On your torque out test, the reason for failure of the bolt head is because you are torquing againts the insert itself, you could try using a longer bolt and thread lock it into the insert and use a washer to provide lateral support for the fastener head and shank.
In my practice most (if not all) of the insert are being pulled out from the opposite end - so "pull-out force" is not important, as printed parts (elements of HEVo) are starting to crackle ominously when screwing then hard :) Also - by inserting M3 bolt from the opposite end while heating insert with soldering iron helps A LOT with keeping them straight. And I used only cheap ones so far.
For me... your channel is the most important one for 3D printing. I say that because you address structural (as opposed to aesthetic) issues, and... you do it in a way that the layman can understand. There are other very technical 3D print channels, but they are useless to me. Their viewers would need to be advanced printers to know what they were even talking about. Thank you for the great work.
Agreed. I'm not a novice and this channel still has the best content. Some of the other popular ones all eventually devolve into "politics" of the 3D printing world - things I have zero interest in.
CNC Kitchen is the best resource channel for 3D printing. A lot of it applies for any printer on the market. I like makers muse is great for fun things to try. Or cool test prints. Bur cnc kitchen is legendary for data being pressnted in an interesting way
If you ever find yourself looking for thorough testing on household items rather than 3D printers, I highly suggest Project Farm. He tests just about everything. Tape, oil, ratchet straps, batteries, drain cleaner, tire sealants, portable tire pumps, nut removers, brush blades, windshield wipers, drywall anchors, the list goes on. Some tests are less scientific than others (the brush blades come to mind, lots of room for human error), but they're more than sufficient for how people actually use them. Plus he never accepts sponsorships.
If that is a possibility, but if it is it will greatly improve performance. I would like to test out if it would be possible to insert them during the print with a pre programmed pause in the g-code for parts where that wouldn't be a possibility, though I think I'll run into problems with the bed level.
Those "cheap" inserts appear to be overmolding inserts, designed to be held in place and have plastic flow around them. When you heat press those, I'm guessing a cavity is formed where the plastic did not flow and create the ledge necessary to mechanically lock into the plastic.
this is 100% what is going on. Those are meant to be in place at the time of molding, not added after the fact. That's why they don't have the lead in chamfer and they push plastic when you try to melt them in. They are not designed for this application, though they will work ok for light loads. When used in their intended application they are equally or more strong than the actual heat set inserts.
I just print the hole smaller than it should be and hold the iron for a bit longer. The only time such a part broke wasn't even in the place of the insert despite it being the place with the most stress applied
Was hoping you would try a pull through test as well as a pull out since that would provide the greatest strength for them all. I.E. insert the insert on one side but screw the bolt into the other. Also, I wonder how well M3 wood inserts would work (if at all) i.ebayimg.com/images/g/2koAAOSwDoNd1PUd/s-l400.jpg
Whenever possible, threaded inserts should be installed on the side of the part opposite from where the fastener is entering it. (Obviously this isn't possible with blind holes) This preferred orientation results in the fastener pulling the insert even more securely into the part. It would be interesting to repeat your pullout tests with your test blocks flipped over. I've found that the plastic part will usually fail catastrophically before the insert is pulled out.
I would think your major advantage to a threaded insert, particularly when you might choose the cheap ones instead of threading directly into plastic, would be for something that doesn't require great strength, but *does* require repeatedly removing and re-installing the fastener. In those conditions, cross-threading and wear would quickly destroy the threads in plastic alone, where the brass insert would happily withstand as many cycles as you'd wish to throw at it without appreciable wear.
Thanks for making this series of videos. I, and probably many other people, find this information invaluable. I’ve always wondered if it was worth buying the more expensive, chamfered, inserts. Your videos are always quick, to the point, and well edited. Well done!
Was just thinking about fitting them into a pocket from the back, The back hole should be sized for the insert but the front sized to just clear the thread thus leaving a shoulder for the insert to pull up to. Love the testing machine by the way
Another potential depending on the design is to make a slot for a regular or square nut to insert them from the side. This essentially embeds the nut in the middle so it is as strong as your piece on pullout. The only issue that I've run into with this (mainly with regular hex nuts, less so with square nuts) is torquing, as the nut may spin through the plastic when tightened as it's only held by 4 sides. I've solved this by simply filling the slot by melting filament which securely locks the other sides of the nut in. Again this is definitely design dependent, requiring the ability to create a slot in the side of the piece to slide the nut in sideways, but it is definitely the strongest when it comes to pull load as the nut is fully embedded in the piece.
@@newolde1 CNC Kitchen did some testing with steel nuts. Putting the nut in a bottom pocket gave over twice the pull-out strength over the side-pocket method! (165kg vs 85kg) Even threading a bolt directly into the plastic was stronger than side-pocket at 118kg! Helicoils and inserts were barely better than direct threading the bolt into the plastic at 120kg vs 118kg so inserts are a waste of time it appears. (maybe useful if inserted differently as some suggested here) There are some surprising results: th-cam.com/video/2wRc1KbEAU8/w-d-xo.html
Surprised that the direct “screw into plastic” method was so effective at pull out resistance. Performed just as expected on the torque-our test. Didn’t think it was going to be any good at that.
I did this with a bunch of M6 screws. I heated a screw with a small heat gun, and threaded the screw in to the PLA to form the threads. After the holes cooled, I stripped a couple of the holes by tightening the too much. The M6 threads are pretty deep, about the same as the wall thickness of the PLA. If I figured out how to create thicker walls around the holes than the rest of the model it would have been better. This video is great because I've wondered about which inserts would work best. Now, I wonder if the results would be the same with bigger screws.
I always just tap right into PLA with a tap set and never had an issue where those threads were the weak point. I bought hundreds of brass inserts to use when the tapped threads fail...but they have yet to fail!
Stefan, I really appreciate all the testing you do. When I have questions about what I can get away with when designing for AM, you are my first source. Thank you so much!
