A Unique 3D Printed Screw Connection 

I've been teaching my friend 3D design using OnShape (I'm Fusion user myself) and it's fun to confuse him by showing at least 3 different ways to do the same thing :) First drawing rough shapes and then contraining them completely is something I do as well. Seemed like a good way to me.

It's an extremely handy way of working -- except when it isn't. Add it to your toolbox and you'll learn by doing when it's appropriate and when not.

I think the draw first, constrain later is the old school method of using CAD programs and often used by experienced CAD users because it makes it more apparent which constraints you need to add. That method uses drawing shapes intentionally in wrong places (that is, you intentionally draw the shape bit offset and intentionally a bit wrong size). Combined with parameter table to define all the measurements you definitely get witchcraft level automation.

Some people are sticklers for constraints, some people just wing it, I'm probably somewhere between. I think constraining when you need to do it is a good compromise, it really depends on what you're doing and what the plans are for moding it later on. I made a fully parametric car holder, everything needed a named parameter to do it right.

@@ozzieenkees LOL, you got it, I suppose I should probably do things by-the-book for the videos, sometimes I get an idea and I just go where it takes me. They're really more about the ideas, but I will have more videos coming which are focused more on design and in those cases I'll try to provide a little better guidance.

Totally agree. That was a decision that sometimes people just want to show it correctly instead to invite you to sit around and enjoy the journey

is Chinese

Stop splayed scarf joints appear in ancient Roman structures, and I am pretty sure that other form of scarf joints appeared even earlier in boat building. I believe that the problem of having only short planks available and needing longer planks is an old one, and several civilizations came out with similar solutions. The Japanese joints try to avoid using metal fittings whenever possible (moreso than the Chinese joinery, because China had more access to metal ores), but "trennels" are nothing new, and using pegs to hold joints together is also not exclusive to the far east.

​@@TDOBrandano there are old woodworking schools in Scandinavia and Slavic countries that pride themselves on not using any nails or metal while constructing huge churches that have joinery in no way inferior to the Japanese one. And that's just one example.

The Chinese taught the Japanese.

Still doesn't beat the speed of a nail gun... And the ironic thing is that residential homes in Japan are made cheaply because regions are so earthquake prone there's no reason to build a house that will last 100 years because it probably won't. Meanwhile in Germany, roofs are constructed to last 100 years, unlike the typical 30 years in the US

(I don't have a soldering iron so unless there's an obstacle I'm unaware of,) I believe a quick, jank way to gain more control would be to replace the soldering tip with a bolt of appropriate size.

​@@krayne2084good idea. But you want to pull out the soldering iron right after the heat set inserts is placed. Every second will heat up the plastic even more and potentially ruin it.

I bought a kit like this but I've already lost 3 or 4 pieces. And honestly I didn't think it was that practical. But a "new" technique that I saw on the Geek Detour channel where you use a screw or a table to align, that was a golden tip.

The insert inserters are wonderful, especially the CNC Kitchen one. A bolt won’t work well because steel does not transmit heat nearly as well as aluminum or brass.

The one I used in the video was one that I modified to fit the M3, it was made from one of the extra tips that came with the iron, it was too short of course, but it's just made from copper and plated, so I can easily whip another one up when I get my lathe put back together. The size is important for sure, the tip expands when heated and it can get hung up inside the hole giving you an awful experience when trying to put these in, I ended up doing some trial and error to get just the right size that worked at 300C.

Absolutely, I think if I could extend those waves into the other body a bit more, it's be even better. I'll add a glue-test to see how they compare as well, I'm really curious

Thanks Tim, I really appreciate that. I want to try to share the ideas and also show how to create them in as short a video as possible, I still have more work to do to make the videos better. Baby steps I suppose.

Oh, I guess that's a good thing. I think I still have a long ways to go to get more comfortable speaking into a camera. It's not like speaking to a person that's for sure.

@@NeedItMakeIt but it's already calm and clear. I think it's fine.

thats good mic helps ALOT ;)

Wow, thank you very much! I still have a long ways to go, I want to do even better in every way possible, I'm still getting comfortable speaking to the camera, it's not a normal thing.

@@NeedItMakeIt I have learned so much from watching your videos, not only that some of the files you even give away for free you can make a few bucks on but you choose not to, and you're growing the channel the proper way not by begging for money every few seconds in every video they're consistent, educational, useful and just plain out excellent I look forward to your videos.........................

was thinking that to, something so simple made it look so complicated.

You got it, I thought about the 2nd more complex shape and whether it could be done another way, and this seemed like the best way. doing 5 sketches would probably give a pretty good result as well, technically I could also have mirrored the surface to save a bit more time too. Lots of options, it's going to open up some options for sure down the line with other designs.

I took it a bit far, there isn't a need for that many sketches, I just like to take things too far... it's a problem that I have.

