Nylon Filament

The bottle cap I'm working on obviously needs to not leak, by sealing to the top of the bottle. It has two problems: it's made of PLA which is rigid, and the underside of the "deck" is irregular because it is printed over a grid of support material. Most bottle caps use a soft plastic gasket or washer where they meet the bottle, maybe nylon or teflon. I figured it is unlikely I would find a suitable part of exactly the right size, so naturally I thought about printing it, and looked into nylon.

Nylon has become available for 3D printing in recent years. From what I read it is rather more difficult to work with and about 50% more expensive than PLA. I need only a tiny part and didn't want to buy a whole spool just for that. But I think I'll find other uses for it, so I ordered a 1-pound spool. A filament called Bridge from Taulman seems to be rather popular because it reduces some of the problems with earlier products.

It occurred to me that maybe I could print the whole bottle cap in nylon and avoid the need to fit a gasket. Rather than print the whole thing, I first printed a "slice", the center section of the bottom, so I could observe the threads. I needed to adjust the depth of the bottom anyway. (I like finding shortcuts to avoid printing the whole part every time.)

With nylon one can't use hairspray for adhesion. The two popular methods are glue stick and blue tape. I tried the glue first. It's easy to apply. My first print stuck well except for one arm. It seems I didn't get the glue over the whole print area, which seemed strange. I tried again and missed again. What's going on?

My printer has a glass heated bed. It has a rectangular "hot zone" printed in it, and the Robo3D logo in the center. Except... it's not centered! All this time I've been applying hairspray roughly on the logo and it's been working OK but only by luck. Reading on line confirmed this. I had noticed that the nine leveling touches didn't line up with the square but had not thought it important. Kind of dumb of the manufacturer to print something that implies alignment and not line it up. So I measured things and made a center mark, applied glue accordingly, and it stuck fine.

After making some minor adjustments I decided to try printing the whole bottom half in nylon. It printed OK, but two problems arose:

• The support material was really hard to get out. It took a lot of cutting with a Dremel saw bit and a lot of pulling.
• The bottom is too flexible. When I close it tight the threads can actually skip and lose the seal.

So I won't be producing the whole thing in nylon. Now I'm working on designing a gasket that will be held in by friction and act as a seal between the bottom of the deck and the glass bottle neck. The whole bottom needs some size adjustment, and since the gasket is also parametric it will resize at the same time.

The other method for adhesion is blue painter's tape (masking tape). It's thicker than glue, so now I'm trying to figure out how to adjust the Z dimension upward. MatterControl shows a Z offset but does not let me change it. If I don't move it up, the nozzle drags a groove in the tape and does not extrude properly. I'm not sure how to only place tape in the center and still have the auto leveling feature work. If there's no tape in the  other eight spots, the algorithm will not adjust the Z. Do people tape the whole bed? That seems excessive. Maybe I need to add tape in all eight leveling spots. 

Work in progress: sports bottle top

This one has taken me a while. We have a couple of these nice glass bottles in rubber sleeves. The problem is that the caps screw off and are not tethered. Not convenient for one-handed use such as when driving. I know, "first world problem", right? My wife suggested a flip-up top as a project. Imagine the flip-up top on your shampoo or syrup bottle, on steroids. It seemed like a good project to learn some techniques. It sounds simple, but it's actually fairly complicated. I didn't find anything like it on Thingiverse, so it's not trivial (or not in much demand). It needs to be relatively attractive.

There are several "moving parts" that need to fit and work just right, and it needs to be 3D-printable:

1. The screw-on part needs to be just the right depth to seat correctly so it will not leak fluid around the neck.
2. The threads need to match the glass bottle neck.
3. The flip-up top needs to snap on and off easily yet securely.
4. The top needs to have a plug that fits into the drinking hole so it will not leak fluid into the cap when closed.
5. The flip-up part (top) needs to be hinged to the screw-on part (bottom).

(I recently bought an electronic caliper. This would have been much easier if I had had it all along!)

I started thinking about it in late January. I've been designing and printing trial parts since mid-February, or about 6 weeks off and on around other projects and real life. I'm working in Fusion 360 and trying to design it completely parametrically. I now have a working design which meets 4 of the 5 requirements. I'll discuss each in turn and then update it as I finish the design and testing.

1. The screw-on part needs to be just the right depth to seat correctly so it will not leak fluid around the neck.

Right now it's about 2mm too deep, so the bottle top doesn't quite seat. But since I'm working with PLA, even if it seated I don't think it would seal tightly, because PLA is hard and doesn't "give". So I'm thinking a nylon gasket or washer up in the top would do the trick. Depending on the thickness I find, I might need to adjust the depth again. I've thought about spreading some sort of food-safe silicone around the inner top to form a gasket. I've also thought about switching to Nylon for the whole project. The texture would be better and it should seal properly, but it's more expensive and issues with moisture in storage.

So this part is not done yet.