I think you nailed it in this one. As tempting as it is to go with the cheap inserts for low strength applications, the amount of time and effort of keeping them aligned is a deal breaker for me, especially if i was trying to do a small production run of parts.
I'm new to 3D printing. I've been watching your videos to get some tips and tricks to improve my skills. I don't normally comment on TH-cam videos but I would like to say your videos have helped me tremendously in improving my skills. Thank you for the time and effort you put into your videos to help all of us. Keep up the good work.
Thank you so much for these great guides. Over the past few years as I've done 3D printing your experiments and wisdom have really helped to smooth out the process for me. I know that making such videos and experiments are a challenge and can be not as fruitful as you might need them to be to pay the bills, but hopefully the small consolation of being a true and wonderful virtual mentor can temper the struggle. Thank you Stephan 🎉
Fantastic tests. I think what this tells me is that just driving the screws directly into the plastic is fine, unless you plan to repeatedly remove the screw and replace it.
Stefan, I totally love your very professional approach to all the project you make. Maybe one if not the most serious 3D printing channel out there. Please keep up the good work. I keep watching your video, even if I dont need it, just to see the level of high quality :) Five stars from me *****
For such a technical hobby the 3d printing community really does have a lot of folk wisdom floating around that is pretty much baseless. Its really nice to have these videos to provide some real information.
@@ArnaudMEURET When I got my first roll of PETG I spent the longest time taking pains to 3d print everything as one part because I had seen loads of people say it was near impossible to glue it. But one day when I was out a family member accidentally broke one of my prints and glued it back together. Turns out super glue works fine on PETG. I learned a lesson about 3d printing common knowledge that day.
@@outsider344 It kind of depends. ABS you can just partially dissolve with acetone and get perfect properties in the joint, as if it had been one piece (disregarding internal structure from infill). Super glue works well if you just want a stiff joint. If you are however using the flexibility of PETG properly gluing it is hard.
@@FreeOfFantasy I agree. But I had seen people saying you just can't glue petg at all. That they avoid printing it because the only way to join pieces is with mechanical fasteners. You do make some compromises with the material properties at the joint when you use superglue but that beats the heck out of having to screw parts together.
As I was searching for threaded inserts for hardwood, your website kept showing up. I follow your channel so I'm delighted to see that SEO for your website is really good.
I would be interested to see how this compares to having inset hex nuts instead. I tend to use nuts to fasten my prints together (similar to the construction of a Prusa) for ease of use since it doesn't require a soldering iron. I could be persuaded to switch if threaded inserts were significantly stronger though. I suppose this largely depends on the amount of infill used.
My guess is the nut will have a higher pullout, but would have lower torque out because it crushes the part. The threaded inserts are touching the bolt head when they failed in torque out.
Hi there! I worked in the recycling industry for years and when we started to use 3d printed parts for our valve systems, we quickly banned all inserts. They just won't hold and become loose due to vibration. You can't loctite them and no other glue seems to work either. The solution was using square nuts in channels. They are not only cheaper but can also handle higher torques than inserts or hex nuts. I hope that helps! Regards, Etna.
Inset hex nuts are my favorite too, either installed after the fact in a slot or captive in the part by installing them during printing by pausing after the last layer before the hole is closed up.
I haven't gotten around to using threaded inserts yet, mainly because I'm too lazy to go buy a new soldering iron lol but somehow preferred to buy hexagonal female standoffs. I use them in the same fashion as inset hex nuts, using friction fit to hold them in place and put in a screw from the opposite side so they stay put.
I'm frankly amazed how well tapping the plastic with the fixing performed. I mean it's not much good for repeatability and removing/reinserting the fixings but it's quite incredible for a single use fixing.
Find ich toll, dass Du 3D-Print-Grundlagenforschung betreibst - scheinst damit so ziemlich der Einzige zu sein, der das macht und dazu noch so systematisch und akribisch - wissenschaftlich eben. Erspart einem 'ne Menge eigener Ausprobiererei.
i model threads in the part in fusion360, with a little bit of chamfer on top, and directly screw in the plastic. In my experience, with 4 perimeters of 0.4 nozzle this is a rather durable connection, even after a half a year of screwing in and out 1-2 times a week threads hold properly, i just make sure not to overtighten them. So, id be glad if you tested torque and pull-out force that this connection can withstand with differing amount of wear on it - perhaps, strength after 0 cycles, 10 cycles, 50 or even more
I hope you know Stefan that you have a great effect on the whole 3D printing community at large. I have seen experts reference your videos many times :)
Once I was able to snap M3 A2 stainless steel bolt when screwing it into bare PLA part with 100% infill and thread lenght 10mm which was created by first drilling recommended size of hole for tapping and then tapped it with the bolt itself. Cheap injection molding inserts, completly useless, come out easily, get blocked at the end - so same results as you had. More proper press-fit type inserts("eBay" ones) work well for me, but I don't use heat like you do, I drill smallest recommended hole for this insert, then press it cold, requires quite a bit of force, at least 50kg, but then stays in quite well. Those special inserts "specially designed for PLA" yeah you're correct, those are inserts for standard plastic, didn't use them tho. And yes I do drill holes in filament, it's stronger as you have solid wall instead of textured wall, it's too much of a pain in the ass to try to print exact hole size, depends on too many factors, too hard to control it, just design 3mm hole, it will print as 3.2mm then you just correct it with 3mm drill.
I bought some threaded inserts from eBay and chose the sort with the short plain section (like the $18 inserts) because I thought it would be eaiser to locate an the plastic was less likely to get "ploughed" into the thread. It's nice to see my "Guess-geneering" actually was right and confirmed by actual testing. Thanks for all the effort you put into this kind of testing, the 3D printing space it way to full of things that are taken to be true because someone saw it somewhere, but for which there is no actual video or blog post showing whether it is true or not, or still true with modern materials and printer designs.
would like to see test of Screws that are specifically made to fasten into plastic, and that self tap the hole. Like those found on many consumer electronics.
