Saturday, June 30, 2018

Dungeons & Dragons dice in wood

One of my girls asked if I could make her a set of Dungeons & Dragons dice. Not knowing anything about them, I sent her shopping on Thingiverse and she settled on this set designed by JakeVav. I printed it in Hatchbox Wood PLA hybrid so I could make it look antique-y.

No, I didn't print them all at once, this is a reunion photo. I tried to print them in groups of 2 or 3 to save setup and print time, but I found that as the extruder jumped from one die to another the filament sagged out of the nozzle and left a little "worm" on the next die. Although I experimented with retraction, coast and wipe settings, the wood filament is just more prone to this side effect. So I reprinted the last few as singletons and they came out much cleaner.

After some cleanup I treated them all with wood stain (submerged for several hours) and added a couple very thin coats of Varathane. They came out pretty darn wood-like, especially the triangular one with some color variation that looks like grain.








Here's the whole set ready to give away.
There were a few problems. Lettering can fail to come out clearly if it is too small. And there were a few "slumps" especially in the overhanging faces (I did not use any support structures).

The 10 face was down to the bed, so the 90 was an overhang. You can see the edge between 90 and 40 came out kind of ragged. But in a project like this that is supposed to look ancient and magical and mysterious, I figured a little irregularity gives it character.



Same here, the 1 was on the bed and so the 7 was an overhang.

In many cases I had to clean the letters out with a very tiny Dremel bit.












Another issue is that the faces on the bed come out glassy-smooth. I sanded them in a single direction to give them some "grain" for the stain to penetrate. The 10 and the 1 above came out pretty well. I should have sanded his 1-2-4 face more deeply - it came out more smooth and yellow than all the rest.






At first I tried to use some GP3D wood filament I have on hand that is a bit lighter in color. But I could not get it to stick to the bed, even though I was using the wonderful Magigoo coating. I was trying to print one of the smaller dice, and it was just not enough surface area to stick to the glass. I gave up on that pretty quickly and switched to the Hatchbox. This is the second project in which I've given up on the GP3D - see my trophy base project. I don't think I'll try it again.

Evil thought: (only after it was all done)  I've been using Simplify3D software for the last 7 months. One of its features is the ability to print different levels of a model with different settings. I printed this with 20% infill. But it would be very easy to print the lower layers with a much denser infill, which would essentially make them "loaded" dice. The bottom face would be much much heavier, making it slightly more likely to land on the bottom, and making the opposite face more likely to come up. So... if you're playing D&D or any other dice game with 3D-printed dice, watch out!

Friday, June 22, 2018

Multicolored folding robot

I saw this video by 3D Printing Nerd and figured I'd give it a try because I'm interested in clever techniques. A single-extruder 3D printer like mine can only print in layers upward (with a few exceptions) and so color changes are limited to horizontal layers. Picture a layer cake with chocolate on the bottom and white on top and frosting in between. I've done a few projects with multiple colors or different filaments, like this trophy base.






These folks at Fab 365 have designed robot models which are opened up into boxes, the sides joined by living hinges. After printing it all flattened out you fold the body into a box, which wraps the colored layers onto all four sides.













By carefully choosing the layers at which to change colors you can highlight parts on each side of the body with contrasts that make them "pop". I chose to do four colors in seven layers.








Other parts that stick up like the arms and legs and head also get the multiple colors, and they end up mostly vertical.

The shoulder parts, the frame on the front, and the jet pack mount on the back are all the same height, so they all get the orange color. The little green controls on the front are just a few layers thick and contrast well with the white background.

The feet and the fingers are about the same height, so they both came out blue. So did the top of the head, unintentionally I don't know if the model was actually designed with multilayer in mind... I think not because some of the arm and leg segments are not horizontal, so the colors cut across them at an angle.

This is all previewed in the slicer software to see exactly where the different features start, and each different colored layer is given its own segment. I needed to pause the print at the start of each layer. I found some G-code on line which was supposed to do that for me: pause the print, raise the head and move it to the side so no plastic will drip on the model while I change filament. It actually didn't work except after the first layer, so I had to manually pause it for each change. Later I found out how I had implemented it in the wrong place - better luck next time.  So I'm pretty amazed that I was able to get a good result in just one try! Now that I know how this works I could use it in my own designs.





The result is quite a patchwork of color. It's Robbie the Robot from the movie "Forbidden Planet"!

