the steel reinforced belts stretched and broke at the worst times and I gave up on the belts. I went to a chain and sprocket system that tuned an alumumum tube that turned the inner frame via a bevel gear printed in bridge nylon. It has not failed since I built it. Info and pictures are still at https://www.thingiverse.com/make:327099
Video of it in action is at
More details are in the replies to this post.

[erjack1]

That's the kind of belt I used. I'll have to try this to see if I can get it to work. I also used it to go around the two big pulleys to make sure the grip was good and wouldn't stretch out. Yay, steel core.

[DrRadium]

I eventually ended up dumping the belt altogether. Even with T2.5 steel reinforced belts, they would slowly stretch and eventually come off or break, usually in the middle of a casting run. So I switched to a chain and bevel gear system, and also changed the PVC tubing to schedule 40 aluminum pipe for improved rigidity. In place of the belt I used a #40 roller chain. The stationary pulley was removed and saved for later. Others may wish to move it to the outer side of the base as a guide for the axle. I didn't need it because I had already put in a bearing here, In place of the original stationary pulley, attached a stationary sprocket I made by welding sprocket to a piece of plate steel with a hole cut in it for the now aluminum pipe. On the outer moving frame, I used a 1/2" aluminum tube with a sprocket attached at one end and the small bevel gear of a set I printed in nylon at the the other. I removed the rotational pulley, and replaced it with the previous stationary pulley to act as a guide and spacer but the axle was extended when I replaced the PVC with aluminum. The other larger bevel gear was attached to this axle. To guide the 1/2" tube, low profile base mounted bearings were used. I published the bevel gears at https://www.thingiverse.com/thing:3161163. Mine ware printed in taulman bridge nylon. I found that it was important to replace both belts, and not try to rely on just one as the bevel gears tended to slip with large molds if only one gear set was working. I've included pictures to help in visualizing.
9/2021 EDIT: this has held up very well. I am still using the original printed bevel gears!

[DrRadium]

Another solution to the problem of the crank is to motorize it. I've wanted to do this for a while but only recently found a inexpensive motor that worked. I found one on eBay that seems to gave enough torque (12VDC, 15rpm, 1.5A 18W with 40Kg cm torque [= 3.92 N m torque]). A picture of how it is connected is attached and a video of it can be found at https://youtu.be/Z7vuvAVYPH

I should probably also note that I added bearings to the shaft between the supporting structure and the first rotating frame. It made the motion smoother, but not nearly as much as I was hoping it would. And I don't think it decreased force needed to turn the frame by much at all, if any. Not sure if I would do it again. Probably would't recommend it unless there was a problem. This bearing can be seen in the picture of the motor mount, but is absent in the picture of the coupling nut attached to the PVC tube that is above.

[erjack1]

The .mov file didn't work. But the motor seems like a great idea. It'd be cool to see how fast it spins the machine. And is it powerful enough to spin the machine with a fairly heavy mold. It's such a big thing it'd be a shame if you couldn't use the full size of it.

[DrRadium]

I changed the system used to hold the mold to the one you used. But before aI did, I ran with the original one that used 2 pieces of 2ftx2ft plywood and 3/8" bolts and spun a mold weighing over 6kg, not counting the resin. The motor I described spun it just fine at 15rpm.

After I switched the mold support to the system you used, I spun again. Unfortunately a heavy 10ft section of 6" PVC pipe that I was using to build a dust evacuation system had fallen in its way and the rotocaster had to try to lift it to continue moving. I rushed in to get the pipe moved but the motor had almost done it by itself and as still moving when i was able to help lift off the pipe. It finished the run without a problem. Since then I haven't been too worried about the motor's capacities!

[DrRadium]

Screws and bolts work great on the PVC. That's how I attached my crank to the PVC axles well.

The video didn't upload properly so I uploaded it to YouTube and edited the original tip. For people who don't want to scroll down here it is: https://youtu.be/Z7vuvAVYPHc

It ran at about 15rpm which is, what I have read, optimal for most rotational casting. I haven't run it on any huge molds yet, but it does work just fine with moderate size ones even with my heavy attachment system (two pieces of plywood, 2x2 corners and 3/8" threaded rod and nuts). So it should work well with a simper (and lighter) attachment system. The motor only cost $40 on eBay. A better (more expensive) one with more than 40Kg cm of torque should be even more reliable. I will post more on how well it works with heavier molds as I work with the caster.

One note, from what I understand, many Chinese motors have overstated torque ratings. So one from the US, Canada, or EU with the same rating may be more powerful. Also, I found during my quest for a motor that some have torque ratings that exceed what the gearing in their gearboxes can handle. So while the gearing ratios will come out to calculate a high enough torque, the gears will break when you try to do it. This was the case with a 19rpm economy gear motor from Servocity. It had a 192Kg cm torque rating. But when I tried to use it to spin this thing, one of the gears teeth broke off before it turned the thing even a 1/4 of a turn. This wasn't a fluke. It happened twice. At the same gear.

[DrRadium]

To attach a handle I made an adapter. A PVC tubing cap and coupling are used with some tubing to make an adaptor. A 3/8" hole is drilled and tapped in the PVC cap and the threaded rod screwed into it. The rod is then stabilized by pouring smooth-cast urethane resin into the adaptor and letting it cure. A 3/8" coupling nut has 1/4" holes drilled and tapped on every other side near one end of the nut. Set screws are used in these to hold a hex rod for adapting to a drill, or to the hex wrench used for garbage disposals as a crank handle. The other end is slipped onto the PVC pipe that this the outer rotational axis and help in place with coarse drywall screws. It isn't pretty but it works. This doesn't show the bearings that were Iater installed here. They didn't make nearly as much difference on force needed to turn as I thought they would. Turning was just a bit smoother, so I don't really recommend it.

[erjack1]

I did figure a way to attach a crank. You can drill holes in the pvc pipe and a the pvc couplers, then use a sturdy nut and bolt to lock it in place. Doing two perpendicular to each other in eases the strength and stability of the crank. It also means you can remove the crank if you need to. Keeping the crank close to the side of the machine ups the leverage so you don't need to work as hard to turn the thing.

[tomatoskins]

Very cool design! What is the benefit of this over other casting methods?

[erjack1]

Less arm work mostly. You can do a solid mold, not just a pour mold. And you can get a nice even layer of material inside the mold because the spinning helps the material coat evenly. Just don't use Smooth-on Task15. It need air to cure properly and a solid mold doesn't allow that.