They'd be pretty good for pull-out (dependent on length, obviously) and fairly lousy for torque. The holes also deteriorate quickly if you take the screws in and out. I actually watched this video whilst buying threaded inserts (and stainless M4 pins) to replace the self tapping screws that hold on my car's pollen filter housing and can no longer be properly tightened. If you will need to remove and refit the screws then self tappers aren't a good choice in any material.
Great video once again. I would have loved to see the embedded-regular-nut technique too, that's my go-to technique. I would suspect it had very high pullout strength, but relatively poor torque-out strength, especially for small nuts.
I've been using M2 & M3 bolts to hold printed pieces together that can't be printed in one go and don't need separating, but I will be watching this video with interest as it's always good to know new engineering techniques. Thank you.
I've been using some I got from Adafruit and have found that, as I iterate, I waste a lot. Thankfully, I've found that you can actually reuse them if you screw a screw in all the way, blast it with a heatgun, and pull. Ruins the part, but if it's a failure anyway who cares? Just thought I'd share in case it helps anyone.
I have the cheap Ali inserts. Mostly for items with repeated screwing and unscrewing with little pullout load. I also design a small indented circle that works with the outer diameter of the insert as a position guide. I also modified a soldering tip to have a cylindrical tip that fits into the insert to keep it aligned and stable and improve heating of the insert.
Wonderful video. I usually use the under-sized hole method. I've also had success with regular M3 nuts thermally set into one side of the part. The measurements of having a nut thermally set in a hexagon pocket in the opposite end of the part would be interesting. This is cheaper than the Chinese brass injection molding inserts, and likely provides greater strength, due to the exposed diameter of the nut being supported by almost the entire thickness of the plastic.
DANKE!!! This very helpful video has helped me prevent a future headache, I haven't need to use these inserts in any prints yet but your video made me think about the kit i purchased. I had purchased a knurled insert kit and they were the nasty cheap ones more made for injection moulding but falsely sold for 3d printers, they have no chamfer and the knurled part is at the bottom and top so they clog with material and don't sit flush while skidding around on the print. I've returned the item and in exchange purchased a chamfered knurled insert kit, I was going to get the ruthex ones but for £16 for 100 m3 pieces I decided with another brand that is £16 for 240 pieces of m3-m5 inserts that are exactly the same in design appearance with the knurling changing direction and a nice chamfer for seating the insert into your work piece.
Would be interesting to see case variation with screw directly to plastic, reasonably tightened, and performing the pull out test after some time: from my experience screwing directly to plastic suffers often from creep or material relaxation over time thus might not be a long term reliable solution. Otherwise it was a great video, as usual! Cheers!
I worked at a POS(Point of Sale) industrial repair shop and most if not all of the units I fixed used the threaded inserts at 1:50. They didn't seem to pull out very often and were VERY strong. The plastic they used was a PC/ABS compound and hardly ever broke.
One thing to consider with the cheap inserts, well all inserts, with 4 inserts working together, like the fan mount, the cheap inserts would give you up to 156kg before they started to fail.
You should never rely on multiple fasteners to do the job unless one is strong enough by itself. The rest are just for support but all of the load will only be applied to one and not spread evenly.
I would be curious to see the effect of hole geometry on the pullout strength of the cheap inserts. For instance, you could try having a conical top to facilitate threaded insert alignment. You could combine this with a narrower diameter below the cone to help grab on the smooth ring of the cheap threaded insert.
Unfortunately the insertion path robs you of most of the freedom there. Because the bottom is flared out, the top is flared out; once you insert the bottom, you have displaced the material and it won't really fill the void around the waist of the insert no matter how you designed it originally, i would think, since the plastic is never fully molten.
Me too, I use the cheap inserts and found good results when reducing the printed hole diameter to something like 3.5mm . Yes it causes excess pla to ooze out when inserting but it definitely makes for better pull out strength. I suspect some experiments will show that its possible to make usable inserts with the cheaper type.
Danke Stefan! I don't have the time to do tests like this with this much fine detail. I'm letting my work colleagues now about your channel. Please keep going.
Good question. From experience, captive nuts perform better (when used on the reverse side of the part, or even embedded inside the part), simply because they have a larger surface area to engage with the plastic.
@@fredgenius Your claim does not make sense with respect to the empirical data given. If you had the hot melt insert on the back side of the print, the plastic would again be the weak link. Look at what the inserts did in the last video when he had infill. The inserts were strong enough to blow apart the plastic well beyond their diameter. Having more surface area is a moot point when the weakest point is the surrounding infill.
@@Anyone700 As I said, it is my experience. Much depends on the material, and structure of the print, I suggest you try for yourself and form your own conclusions.
I work in a plastics factory. We use an insert very similar to the Ruthex. We load it into the tool before the part is injection molded. We also insert similar inserts into parts after molding with a sonic welder.
@@timha4102 Yes, I use this trick to help get the insert square, so the screw is perpendicular to the surface. I use a long stainless screw so I don't burn my fingers, and easier to judge the correct angle.
I use chip board screws direct into material, 6 sided holes in the 3d print - the hex holes seem grip really well. I set short axis of hex hole (parallel sides) to inner diameter of the screw (inside thread, measured at top of screw with calipers) Seem to work :) Thanks again for all you hard work.
Sometimes we don't need a lot of holding strength but instead need the ability to insert and remove the screws a number of times without damage to the plastic. In this case the cheap inserts could be more than sufficient.
@@TheMidnightSmith if you don't need a lot of holding strength the cheap ones are good enough. Now that i know that cheap ones holds up to 30KG i surely going use then for hold up to 15KG...
Thank you ! If you take a look at the PLA caracteristics you purchase you will see the % of extend or retract it have so it help to correct the mesurement when you need a part at the right size.