The living hinges were a little weak, and the "tab A into slot B" design that was supposed to hold it together were not sized right. So I used Super Glue and it all came out fine.








As a bonus, the design includes movable arms and legs. The shoulders actually each include two nested ball-and-socket joints, so the arms can raise and lower, and also rotate, giving them an impressive range of motion. (But on my print, one joint fused and can only rotate.)  I'll remember this joint idea for some of my own designs.

Wednesday, June 20, 2018

Apple device charging stand


My daughter asked me to print this charging base for Apple iPhone, Watch, and Airbuds. It's on Thingiverse and designed by Raphael Barthe.











There are some cable tracks on the bottom and they tunnel up through the stand to reach each device holder.












I chose to print the version with the watch shelf separated to be glued on later. This avoids the support issue. 





I printed it in white eSUN PLA. It came out great! The second time. The first time, the filament was crossed under itself on the spool which led to a tangle.


Sunday, April 8, 2018

3D Toy from a Drawing

I saw a new video by I Like to Make Stuff called Turning a drawing into a toy using 3d printing! About that time my nephew send me his Flat Stanley image. This is a school project in which kids send a picture of themselves around the country or world and have people take pictures of the Flat Person (hopefully) having fun in various situations. I took Flat Lukas to work, to Disneyland, to church, to the dentist etc. In the Flat Stanley books, he was flattened by a steamroller or a bulletin board. Eventually he is returned to 3D-hood by someone with a tire pump. I figured in the 21st century maybe Flat Lukas could be transformed by a 3D printer, so that video was timely.

I imported his artwork, drew sketches around it, broke it into parts for moveable arms, legs and head, made each part thick and rounded, and added pegs and holes so it could fit together. It turned out to be 10 parts, which I printed in 4 colors. The finished "action figure" is about 8 inches tall.

To get parts in skin tone, I used some GP3D Wood PLA that I had on hand from my trophy project. It came out OK, but that material is a pain to print and I'll probably never use up the remaining material. I had to print the head 4 times. This stuff thoroughly clogged my print head. I had never had to remove the "hot end" before but I did this time! It clogged a second time and I was able to unclog it with a "hot pull" and a *lot* of pressure.

Sticking parts to the print bed is a tricky business. Sometimes they don't stick well enough, and sometimes they won't come loose. See this post about damaging the glass bed of my printer. I read some good reviews of a product called Magigoo so I ordered some. It worked really well with most of the parts, but I also had some issues with the Z height preventing adhesion. The instructions and reviews are unclear as to how often one needs to apply it, that is how long it will last. Some people say they reapply after 100 prints - that seems extreme. I found it is sometimes hard to tell if it has applied or how thickly... need to get the light just right. But all in all I like it.

Here's a summary video of the design process in Fusion 360, using its timeline playback feature. Ssshhh - it's a surprise! I'll be shipping 3D Lukas to him soon. I'm posting the video here so he and his family can watch the video when the package arrives.





Saturday, February 10, 2018

Cable holder in magenta PET-G, but at a price

One of my daughters saw the phone cord holder I had made (designed by Lay3rWorks) and wanted a smaller version. Fortunately it was provided as a Fusion 360 file so I could import it and make changes. Although it was a parametric design, it wasn't fully automatic so I had to manually split the parts and recombine them. Not a very big deal.

She liked the magenta PET-G I have (by eSUN) so that's what I used. The first one had a problem: I had used a 15% infill and the top deck did not bridge that nicely, so it ended up with holes. Usable, but ugly. It also had a moderate amount of stringing between the upright dividers. And there were a couple of dark patches, like scorched plastic.

So I put some more thought into the settings. Now that I'm using Simplify3D I have more options.

  • I set the base to 50% infill both to fix the upper deck, and to give the product more weight and a lower center of gravity. The upright dividers are joined to the base by a pretty narrow rectangle, and I was a bit worried about it breaking in the long term, so I set several layers around the joints to be 100% solid. Then the rest of the layers (most of the dividers) back to 15%.
  • I had been using 245C for all my PET-G prints so far, and that's in the upper end of the recommended range. That can cause stringing. So I varied the temperature in the different sections of the print. The base was still 245C for good adhesion. I reduced the joint area to 240C, and the dividers to 235C to try to reduce stringing and scorching. The bed temp remained at 85C throughout.
Here's the result:

I was away for a while after the second print finished, so the print had time to cool on the bed. It popped off easily by hand, no razor blade needed. I've found that PET-G usually comes off nicely after it cools.