Thanks, I can't take much credit here, I've found that when printing PETG-CF on either the Bambu X1C or the QIDI Q1 Pro produces results like this, the Bambu auto-calibrates... well it normally does when using the correct buildplate. The QIDI needs the manual calibration, but if you stick with the same filament, it seems to be a one-and-done. The only issue with PETG-CF or any filament with CF in it is that layer bonding is dramatically reduced, it's a stiff material, but it acts more like a piece of wood that wants to split along the layer lines more easily. Basic PETG may be more flexible, but it has far better layer bonding.

May I ask for a link to the filament?

He could even make the hole smaller towards the side the screw comes in so the insert really has something to get pulled against in the hole.

That would work, but only if you weren't concerned about the finish of the opposite side, as implied in the video.

@@markbooth3066 The opposite side could even be closed. With 3D printers you can print the whole, interrupt the print, drop in the nut and continue printing so the nut is captured securely inside.

That is technically correct@@AlJay0032, which is, of course, the best kind of correct. *8')

I hear ya! I'm starting to finally wrap my head around it, but it is a bit of a shift, I've seen some pretty incredible designs that have used surfaces, and that was pretty much what inspired me to try it out. There are some parts that are just too hard to model using the solid options alone; curves in many directions seem to be where the surfaces really shine.

Thanks for that! I'm trying to include some Fusion stuff more often and present it so that I can use my hands more often, I think I'm going to need to use a VR Headset so I can present directly inside the program. I'm still working on trying to come up with the best approach that is clear and quick.

Well if that's the case I have 100s more of these kinds of videos to make! I really want to come up with the best way to present information on-screen, a simple screen capture is pretty plain and I want to use my hands and interact a lot more. I was thinking of a VR setup where I could be right inside the software, I wonder if that's doable?

Thanks for that! I'm working on trying to make better and better videos; more info in less time and hopefully I'll get more comfortable on camera to smile on occasion... Lots of room for improvement, luckily there is no shortage of video ideas.

Yes that would have been really good, I was thinking that I should set all of the connections at 45 degrees for the testing I'll have to do a test run to see how each connection will turn out when printed, some will print better than other's I'm sure. I also will switch to Basic PETG because layer bonding is far better than anything with CF or GF in it.

A pull test would be good for sure, I'll try to integrate that into my testing somehow!

Thanks! Hopefully they're a good combination of info and entertainment. I know a lot of people want more Fusion tutorials, I'm trying to come up with a better method to present these in as short a video as possible while also having a good amount of interesting information. Simple screen captures aren't good enough, so more work is needed to get something better.

@@NeedItMakeIt I love the concept of improving and fine-tuning cool wood joinery for 3D printing. Personally, I'm not too concerned with tutorials in a specific program.

Several wood joint designs don't need screws and you could copy those. But I think this would need a lot of tuning to get a perfect fit because the tolerances vary between printers and print profiles etc, there's a reason people use screws. I bought a bunch of M4 nuts and bolts, copy-paste the cutouts and holes into the part and it just works every time

​@@SaHaRaSquadIdk, like those big locking screw things Ikea is fond of, could you create printed in place plugs or geometry that you then tighten with a screwdriver that pulls two pieces together and locks into place?

That's right, it's not going to use the same amount of material, just as a scarf will also use mor material than a butt-joint would. I can certainly speak to this when I get closer to finalizing the testing video. It's not going to be a perfect apples to apples, it's more about sharing interesting ideas. Valid point though and I will make a note to speak to this in the follow up video.

I'm glad you thought so, thanks for that!

i like your comment, in mine i suggested designing it so that hole goes all the way through, so you can insert the heat set insert from the back side, that way you are pulling against the plastic on the back side of the heat set insert, not just pulling the insert directly out of the surface of the print. essentially clamping both prints between the heat set insert and the head of the machine screw. or even replace the insert with an m3 nut utilizing the same idea, since the nut may have better surface area on its front face.

@@BrandonTenbrink I agree, that's a better way to use heatset inserts since you can get a much higher clamping force for the screw. I'm sure there's formulas that can dictate what circumstances each configuration will come out on top. I think for this kind of joint you may also be considering the aesthetic appeal of it too, which might lend towards a blind hole setup. For m3 nuts, I find them to be very weak for this application, as there isn't much of a good surface resisting the force of spinning that's actually making sure you can apply torque to it. This is why I prefer using square nuts in those circumstances!

This is not really true. For one thing, the insert is melted into the plastic - unlike a screw which sits on the threads. There is far more contact between the insert and the item, and the plastic itself sticks to the metal as well. Also, if you install the insert correctly then it will typically have a small bulge / lip of plastic over the top that resists pullout. And inserts have far better wear resistance if it's something that will be taken apart and reassembled.