@CNC Kitchen I just clicked on the aliexpress search link elsewhere in the comments and I saw something interesting. They have threaded inserts that are like solid helicoils. They are threaded inserts that are threaded on the outside as well. A M3 insert is threaded as M5 on the outside. Your testing showed that self tapped screws held on almost as well as an insert. Extrapolating from there a larger self tapped insert would hold on at least as well and also have the advantage of more surface area of the larger outside screw size so I would expect it would hold even better. I wonder how it would perform in the torque out test though.
You'll probably get more useful torque test results if you were bolting down a stack of washers rather than just the bolt. The longer exposed thread of the screw will be allowed to stretch and load up the insert properly
Nice test Stefan thanks. Looks like the expansive ones are the ones to get. Not only they perform better, but also locate correctly. If a part has many inserts, accurate positioning becomes very important.
Never got a notification off this one, glad I've just noticed it!! Always have used the eBay ones, but they are quite expensive. You also can reuse them, use the soldering iron to pull them out and let the pla burn off.
The undersized hole was seriously impressive! Seem like it's worth spending a little more on fancy inserts if you can. Definitely going to keep this video in mind for future projects!
I was wondering about the pullout strength of bolts heated to 210 degrees C then screwed directly into PLA. It seems like this would give higher pullout strength because the PLA can flow more fully into the thread form. Another brilliant video Stefan.
Thank you for the video, honestly only ever used the Injection molding inserts and just figured they were all about the same. One difference between how I use them and how you tested could make for a video idea. When I design holes, I make the octagonal rather than circular..... they seem to be printed more accurately in size, and since they are printed so small, there's little difference in the viewed shape made. Too that, I usually design object to accept the insert in from the rear or the part, the chase the threads with a tap. IDK if it helps with strength, but my thought was that I'm essentially trying to pull the inset through the plastic when tightened, rather than pulling against the direction of insertion.
Great testing, I was curious of the quality I should get and now I know. Best part is that you sell em on your website, so I can support the creator & get quality inserts in one transaction!
I just stumbled on your channel recently, and this was a great video. Thanks for your work on it. Something worth noting about screwing directly into the PLA: since you're effectively cutting threads out of plastic using your screw, those threads likely won't stand up very well to the wear and tear of removing and re-inserting a screw multiple times. If you're looking at a situation where you need to add and remove screws from a part multiple times, you'll almost certainly want to use an insert.
damn that 0.1 difference is great. most printers I see make it 0.5~ smaller. mine used to do 0.3mm i think? i'll check again EDIT: huh mine can do 0.1 too with ease. I was wrong this hole time....
I use hot inserts similar to the ones you prefer for 1/4"-20 UNC tripod mounts. For other applications I print holes and hand tap as it works better for me than forcing a machine screw. For holes of 2mm or less, I use wood screws as machine screws are not intended to be self tapping into soft materials.
Very interesting results, I expected the cheap inserts to be much worse, but surprisingly not bad in comparison. The raw plastic results were the most surprising, I didn't realize pla was that strong.
I bought some of the injection moulding inserts a couple weeks ago when the original threaded inserts video appeared in my feed. I was going to use them for a Redox keyboard, and I still am. However, I now know that they're best suited for threads where there isn't much of an expected load. I've had to get the bottom case parts reprinted because I was trying to force an M3 threaded insert into a 3mm inner diameter (I didn't RTFM). I've been able to get the 3mm inner diameter increased by 0.45mm, which makes the inner diameter of the hole only 0.05mm smaller than an M2 threaded insert outer diameter. Obviously that's only on paper, and in the real world, it'll be more like 0.08mm smaller. But at least it'll insert without gunking up too much
@@Ce3pyo good point Mr. Hartmann, I however always screw directly into plastic (or to a nut behind it). Inserts are just not ever absolutely required, really, so I've never needed them..
I am grateful for your work! There are also many 3d printing engineers in Russia and your videos are a very useful tool, especially tests and comparisons. Выражаю благодарность за ваш труд! В России тоже много инженеров, занимающихся 3д печатью, и ваши видео являются очень полезным пособием, в особенности тесты и сравнения.
Damn, these videos are always to the point and address all the points I'm intested in on a certain issue, definitely the best ressources for learning to construct structurally sound 3d printable designs
I have these same inserts, and I had came to this conclusion as well, the 18c ones are excellent and are my preference, easy to insert, and hard to pull out.
When I'm screwing directly into PLA, I first heat up the bolt a bit with a butane torch (to about 60-80°c). It makes screwing it in a lot easier, and I think it helps the thread form better, since the plastic that's pushed out can fill any gaps made by extending the hole. It also prevents the plastic from splitting or cracking. I'd love to see a comparison between a bolt that was screwed in warm VS one screwed in cold.
For my 3D prints I've been using flange-head press in inserts backwards, and heated into the 3D printed part. This is so the flange head is hopefully spreading the force on the flange, and gas to pull through the print to fail rather than away from the side it was inserted on.
M3 screws are not that big. A random screw calculator at www.amesweb.info/Screws/Metric_Bolt_Grades_Strength.aspx gives 2.1 Nm minimum for the best 12.9 grade, so these are quality indeed IMHO.
lol, people never understood why i told them to always model with as many sides as you can with a circle. Some cad programs allow you to pick the amount of line segments for circles, IE sketchup. so i make mine with 96 sides. This makes it less severe for dimensional shrinking. 3D printing is great but its not ment to replace certain processes in the machining world. Its ment to compliment and add to the manufacture process. I have been watching these videos of yours for a while and i thought i had subscribed, i guess not, but today i did. :)
Thanks for this one! I just came across this while shopping for some heat set inserts for my prints. I ordered a ruthex assortment and their tips as well. Thanks again!
Great work Stefan ! Perhaps you would consider testing a Helicoil insert thread as well ? Helicoils should have the advantages of a plain tapped thread, but with the additional strength of a larger diameter thread.