But the print had some weird lumps on the bottom, in a slightly different color. Chips of glass, stuck very tightly! I had to use a perpendicular X-acto knife to clean them off.
So now my print bed is chipped, right in the middle. Two big chips and a little one. Plus a scratch.

I imagine I've set the stage for the chips by using a putty knife to pry or tap off other prints that have been very stuck. PLA can do that. Often PLA will self-unstick if I let it cool thoroughly, but I've learned that only recently. And in trial-and-error printing, it's inconvenient to wait 30 minutes or so to remove a trial print.  So I imagine I had caused some tiny chips. Repeated heat stress probably caused tiny cracks to propagate. With PET-G I use a much higher bet temp, 85C. But I hadn't done that since in two months, so maybe the glass was damaged during that time and was prone to chip the next time it got really hot?

It seems unlikely that the PET-G just really stuck in those spots and caused the break, since I did not apply a lot of pressure to remove the print. I'll have to do some research to see if this is common. Will these chips now propagate worse? Will that scratch turn into a crack? I know from limited experience that that's how we cut glass: scratch a line with a tool and then stress it. I'll do some test prints, and will have to avoid the center of the bed.

I could just buy a replacement bed from Robo3D, but I think there are options. I have seen mention of some removable, flexible bed materials which make prints pop off more easily. So I'll do some reading before buying.


Saturday, January 27, 2018

VR adapter for OWL Stereoscopic Viewer

Background

This project requires some explanation. For Christmas I received a copy of Queen In 3-D, the new book by Queen's guitarist Dr. Brian May. It seems he carried stereoscopic cameras throughout Queen's career and took a bunch of photos backstage and out and about. It's a terrific book. It comes with the OWL viewer, produced by Brian's London Stereoscopic Company.

LSC also developed an adapter which attaches a smartphone to the OWL and enables you to watch virtual reality videos. It's like a nicer version of the Google Cardboard viewer. The design is brilliant - very simple and elegant! I wrote to LSC to ask if the adapter is available separately since I already have the OWL, but got no response. Since I have a 3D printer, it should be easy to print a 3D adapter. No one has published one on Thingiverse, and out of respect for Brian's and LSC's intellectual property rights, I won't publish mine either.

Design

It's a simple rectangular plate that slips into the OWL viewer where one would normally put a stereo photo card. On the back I made some circular holders where I mounted some round Neodymium magnets (ceramic magnets are not strong enough). My iPhone already has a magnet on the back that I use with a magnetic mount in my car. So it will just stick right to the plate. I designed it parametrically so I can adjust the size for a snug fit in the OWL.






Printing

I printed a few partial samples to fine-tune the dimensions by a few millimeters. Here is a sample
with large chunks of the back plate cut out to reduce the print time but still give me the full width and height so I can test the fit. I used blue Hatchbox PLA filament for the test prints, and black Shaxon PLA to give a darker background to the videos.




Final Product


















Using it

I was puzzled for a while about what software the smartphone needs to play VR. Then I learned that the YouTube app itself includes a VR mode (intended for Google Cardboard). For VR 360-degree content it displays the left and right halves and responds to motion of the phone to let you look around. For regular videos it simulates 3D by splitting the video into left and right, but can't respond to motion because the video is just a straight-ahead view.

I imagine there are dedicated VR apps for games and simulations etc. but I haven't looked into those yet. I saw a reference on Brian May's web site to some astronomical VR video's he has produced... he is also an astrophysics after all. I'll have to go back and find that link!

Friday, December 29, 2017

Christmas light storage frames

Here is a short and simple project. I bought new Christmas lights this year and decided to make some frames to wrap them around instead of the cruddy old cardboard I had been using. These have "ears" to keep the wires from falling off, and little clips on the sides to grab the plugs.



I recently listened to a blog which said many projects can do with much less infill than we often think. Since I needed to print several of these, I decided to vary the infill and see how it worked out. I went all the way down to 0%, forgetting about the issue of bridging. Since the sides are not that far apart, even with some drooping bridges, the the second and third layers fixed it and it came out with nearly no surface gaps. (For a project like this appearance doesn't really matter anyway.) So as little as 5% infill worked just fine.