@@skipfred Go ahead and test it yourself, I've done a ton of testing wrt using printed/cut threads vs using inserts. If you look at what I said, I considerably lengthened the amount of plastic that you'd need for the direct thread method. This reflects the difference in surface area per axis length between the two methods. I haven't worked out the ratio of surface area to diameter for a screw or an insert, but in general I've found you only need a printed thread with 2x the axis length of the insert in order for an M4 or larger to be superior to the insert. Of course there are different insert styles, and I'm using the double row chevron style, similar to the video. Spool style inserts are straight up garbage and should be avoided. I wouldn't use printed or cut threads in plastic if the screws are under M4 or 6-32. Too small of features for the polymer to retain strength after insertion. You also need to consider that only the portions of knurling that are opposing the pulling force are really contributing to the holding force of the heatset insert. Yes you get some extra from the thermal bonding between plastic and brass, but brass is a pretty low surface energy material to begin with so it's not like that's going to do a lot. Depending on the polymer, you are also quite likely to overheat the plastic that is directly in contact with the heatset insert, greatly degrading the polymer structure, and making a severe weak point. While in many cases this doesn't matter, it sure does for pullout strength. What the other gentleman suggested (inserting the heatset insert from the opposite side of the screw) is far superior to any small bulge or lip of plastic over the top that resists pullout. Additionally, I'd argue that if you have a bulge from a heatset insert, you have performed the installation incorrectly, as most are designed to be flush or subflush mounted. Wear resistance is a good consideration, but I find the cheapo brass to be a bit cheesy. There have been plenty of tests comparing cut wood threads to insert threads, but I haven't done any longevity tests between cut or printed threads in plastic and insert threads. Keep in mind you only have a few mm of actual thread engagement with the inserts.

Is that a typical printed part? Most of my parts are quite massive. I'm sure everyone has completely different parts and I won't pretend to be able to predict what people want to design and print.

Oh, that's a cool idea! That does give me an idea, I could make the connections 3D printed and have the rods made from wood. Most of the forces would be compressive it could be quite a strong frame.

I was just thinking that with an appropriate layer height, and slicing orientation, you may be able to use the layers themselves as barbs.

@@markbooth3066 No, don't do that, because the layers are planes of cleavage, and putting shear stress on them in the plane will make them split.

I was assuming you'd meet the barbed surfaces at ~45deg angle to each other, so the barbs were at a diagonal to the plane, rather than in the plane, but maybe I'm missing something here.

@@markbooth3066 Don't have the barbs stick to any particular plane.

?

I have them in my hands right now and just tested them again to see if I missed anything, and I don't feel any wobbling. We'll be testing everything anyway to see which ones really come out on top so it doesn't matter what I think, the proof will be in the pudding. I'd prefer to not use any fasteners so if this isn't the best connection, that's fine with me. Most of what I do here on YT is just share ideas they don't need to be good, ideally they're just different from what anyone else has done.

That's a good point, it really depends on print orientation, the other issue that I was concerned about was the severe angle acting more as a lever, again it all depends on the way it is printed where it would be a problem or not. In the example from the video, the print should have been rotated 90 degrees for the best results. Also printing with anything CF or GF will cause a dramatic drop in layer adhesion. It's best to stick with Basic filaments.

Thanks, I don't know it it can beat some of the other connections, but it's an interesting approach and maybe it'll be a useful tool to use for other applications as well. I can already think of a few times where this tool would do a much better job than just using the solids options.

I just suggested that too. Another advantage is if you use something like a pinecil, the insert tip from CNC Kitchen has a pretty thin shaft which would have helped a lot here.

@@conorstewart2214 I looked up the ones from cnc kitchen, I have the "Virtjoule" branded ones off Amazon, they look pretty similar in length though it's hard to tell. I wonder if anyone makes even longer ones for really deep holes

Thank you, many many more to come, I have lots more ideas for design videos too, I also want to explore the latest tech as well a bit more, but overall just sharing ideas and trying to make things ourselves is the goal.

You're very welcome! I definitely needed to plan ahead more on that insert.... I have my metal lathe in pieces while I restore it but after I have it together, I will make a new custom tip.

I guess hell make his live easier if he moves to Holland 😂

One larger bolt or something that can go through completely to keep tension would be great. I like your thought process here, let me see what I can come up with!

I'm curious too and I want to test in Torsion along with bending, I'll test deflection under load and I'll take each to the breaking point. I've tested in the past, but these tests will be better and should focus the force on the connection to make sure that the connection is the weak point... well that's the idea anyway. I'm sure Stephan @ CNC kitchen could take these ideas to the next level!

Well, you sir saving me some time 👌

Same

Appreciated.

Thank you for your service 🫡

Saved me time!

Agreed-it makes changing any of those numbers MUCH easier. If a number gets used across multiple sketches or operations it should be a parameter. Within a sketch, just click the reference number a dimension should equal. Updating the original number will update all references to it.

Yes, better to use parameters for long-term best practice. Sometimes the ideas just flow and I just go with it.

Hey, thanks for that!

Thanks! Another fun idea, and using surfaces opens up some possibilities for us to create more flowing and complex shapes.