Hi, I use the cheap ones for when needed. More like multiple loosen/tighten threads. Because pla wears out. But I have another approach. I make the hole slightly smaller. Heating up a little and then pull them in with a bold on the other side. Not completely through the hole and a little plastic in front so it's locked up. And before pulling it in a drop of CA glue. Maybe you can test something like this too. Yesterday I saw your epoxy / CA glue video. I also thought CA would improve strength between layers but turns out it doesn't. For expoxy I use Z-poxy but probably much more expensive then HK. Although it less yellow I think. After a while it still colors yellow due sunlight. And try some carbon fiber or glascoat with CA instead of epoxy in your next test. Keep up the good work.
If you ever do any repeat testing. In the past, I've also added hexagon holes in the bottom of the part for a standard hex nut, also a pocket that the nut slides into.
Splendid work! Your engineering-based experimental approach has greatly informed my own choices: My Sidewinder X1 arrives tomorrow! Thanks to your investigations, I believe I can make truly strong and reliable large parts that will also support repeated assembly. Thanks!
great video, was curious of the process, I was prototyping a handle cover for something with some PLA and with just set screws the plastic of course ended up cracking away at the layers when you over torqued it as you'd assume, so also got these heat inserts to use with the set screws and I got the same style of the amazon ones you showed in this video, and they work much better. Very informative video, great content.
Don't forget to share this video on Facebook, Reddit, Twitter and other social media!
Impressive analysis committee and methodology as usual Stefan!
On your torque out test, the reason for failure of the bolt head is because you are torquing againts the insert itself, you could try using a longer bolt and thread lock it into the insert and use a washer to provide lateral support for the fastener head and shank.
In my practice most (if not all) of the insert are being pulled out from the opposite end - so "pull-out force" is not important, as printed parts (elements of HEVo) are starting to crackle ominously when screwing then hard :)
Also - by inserting M3 bolt from the opposite end while heating insert with soldering iron helps A LOT with keeping them straight. And I used only cheap ones so far.
Just a heads up, your M3 aliexpress affiliate link is broken.
There is a problem with the audio in the video, around 1:05 it cuts Your voice and then returns to normal after a second
Great video anyway :D
For me... your channel is the most important one for 3D printing. I say that because you address structural (as opposed to aesthetic) issues, and... you do it in a way that the layman can understand. There are other very technical 3D print channels, but they are useless to me. Their viewers would need to be advanced printers to know what they were even talking about. Thank you for the great work.
Agreed. I'm not a novice and this channel still has the best content. Some of the other popular ones all eventually devolve into "politics" of the 3D printing world - things I have zero interest in.
CNC Kitchen is the best resource channel for 3D printing. A lot of it applies for any printer on the market. I like makers muse is great for fun things to try. Or cool test prints. Bur cnc kitchen is legendary for data being pressnted in an interesting way
If you ever find yourself looking for thorough testing on household items rather than 3D printers, I highly suggest Project Farm. He tests just about everything. Tape, oil, ratchet straps, batteries, drain cleaner, tire sealants, portable tire pumps, nut removers, brush blades, windshield wipers, drywall anchors, the list goes on. Some tests are less scientific than others (the brush blades come to mind, lots of room for human error), but they're more than sufficient for how people actually use them. Plus he never accepts sponsorships.
@@The_Forge_Master however he has the personality of a small pencil.
Agreed. CNC kitchen will turn you into an advanced printer and you won’t even notice it.
As an ex aircraft engineer, I've always fitted inserts from the back side for a superior fixing.
thank you for sharing you experienced tips :)
If that is a possibility, but if it is it will greatly improve performance. I would like to test out if it would be possible to insert them during the print with a pre programmed pause in the g-code for parts where that wouldn't be a possibility, though I think I'll run into problems with the bed level.
You mean so the thread goes from the other side through the plastic?
@@EGGSHL I think he means he'll insert it from the back side so the force on the end product will try to pull it further in.
not always possible but good when it is.
Thank you again. You're the only one who provides reliable, real-world data for this sort of thing, and it helps me when I design my projects.
Yeah- what Darrin said!
Now i wish I've seen this video 2 weeks ago, just had a project, where this would have been the perfect method...
Those "cheap" inserts appear to be overmolding inserts, designed to be held in place and have plastic flow around them. When you heat press those, I'm guessing a cavity is formed where the plastic did not flow and create the ledge necessary to mechanically lock into the plastic.
this is 100% what is going on. Those are meant to be in place at the time of molding, not added after the fact. That's why they don't have the lead in chamfer and they push plastic when you try to melt them in. They are not designed for this application, though they will work ok for light loads. When used in their intended application they are equally or more strong than the actual heat set inserts.
Perhaps it would be worthwhile to experiment with holding the soldering iron for longer to see if melting the plastic more helps with that issue.
agree
I just print the hole smaller than it should be and hold the iron for a bit longer. The only time such a part broke wasn't even in the place of the insert despite it being the place with the most stress applied
Was hoping you would try a pull through test as well as a pull out since that would provide the greatest strength for them all. I.E. insert the insert on one side but screw the bolt into the other.
Also, I wonder how well M3 wood inserts would work (if at all) i.ebayimg.com/images/g/2koAAOSwDoNd1PUd/s-l400.jpg
Whenever possible, threaded inserts should be installed on the side of the part opposite from where the fastener is entering it. (Obviously this isn't possible with blind holes) This preferred orientation results in the fastener pulling the insert even more securely into the part. It would be interesting to repeat your pullout tests with your test blocks flipped over. I've found that the plastic part will usually fail catastrophically before the insert is pulled out.
I would think your major advantage to a threaded insert, particularly when you might choose the cheap ones instead of threading directly into plastic, would be for something that doesn't require great strength, but *does* require repeatedly removing and re-installing the fastener. In those conditions, cross-threading and wear would quickly destroy the threads in plastic alone, where the brass insert would happily withstand as many cycles as you'd wish to throw at it without appreciable wear.