Each of these used about 72 cents of PLA, and took about an hour to print. That really doesn't make much sense... I could probably have bought something like this for less... but this was more fun and printed while I was doing other stuff.

Thursday, December 21, 2017

Wood filaments, finishing, Simplify3D

Someone where I work asked if I could make a base for mounting an unusual object as a trophy. I remembered that wood-infused filaments are available, so I said I'd give it a try. It turned into quite a project, and I learned quite a few new things:
  • Printing with wood filament
  • Switching filaments during printing
  • "Engraved" lettering
  • Section analysis in Fusion 360
  • Sanding, staining, varnishing and gluing wood-infused PLA
  • Simplify3D instead of MatterControl

Filaments

Globs that would need to be ground off and sanded.
Because I had a timeline of about 3 weeks, I bought the only wood filament available at my local retail store, GP3D wood light, which says it is 40% wood. It prints a lot like PLA, but it's more brittle and rough: one time the source filament actually broke while I was printing. It tended to produce little globs on the surfaces. My model had "engraved" or negative-space lettering, which I figured would be tricky, so I printed several chunks of my model with the wood, trying to get good smooth results. Time was running short and I was not getting the results I wanted, and I could find no user discussions of the material on line, so I ordered some Hatchbox wood filament that looked very popular.

While waiting for the Hatchbox to arrive, I figured out that I was using too high a temperature, which was causing some material to scorch inside the nozzle and then come out as a hard lump. Once I dropped to the low end of the temperature range, 190C, that seemed to settle down and I started getting smoother prints.

The Hatchbox printed nicely the first time, and I ended up using it for the final product. To be fair, the GP3D probably would have done fine as well with the lower temp and (as discussed below) Simplify3D. The Hatchbox seemed to come out smoother and more plastic-looking; it does not say what its wood percentage is. So maybe I'll give the GP3d another try on another project. Being a more woody texture, the GP3D took stain better.

Switching filaments

Contrast layer samples, between wood layers, after staining
Rather than a monolith of wood, I wanted to insert a contrasting layer to make it a little classier. I printed some samples with glossy black PLA, and also with some rather translucent PET-G in magenta or green. Eventually I decided to go with the black. The contrast layer was just a short ways down from the top deck, so I could "engrave" letters all the way to it, so the letters would have a black background.

Switching filaments is certainly possible, but there is a lot of discussion on the Internet on who to do it well. I learned a few things:
  • By default if you just pause MatterControl, it leaves the head in place. It can ooze filament, and when you insert new filament some needs to come out to get it flowing. This could land on and mess up the previous layer.
  • By default the stepper motors release after 60 seconds, so if you can't switch in that amount of time, the head or bed can move and ruin the print when it resumes. I spoiled one after 4 hours of printing, so I went looking for a solution.
  • I found some custom G-Code to add to the pause and resume actions. I set the pause to move the head up 20 mm and lock the motors for 10 minutes, and set the resume to lower the head back down. This worked well. I also slipped a piece of paper on top of the model to catch the ooze.
  • Pausing is not an instant action. I wanted a clean transition of colors, so I learned to preview the layers to see where the head would be at various times, so I could select an optimal location inside the model, not on a perimeter. The different colors would be hidden in the infill. Depending on the specific situation, sometimes it's in the layer before, and sometimes in the next layer.
The final model was going to take over 6 hours to print. I needed to switch colors after about 4 hours. In one of my test prints, and then in my second attempt at the final model, when I paused to make the switch the head moved up but would not stop extruding! There was nothing I could do to stop it and I had to cancel the print.

Engraved lettering

My design has a couple of "nameplate" surfaces on the sides with lettering cut into them at an angle. One of them includes a company logo which I had to learn how to import into Fusion360. The first tries used lettering way too small, which came out rough and globby, so I went with a larger font and simplified what I wrote (I had the design flexibility to do that). It also has some letters cut into the top surface, all the way to the contrast layer, which would give them a nice dark background.

Section analysis in Fusion 360



To attach the object to the trophy base, I designed a little round pedestal which would fit into a hole on the top. I wanted to see how all the pedestal plug fit, and how close the bottom came to the bottom of the hole so I could glue it in. I found the section analysis option, which was very helpful.