2. The threads need to match the glass bottle neck.

This turned out to be tricky. None of the threads that come in Fusion 360 seemed to be quite right. It turns out that glass bottle neck thread standards are established by the Glass Packaging Institute. All the info is freely available on their web site, if a bit hard to interpret. Eventually I figured out the neck size and thread specs in their terms. Then I had to learn to create threads "by hand" in Fusion 360, following tutorials and examples. It requires creating a spiral of the right length and number of turns, creating a profile for the thread, and then "sweep"ing the thread profile along the spiral. It worked.

I printed some little slices of the bottom in order to more quickly test the threads without printing the whole model.

Later I'll talk about the printing orientation. It turns out the design needs support material inside the bottom adjacent to the threads. I was afraid the support would be difficult to remove and jam up the threads, but it actually worked out OK.

3. The flip-up top needs to snap on and off easily yet securely.

I designed a "step" around the top of the bottom, and an extension (not sure what to call it) around the bottom of the top, so the top will snap down onto the bottom. The diameters and heights need to be just right. I used a bit of chamfer to make them seat into each other. With a tolerance in the parameters, I can adjust the tightness of the fit. The latest sample is fitting just a little too tight, but I think it'll break itself in.

4. The top needs to have a plug that fits into the drinking hole so it will not leak fluid into the cap when closed.

Kind of like the rim, these need to be the right height and diameter so the plug will fit the hole. The centers of the hole and plug need to be the same distance from the center (or rim) so they will fit together when the top rotates down on its hinge. This fit contributes to the total friction holding it closed. There's a chamfer on the plug so it will guide itself in. So far so good. Sometimes it needs some filing to smooth it out after printing. 

5. The flip-up part (top) needs to be hinged to the screw-on part (bottom).

This has been challenging and strongly influences the overall design and printing orientation. I'm now on my third hinge idea. I wanted to figure it out without minimal influence from others's design.

1. Many such objects use a "living hinge" which is basically a thin section of the part which will flex without breaking. I printed a couple trials, then I found some (often-repeated) guidance on living hinge dimensions. I tried a couple variations of this fastened to the top of the bottom and the bottom of the top. Again I used slicing to print just the relevant section without the full depth of the bottom and top. It worked... kind of, with a couple of problems. PLA is just too stiff for this to work well. It fractures instead of flexing. My samples would break after about 300 bends (150 open-close cycles).
   
2. I then designed a separate living hinge with tabs that would fit into slots on the top and bottom. This tried to solve two problems: the hinge would be replaceable after it inevitable broke. And it would allow the top and bottom to be printed separately, which would solve a couple of other problems. This kind of worked but required some post-print filing to get the tabs to insert easily.
   
   
   
3. About that time I saw someone's design for an object with what I'll call a "ball and socket hinge". I think of it as a bump hinge. The bottom has hemispherical bumps on the ends of a long piece. The top has hemispherical holes in arms. The bumps and holes differ in size by a tiny margin. When printed, the bumps are in the holes but (hopefully) separated from it by a tiny tolerance. (3D printing has problems with spherical shapes due to the overhang issue, but when they are this tiny that's not a factor.) If it all works, it will come off the printer as a working hinge. In practice there has been a tiny amount of fusion that easily breaks loose. Nice!

Except... I just opened it forcefully and one of the holes broke out. PLA is brittle when small. I'll need to adjust it to give it a little more material around the holes. Nylon is looking better and better...








Now some thoughts about the printing orientation:

• If the bottom is printed right-side-up, it needs support material because the big "deck" that seals the bottle and supports the drinking hole is a big bridge.
• If the bottom is printed upside-down, the drinking spout would cause the deck to be slightly off the bed. Ugly support problems.
• If the top is printed right-side up, the deck supporting the plug would need thin support. Yuck.
• If the top is printed upside-down, it works out best. The top-of-the-top would be on the bed and so be nice and smooth. The deck would need support, but that turns out to be completely internal to the cylinder.

So the chosen solution is to print the bottom right-side up, with the top upside-down (open position), hinged together. The only downside of this is that to print it, the top needs to be be as tall as the bottom, which it otherwise would not need to be. That makes the whole assembly a little taller than aesthetically ideal. Oh well, I'll live with it.

I'm thinking to make the top not so tall-and-ugly by embossing my initial onto the top of it, in a fat font which will not create a lot of overhang. That will kind of visually break it up. And I'm thinking of printing the initial in a different color to add interest. The latest version of Mattercontrol enables you to pause at a specific layer to change filaments, so this might work out well. The bottom of the bottom would also be printed in that other color, which would give the whole thing a three-stripe appearance which might be cool.