Thanks for making this series of videos. I, and probably many other people, find this information invaluable. I’ve always wondered if it was worth buying the more expensive, chamfered, inserts. Your videos are always quick, to the point, and well edited. Well done!
Was just thinking about fitting them into a pocket from the back, The back hole should be sized for the insert but the front sized to just clear the thread thus leaving a shoulder for the insert to pull up to.
Love the testing machine by the way
I'm using your suggested method whenever possible and it works very well even with the cheapest inserts.
Another potential depending on the design is to make a slot for a regular or square nut to insert them from the side. This essentially embeds the nut in the middle so it is as strong as your piece on pullout. The only issue that I've run into with this (mainly with regular hex nuts, less so with square nuts) is torquing, as the nut may spin through the plastic when tightened as it's only held by 4 sides. I've solved this by simply filling the slot by melting filament which securely locks the other sides of the nut in. Again this is definitely design dependent, requiring the ability to create a slot in the side of the piece to slide the nut in sideways, but it is definitely the strongest when it comes to pull load as the nut is fully embedded in the piece.
Absolutely.
@@newolde1 CNC Kitchen did some testing with steel nuts. Putting the nut in a bottom pocket gave over twice the pull-out strength over the side-pocket method! (165kg vs 85kg) Even threading a bolt directly into the plastic was stronger than side-pocket at 118kg! Helicoils and inserts were barely better than direct threading the bolt into the plastic at 120kg vs 118kg so inserts are a waste of time it appears. (maybe useful if inserted differently as some suggested here) There are some surprising results: th-cam.com/video/2wRc1KbEAU8/w-d-xo.html
Surprised that the direct “screw into plastic” method was so effective at pull out resistance. Performed just as expected on the torque-our test. Didn’t think it was going to be any good at that.
I did this with a bunch of M6 screws. I heated a screw with a small heat gun, and threaded the screw in to the PLA to form the threads. After the holes cooled, I stripped a couple of the holes by tightening the too much. The M6 threads are pretty deep, about the same as the wall thickness of the PLA. If I figured out how to create thicker walls around the holes than the rest of the model it would have been better. This video is great because I've wondered about which inserts would work best. Now, I wonder if the results would be the same with bigger screws.
@@microdesigns2000 Generally speaking, the bigger you go, the easier the pull-out will be, but the harder the torque-out will be.
I always just tap right into PLA with a tap set and never had an issue where those threads were the weak point.
I bought hundreds of brass inserts to use when the tapped threads fail...but they have yet to fail!
Stefan, I really appreciate all the testing you do. When I have questions about what I can get away with when designing for AM, you are my first source. Thank you so much!
I think you nailed it in this one. As tempting as it is to go with the cheap inserts for low strength applications, the amount of time and effort of keeping them aligned is a deal breaker for me, especially if i was trying to do a small production run of parts.
I'm new to 3D printing. I've been watching your videos to get some tips and tricks to improve my skills. I don't normally comment on TH-cam videos but I would like to say your videos have helped me tremendously in improving my skills. Thank you for the time and effort you put into your videos to help all of us. Keep up the good work.
I recognize those stickers at 6:56! Thank you for the feature Stefan! Also great video on threaded inserts!
I don't even have a 3D printer, but I think it's amazing that you did this test!
Those animated graphs next to the pull test are really nice!
Thank you so much for these great guides. Over the past few years as I've done 3D printing your experiments and wisdom have really helped to smooth out the process for me. I know that making such videos and experiments are a challenge and can be not as fruitful as you might need them to be to pay the bills, but hopefully the small consolation of being a true and wonderful virtual mentor can temper the struggle. Thank you Stephan 🎉
Fantastic tests. I think what this tells me is that just driving the screws directly into the plastic is fine, unless you plan to repeatedly remove the screw and replace it.
Stefan, I totally love your very professional approach to all the project you make. Maybe one if not the most serious 3D printing channel out there. Please keep up the good work. I keep watching your video, even if I dont need it, just to see the level of high quality :) Five stars from me *****
For such a technical hobby the 3d printing community really does have a lot of folk wisdom floating around that is pretty much baseless. Its really nice to have these videos to provide some real information.
So true. DIY 3D printing is the poster child of Cargo Cult.
@@ArnaudMEURET When I got my first roll of PETG I spent the longest time taking pains to 3d print everything as one part because I had seen loads of people say it was near impossible to glue it. But one day when I was out a family member accidentally broke one of my prints and glued it back together. Turns out super glue works fine on PETG. I learned a lesson about 3d printing common knowledge that day.
@@outsider344 It kind of depends. ABS you can just partially dissolve with acetone and get perfect properties in the joint, as if it had been one piece (disregarding internal structure from infill). Super glue works well if you just want a stiff joint. If you are however using the flexibility of PETG properly gluing it is hard.
@@FreeOfFantasy I agree. But I had seen people saying you just can't glue petg at all. That they avoid printing it because the only way to join pieces is with mechanical fasteners. You do make some compromises with the material properties at the joint when you use superglue but that beats the heck out of having to screw parts together.
As I was searching for threaded inserts for hardwood, your website kept showing up. I follow your channel so I'm delighted to see that SEO for your website is really good.
I would be interested to see how this compares to having inset hex nuts instead. I tend to use nuts to fasten my prints together (similar to the construction of a Prusa) for ease of use since it doesn't require a soldering iron. I could be persuaded to switch if threaded inserts were significantly stronger though. I suppose this largely depends on the amount of infill used.
My guess is the nut will have a higher pullout, but would have lower torque out because it crushes the part. The threaded inserts are touching the bolt head when they failed in torque out.
Hi there!