Finishing

The globby lettering needed some cleanup with my Dremel, and some sanding to get rid of the "pimples". Since it's 40% wood, the material sands better than regular PLA does. Regular PLA tends to turn white when sanded. This does a bit too, but not bad. Unfortunately it is still plastic, so when I trimmed some things with an Xacto knife, they tended to stick and remain as fibers, rather than cutting off cleanly as wood would.

The wood filament is pretty light in color. That's nice for some things, but I wanted a richer color for a trophy. I tested some stain that I had on hand. The stain tended to penetrate the grooves between the printed layers, which looked like wood grain. But it did not penetrate much in the smoother areas, even after 30 to 60 minutes, especially where I had sanded. I needed to leave the sanded areas a bit rough if I wanted them to stain.

I had some Varathane on hand, which is really a plastic and not a varnish. I was concerned that its solvent might dissolve the PLA, but it worked just fine.

I knew I would need to glue some of the final parts together, so I looked to see what was compatible with PLA. A very comprehensive article pointed me to Loctite Gel Control. It worked great!

Software

I didn't think that the runaway extrusion was the fault of the printer, but rather of the software. It became apparent that I was going to need to reprint just the upper sections of the model due to the aborted filament switch. I also was not thrilled with the little bumps caused by filament oozing when retracting on moves. The wood filament was not sticking to the bed well, causing wrinkles and flaws. And I was having trouble with the completeness of the top layers, and was futzing with extrusion factors to try to fix it. 

I've been using MatterControl, which came free with my Robo3D printer. Some reading earlier and when trying to solve these issues led me to Simplify3D as a more professional option. Many people have raved about it, so I bought a copy, although it is $150. If there was a chance it was going to help me not foul up a 6-hour print (I only had a few days left for this job), it was worth it.

I printed a test object (in PET-G, since that's what was in the printer at the time) and I was shocked at
how smooth the print came out. The sides and top were silky smooth, no nubs. The lettering was sharp and had nearly no globs. Nice!

For my next try at the full model, I used S3D. The sides and the lettering on the nameplates came out perfectly smooth, dramatically better than what I had achieved before. I saw that there was still a lot of discussion on how to properly change filaments, so instead of attempting that I stopped the print, and printed the contrast layer and top deck separately, and planned to glue them together. S3D's ability to start and stop at certain points is easier to use than MatterControl's, though I wish it were specified in layer numbers rather than measurements.

What a huge difference in print quality! I could see it right from the start, with great adhesion and smoothness in the first layer, and perfect smoothness all through the infill. The lettering was incredible. I spent maybe two minutes cleaning up a couple tiny globs with an Xacto knife, and NO Dremel work at all. I decided I did not need to sand the nameplates at all either, which would let them stain better. As one of the reviewers pointed out, the saving in post-processing time alone justifies switching to Simplify3D. I'm sold! (Though it does have a few quirks...)

The finished product

The first finished iteration was from MatterControl with a lot of cleanup. The customer loved it. "Just one thing... could you move this to there instead?" Um... sure... I'll just do another 6-hour print, stain, varnish...  The second iteration was done with Simplify3D, which required very little cleanup. It came out much smoother, with much sharper lettering, though slightly less woody-looking due to the poorer stain penetration. The customer loved the final result!


Friday, October 27, 2017

Replacement base for a smoke alarm

A smoke alarm in our house is held onto its base by two thin tabs of plastic that eventually broke. I found a design on Thingiverse by buwprinter (from Germany). It was obviously not made for my model - there are zillions out there - but I thought it might work. It was provided as an STL file, not a modifiable design. MatterControl has a scale feature, but it's an overall resize not intended for specific accurate needs. So I had to learn a couple of things in order to resize it accurately in Fusion 360:
  • How to import a file.
  • How to scale a mesh design.
The overall shape and appearance of the base is not that important. What matters is the spacing of the locking tabs that hold the alarm in place. I measured that carefully with calipers and then used the scale ratio feature and hoped for the best.

Also, this was the first time I was switching my printer from a high-temperature plastic (PETG) to a lower temp (PLA). I had heard that this could be an issue, so recently I bought some cleaning filament. It's an odd size, so you don't feed it through with the motor, you push it by hand. I was surprised how much I had to push through before the result was fully clear with no trace of green from the PETG.