My next steps include:

• Figure out the gasket issue, and/or adjust the bottom depth, and/or switch to all Nylon.
• Maybe adjust the hinge size for strength. I might reverse the parts so the bumps are on the arms and the holes in the "shaft", which would shift the strain to the heftier part. But in Nylon it might flex and be more resilient, which might solve the problem.
• I'd like to cut a little depression in the front where the top and bottom come together, to give one's thumb a little purchase when flipping the top open. It can't be too deep... I might need to beef up the inside a little to allow it to intrude on the cylinder wall. Still thinking about this part.
• Then figure out the initial on the top of the top.

Finished: Table corner protectors

Here's the complete set of four tabletop corner protectors I described here. It'll probably be several months before I actually get to try them out, because the table's in use for craft work instead of playing pool.

Work in progress: Table corner protectors

We have a removable table that we put on our pool table when we want to use it for food service or craft work. I recently re-covered it with a padded plastic material because the old stuff had shredded over the years, mainly when it was dragged away and stored in the garage. I don't want the new material to be ruined right away, so I decided to 3D print some removable corner protectors for when it's not in use.

I had made a much smaller protector for a metal edge on my workbench and thought since it was parametrically designed I could reuse the file. Unfortunately the previous design assumed equal thickness on each leg. Due to the table top being effectively "upholstered", the thicknesses vary, and so my first attempt wouldn't stick on one side, and crimped the plastic on the other. A millimeter or two difference is needed between the legs. Fortunately I had recently bought an electronic caliper which made this all so much easier!

I ended up redesigning the whole thing with more variables. I also found a subtlety of CAD design. I had done two simple extrusions for the two legs. Because the resulting file is the union of all the bodies, the thickness of the overlapping center section was always the narrower of the two (i.e. the largest body in space), so it was never going to fit. I had watched a tutorial on Autodesk's web site showing how there are multiple ways to accomplish the same geometry, so I redesigned it in three bodies with a subtracted channel instead of a single extruded profile. My third try fit perfectly!

I realized that all four corners would be different, so I'd need to mark them. Rather than write on them with a Sharpie, I decided to learn how to use Fusion 360's text feature. I did reverse-embossed (subtracted) lettering. Cool! This one is only cut halfway, but I rather like having them cut all the way through, so I'll do the other three that way.

These are some of the largest objects I have printed so far. At low quality, with 20% infill, this one took over 3 hours. (I've had longer prints on other things, but at higher resolution.)

A stand for my Robo3D R1+

For a while I had the printer on my desk, but I needed the space back. After quite a bit of searching I found this little 5-drawer cabinet that is the perfect size, about an inch bigger than the feet of the printer. My tools & supplies all fit in the top drawer.

The manufacturer is Winsome, the product name is Halifax. It's sold by lots of outlets, and the best price and availability I could find was from a furniture dealer named Goedeker's.

Ceiling fan remote control holder

We have ceiling fans in a couple of rooms, but the remote controls lived on the bed or some furniture far from the door, and with no other lights in the room it was pretty inconvenient. So for my first original project I designed a bracket to mount the remote on the wall just inside the door.

This was before I learned about parametric design, so it's all hard-dimensioned in Fusion 360. The first couple of versions were based on combining rectangular components and came out pretty ugly. I had problems getting them to penetrate each other well enough that they would result in good joints when printed.

Then I learned about spline curves and such and totally redesigned it and it looks a lot nicer. I learned about holes and countersinking, and centering and copying features. One remaining issue is that there's a "bridge" between the two sides where the back joins in, and the filament droops a bit. But it's on the bottom and the back, and the remote hides it completely, so I just left it that way.

It turned out the remote in another room (same manufacturer, different model) fits nicely in it, so I printed a couple more for that room. Nice to be able to just reprint them on demand!

I've shared this design on Thingiverse here.

No, I didn't print it in blue for actual use. This is a trial print!

iPhone 7 Megaphone Base

I got the idea for this from a giveaway iPhone base made of rubber that I saw briefly. The idea was cool, but it was so small it didn't have much effect. This seemed like a good first moderately complex project.

This was my first "parametric" design with Fusion 360. The upper part is parameterized so it can be adjusted easily to deal with the wide variety of covers that can be applied to iPhones. The lower part is fixed because it was really complex to design.

The first version had two megaphones, one for each phone speaker. I designed the right half and then mirrored it. Because of the complexity I printed the right half a few times. I got the idea to print it at 1/2 size so I could more quickly evaluate certain aspects. That means 1/8 the total volume, 1/8 the print time.








It came out OK but was a little too tight on my phone, and the sound amplification was not that great.



















Because of the concavity of the megaphone it needed to be printed on its back. I have a hole in the back to allow for the charging cord to come through, and that meant a little support material was needed. This was my first experience with support. It worked out pretty well, since it was all in the back where it couldn't be seen, so little to no cleanup was needed.

I reworked it to use a single big megaphone. This improved the sound amplification and in some ways simplified the design. It also made it bigger and so a little more stable for the phone to stand in.

I finished this in late January, about 3 weeks after getting started with printing and design.

Later I shared the STL and Fusion 360 files here on Thingiverse.