I worked in the recycling industry for years and when we started to use 3d printed parts for our valve systems, we quickly banned all inserts. They just won't hold and become loose due to vibration. You can't loctite them and no other glue seems to work either. The solution was using square nuts in channels. They are not only cheaper but can also handle higher torques than inserts or hex nuts. I hope that helps!
Regards,
Etna.
Inset hex nuts are my favorite too, either installed after the fact in a slot or captive in the part by installing them during printing by pausing after the last layer before the hole is closed up.
There is also the question of square and hex nuts. I would guess that square work better in this application.
I haven't gotten around to using threaded inserts yet, mainly because I'm too lazy to go buy a new soldering iron lol but somehow preferred to buy hexagonal female standoffs. I use them in the same fashion as inset hex nuts, using friction fit to hold them in place and put in a screw from the opposite side so they stay put.
I'm frankly amazed how well tapping the plastic with the fixing performed. I mean it's not much good for repeatability and removing/reinserting the fixings but it's quite incredible for a single use fixing.
Find ich toll, dass Du 3D-Print-Grundlagenforschung betreibst - scheinst damit so ziemlich der Einzige zu sein, der das macht und dazu noch so systematisch und akribisch - wissenschaftlich eben. Erspart einem 'ne Menge eigener Ausprobiererei.
i model threads in the part in fusion360, with a little bit of chamfer on top, and directly screw in the plastic. In my experience, with 4 perimeters of 0.4 nozzle this is a rather durable connection, even after a half a year of screwing in and out 1-2 times a week threads hold properly, i just make sure not to overtighten them.
So, id be glad if you tested torque and pull-out force that this connection can withstand with differing amount of wear on it - perhaps, strength after 0 cycles, 10 cycles, 50 or even more
Also self-tapping and regular screws with no thread would be interesting.
I hope you know Stefan that you have a great effect on the whole 3D printing community at large. I have seen experts reference your videos many times :)
Once I was able to snap M3 A2 stainless steel bolt when screwing it into bare PLA part with 100% infill and thread lenght 10mm which was created by first drilling recommended size of hole for tapping and then tapped it with the bolt itself.
Cheap injection molding inserts, completly useless, come out easily, get blocked at the end - so same results as you had.
More proper press-fit type inserts("eBay" ones) work well for me, but I don't use heat like you do, I drill smallest recommended hole for this insert, then press it cold, requires quite a bit of force, at least 50kg, but then stays in quite well.
Those special inserts "specially designed for PLA" yeah you're correct, those are inserts for standard plastic, didn't use them tho.
And yes I do drill holes in filament, it's stronger as you have solid wall instead of textured wall, it's too much of a pain in the ass to try to print exact hole size, depends on too many factors, too hard to control it, just design 3mm hole, it will print as 3.2mm then you just correct it with 3mm drill.
I bought some threaded inserts from eBay and chose the sort with the short plain section (like the $18 inserts) because I thought it would be eaiser to locate an the plastic was less likely to get "ploughed" into the thread. It's nice to see my "Guess-geneering" actually was right and confirmed by actual testing. Thanks for all the effort you put into this kind of testing, the 3D printing space it way to full of things that are taken to be true because someone saw it somewhere, but for which there is no actual video or blog post showing whether it is true or not, or still true with modern materials and printer designs.
would like to see test of Screws that are specifically made to fasten into plastic, and that self tap the hole. Like those found on many consumer electronics.
They'd be pretty good for pull-out (dependent on length, obviously) and fairly lousy for torque. The holes also deteriorate quickly if you take the screws in and out.
I actually watched this video whilst buying threaded inserts (and stainless M4 pins) to replace the self tapping screws that hold on my car's pollen filter housing and can no longer be properly tightened.
If you will need to remove and refit the screws then self tappers aren't a good choice in any material.
You and teaching tech are my best resources as I get into this hobby. Thank you for all you do.
Great video once again. I would have loved to see the embedded-regular-nut technique too, that's my go-to technique. I would suspect it had very high pullout strength, but relatively poor torque-out strength, especially for small nuts.
I've been using M2 & M3 bolts to hold printed pieces together that can't be printed in one go and don't need separating, but I will be watching this video with interest as it's always good to know new engineering techniques. Thank you.
I've been using some I got from Adafruit and have found that, as I iterate, I waste a lot. Thankfully, I've found that you can actually reuse them if you screw a screw in all the way, blast it with a heatgun, and pull. Ruins the part, but if it's a failure anyway who cares? Just thought I'd share in case it helps anyone.
Good idea!
This is actually where the cheap injection molding inserts excel - they are so easy to pull out that you can reuse them again and again... -.-
I have the cheap Ali inserts. Mostly for items with repeated screwing and unscrewing with little pullout load.
I also design a small indented circle that works with the outer diameter of the insert as a position guide. I also modified a soldering tip to have a cylindrical tip that fits into the insert to keep it aligned and stable and improve heating of the insert.
Wonderful video. I usually use the under-sized hole method. I've also had success with regular M3 nuts thermally set into one side of the part. The measurements of having a nut thermally set in a hexagon pocket in the opposite end of the part would be interesting. This is cheaper than the Chinese brass injection molding inserts, and likely provides greater strength, due to the exposed diameter of the nut being supported by almost the entire thickness of the plastic.
DANKE!!!
This very helpful video has helped me prevent a future headache, I haven't need to use these inserts in any prints yet but your video made me think about the kit i purchased.
I had purchased a knurled insert kit and they were the nasty cheap ones more made for injection moulding but falsely sold for 3d printers, they have no chamfer and the knurled part is at the bottom and top so they clog with material and don't sit flush while skidding around on the print. I've returned the item and in exchange purchased a chamfered knurled insert kit, I was going to get the ruthex ones but for £16 for 100 m3 pieces I decided with another brand that is £16 for 240 pieces of m3-m5 inserts that are exactly the same in design appearance with the knurling changing direction and a nice chamfer for seating the insert into your work piece.