I noticed that the PLA was not extruding straight out but rather curling up as if it were sticking on one side. I had to scrape the outsize of the nozzle to get rid of some scorched debris, and also tried some advice about pushing and pulling filament getting out a few times until getting a nice clean "cone" on the end. Seems to be printing OK now.

The diameter of the locking tabs came out perfect, but each tab was a bit too long for the slots. I ground off about 1/8" from each with my Dremel tool and then it locked on just fine. Here's the final result:


Wednesday, October 25, 2017

Jack-o-lanterns


Just for fun since I have orange and green PLA on hand. An LED tea light fits inside. I downloaded this from Thingiverse, provided by mb20music.

Lamp repair - again - with metal and PET-G

There was just too much stress on the plastic threaded tube, and the beefier version broke in the same place again. That vertical tube needs to be metal, so I shopped on line and found some 6" long nipples of the same diameter and thread as the original one. That meant that the "base" could be simplified. I decided to stop using a captive metal nut but to build threads right into it. That meant no more support to trim out!

About that time I saw an ad for PET-G filament, describing it as just a tough as ABS but without the fume problem. It is translucent, and comes in some nice rich colors. I ordered some red and green, thinking I might make some Christmas ornaments. PET-G requires hotter extruder and bed temperatures, and writers said it can be finicky.

I read up on how to reduce stringing and - I guess you could call it "globbing" - where oozing plastic goes where it's not supposed to. That led me to reading about calibrating my printer's feed speed. Huh? I didn't know I needed to do that. There's a process for marking a measured length of filament and telling the software to extrude that length, and then checking the accuracy. Mine was off quite a bit, feeding 92.5mm when 100mm was called for. I'm not sure what effects that may have had on my printing so far, but I went through the steps and corrected it through a steps-per-mm setting in the firmware. Test prints gave good results with the PET-G.

The part wasn't quite working, though. There was no good way to keep the top wedge in place in the tube while turning the screw to pull it down. Back to the electronic drawing board. I realized I needed a way for the base to keep the side wedges in place, and for them in turn to keep the top wedge in place. For that to work and still allow the side wedges to slide out and up/down, I designed a tab-and-slot arrangement:

Two wedges sit on the base with the tabs in the slots, free to move but not rotate. Then the top wedge fits between the side wedges, free to move down on the screw but not free to rotate. As long as I hold the base in place and tighten the screw, it should all stay aligned as it expands.
Only problem is, when I created threads of the right size with Fusion 360, the part was too tight on the screw. I fought this quite a while before figuring out I could expand the threads slightly. With the help of a video from Autodesk or someone, I learned to push back three parts of the four faces that make up threads. A little trial end error led me to expand by 0.6mm, which makes for a smooth-turning thread with no slop.

Here is the final assembly ready to insert into the lamp tube. The "top wedge" on the left side is threaded, and the "base" on the right is not, it sits against the bottom of the tube and the nipple pulls down through it. The wedges sit loosely between the two. The nut on the right is just to hold it all together temporarily. It worked great. Holding the base stationary and turning the nipple (with two nuts locked on the threads) locked it in place with no trouble at all. Then I removed the nuts, put the big lamp base and washer and nuts on the nipple, and tightened it all as much as I dared. It feels very secure.

I initially printed some of the parts at 99% infill to make them as tough as possible. But I needed to do a couple trial prints, and went to 10% for speed. Well, the PET-G is *very* tough. I bent and pushed on it and it's quite rigid. So I used the final test print at 10%.

Another great thing about PET-G: if I let it cool to room temperature, it self-detatches from the bed! It just pops right off. PLA never does that. Yet during the print it sticks really well even with a light, leftover amount of hairspray. No curling, no coming loose. Really nice.

So I've put the lamp back in service again - hopefully for the last time.

Sunday, October 1, 2017

Second try at the pole lamp connector


Just as with software development, I can't always anticipate everything and turn out a perfect design the first time. Fortunately the low cost of 3D printing materials make it feasible to try several versions as I refine a design. I'll be honest on this blog about what works and what doesn't.

The first version of this repair insert didn't work because there was no good way to fasten the PVC pipe into the pole. Glue didn't set up around the shims. Because the first inch or so at the bottom of the pole is a smaller diameter than the rest, I can't just insert a solid pipe or something - anything that can get past that first constriction will slop around above it.

I was working on ideas based on a "toggle bolt" concept, when I found on Thingiverse someone's design for an expanding part kind of like what bicycle handlebars sometimes use to connect to their stem: wedges that move outward as a screw pulls them together. One problem with that idea is that it goes off center as it expands, which would not be good for a pole that needs to stand up straight. Also the center screw needs to be hollow to allow the lamp cord to pass through. So I got to work on my own design.

In this pic the black part is what sits at the bottom of the pole. I added a flange below the pole, sitting on the top of the base, to give it some more stability, and since that's going to be visible I printed it in black. It still has a place to hold the nut for the metal nipple that extends into the base. That's what takes a lot of the stress, so I didn't want to make that out of PLA, I'll continue to use the original metal nipple. Now the stem that goes upward is threaded so it can engage the wedge-shaped nut. As that is tightened, it will pull the two-sided wedge down between the two loose wedges, forcing them apart, gripping the inside of the tube.

The threads on the two plastic parts were quite tight and needed some cleaning out with the metal nipple and a nut, and soap to smooth the way. That's OK, I don't want it to slop around.

Not shown is a slot (kerf) in the bottom of the flange in case I needed to turn it with a tool, but I never used it.

Inserting the wedge into the pole.




Slipping the loose wedges in at the same time. I needed to expand the wedge a little at a time as it went in, because there was no way to hold the main wedge/nut once it was in. Carefully adding friction  seemed to work.

Unfortunately after I added the base (which is about 5 pounds) and stood it up, it soon broke as I tried to move it around on the floor. Back to the drawing board! The design seemed sound, but the parts must have had some weak spots.

The threaded tube broke right above where it connects to the main insert above the nut. I identified two aspects that I could improve:

1. Looking at it through a magnifying glass, I realized that the threaded tube was infilled, i.e. not solid plastic. That's how most 3D parts are printed by default, and I had set it for 60% infill (or 40% empty space). For the next iteration I made the whole black part 100% solid. I also added a chamfer around the base, to give it a little buttress in case it wanted to flex. That meant the side wedges needed a little more clearance, so I expanded their center cutout a little bit.

2. I speculated that the wedges were not seated down on the top of the black cylinder. If so, that would allow the leverage of the pole to flex the threaded tube from side to side. If the wedges were to seat down firmly on the cylinder, it should be rigid and resist that bending moment. If the point of the wedge-nut bottomed out before the wedges were tight, it would not be as strong as it could be. So I truncated the point of the wedge - it's noticeably shorter and rounder in this pic. By ensuring it would force the wedges outward, and by starting with it screwed down a little further, I hoped to ensure they would seat all the way down.

That all went really well. I inserted and tightened it by hand as much as I possibly could. It seemed to get really tight and never approached stripping the plastic threads - they're quite long and tight, so that's not really much of a problem.




Here's the repaired lamp standing upright. It wobbles just a little bit - that's nearly six feet of leverage on the bottom joint. I speculate that if someone were to drag it sideways by grabbing the top of the pole, it could still break the threaded part - it is hollow after all, to allow for the cord. But as long as we lift it up vertically if we need to move it, I think it will be OK.

Saturday, September 9, 2017

Insert for repairing a pole lamp

We have a nice Tiffany-style pole lamp, but the bottom of the pole started to break, so it wobbled on the base. (Simple metal fatigue; it is probably 30 years old or more. I decided to make an insert for the base screw to fit into. The inside of the tube is an odd shape, smaller in diameter at the mouth, making it hard to use a single piece of pipe or something without it wobbling. My solution is a piece of PVC pipe wedged and glued in place up the tube, and a 3D printed piece that fits snugly in the mouth, slides 2.5 inches up the PVC pipe for stability, and captures the nut securely. Oh yeah, with a hole through the center for the cord to go through. Not a glamorous project, just a fix-it job that needs a custom part one that no one sells.

I printed it in PLA with 60% infill for strength, which made the upper tube essentially solid. The nut slides into a hole that is a hex shape on four sides.




The part with the captured nut and the base nipple. Why orange? 'Cuz that's what was in the printer.



Inserted snugly into the PVC - I needed to sand down the little PLA bumps to get it to fit.



Going into the pole. Actually I wedged some split dowels around the PVC to center it and stabilize it in the pole. I inserted Liquid Nails glue as I assembled it.



All assembled. I'll let the glue dry until tomorrow. When I put the base on we'll see how well it holds up.