Would be interesting to see case variation with screw directly to plastic, reasonably tightened, and performing the pull out test after some time: from my experience screwing directly to plastic suffers often from creep or material relaxation over time thus might not be a long term reliable solution.
Otherwise it was a great video, as usual! Cheers!
I worked at a POS(Point of Sale) industrial repair shop and most if not all of the units I fixed used the threaded inserts at 1:50. They didn't seem to pull out very often and were VERY strong. The plastic they used was a PC/ABS compound and hardly ever broke.
One thing to consider with the cheap inserts, well all inserts, with 4 inserts working together, like the fan mount, the cheap inserts would give you up to 156kg before they started to fail.
You should never rely on multiple fasteners to do the job unless one is strong enough by itself. The rest are just for support but all of the load will only be applied to one and not spread evenly.
Doesn't rarely ever workout like that in reality. It's also a very twisted way of thinking.
I appreciate the work you put in to these tests.
Thanks!
I would be curious to see the effect of hole geometry on the pullout strength of the cheap inserts. For instance, you could try having a conical top to facilitate threaded insert alignment. You could combine this with a narrower diameter below the cone to help grab on the smooth ring of the cheap threaded insert.
Unfortunately the insertion path robs you of most of the freedom there. Because the bottom is flared out, the top is flared out; once you insert the bottom, you have displaced the material and it won't really fill the void around the waist of the insert no matter how you designed it originally, i would think, since the plastic is never fully molten.
Me too, I use the cheap inserts and found good results when reducing the printed hole diameter to something like 3.5mm . Yes it causes excess pla to ooze out when inserting but it definitely makes for better pull out strength. I suspect some experiments will show that its possible to make usable inserts with the cheaper type.
Danke Stefan! I don't have the time to do tests like this with this much fine detail. I'm letting my work colleagues now about your channel. Please keep going.
For the next video
How well do embedding an ordinary nut compare
Good question. From experience, captive nuts perform better (when used on the reverse side of the part, or even embedded inside the part), simply because they have a larger surface area to engage with the plastic.
@@fredgenius captured hex vs captured square nuts!
They can be fiddly as hell, though.
@@fredgenius Your claim does not make sense with respect to the empirical data given. If you had the hot melt insert on the back side of the print, the plastic would again be the weak link. Look at what the inserts did in the last video when he had infill. The inserts were strong enough to blow apart the plastic well beyond their diameter. Having more surface area is a moot point when the weakest point is the surrounding infill.
@@Anyone700 As I said, it is my experience. Much depends on the material, and structure of the print, I suggest you try for yourself and form your own conclusions.
I work in a plastics factory. We use an insert very similar to the Ruthex. We load it into the tool before the part is injection molded.
We also insert similar inserts into parts after molding with a sonic welder.
I fit them from the back of the part where possible, then they perform much better!
That's a great idea! You can even make so that the hole has a smaller diameter on the end. Awesome, thanks!
@@runklestiltskin_2407 You can also make insert hole with alternate layers smaller/larger, and glue the insert in with epoxy...
Same here. I fit them in through the backside and then pull them in ffom the front with a screw. So you don‘t need the space for the soldering iron.
@@timha4102 Yes, I use this trick to help get the insert square, so the screw is perpendicular to the surface. I use a long stainless screw so I don't burn my fingers, and easier to judge the correct angle.
thats also an idea that i thought About an had liked to see as a test
I use chip board screws direct into material, 6 sided holes in the 3d print - the hex holes seem grip really well.
I set short axis of hex hole (parallel sides) to inner diameter of the screw (inside thread, measured at top of screw with calipers)
Seem to work :)
Thanks again for all you hard work.
Sometimes we don't need a lot of holding strength but instead need the ability to insert and remove the screws a number of times without damage to the plastic. In this case the cheap inserts could be more than sufficient.
Did you watch the entire video?
@@TheMidnightSmith if you don't need a lot of holding strength the cheap ones are good enough.
Now that i know that cheap ones holds up to 30KG i surely going use then for hold up to 15KG...
if it has no chamfer then it's a pain in ass to work with!
@@KiR_3d you can maybe remedy this by adding a slight counter sink to your design.
@@LopezfsS Sure, but it can be hard to implement if you have thin walls (actual walls) for example. Then on one end it will have almost no grip.
Thank you ! If you take a look at the PLA caracteristics you purchase you will see the % of extend or retract it have so it help to correct the mesurement when you need a part at the right size.
aww hell yeah. I was hoping you'd do a video on this. You do such great breakdowns.
@CNC Kitchen I just clicked on the aliexpress search link elsewhere in the comments and I saw something interesting. They have threaded inserts that are like solid helicoils. They are threaded inserts that are threaded on the outside as well. A M3 insert is threaded as M5 on the outside. Your testing showed that self tapped screws held on almost as well as an insert. Extrapolating from there a larger self tapped insert would hold on at least as well and also have the advantage of more surface area of the larger outside screw size so I would expect it would hold even better. I wonder how it would perform in the torque out test though.
Stefan, thanks for the research!
You'll probably get more useful torque test results if you were bolting down a stack of washers rather than just the bolt. The longer exposed thread of the screw will be allowed to stretch and load up the insert properly
McMaster Carr has a huge selection of quality inserts for plastics.
No McMaster in Germany unfortunately.
@@CNCKitchen That sucks, McMaster Carr is such an amazing resource.
Nice test Stefan thanks. Looks like the expansive ones are the ones to get. Not only they perform better, but also locate correctly. If a part has many inserts, accurate positioning becomes very important.
Those "Ruthex" Specially designed for 3d printing are just standard ones and can be bought everywhere and is not made by them lol!
Never got a notification off this one, glad I've just noticed it!! Always have used the eBay ones, but they are quite expensive. You also can reuse them, use the soldering iron to pull them out and let the pla burn off.
Outstanding work lad!
The undersized hole was seriously impressive!
Seem like it's worth spending a little more on fancy inserts if you can. Definitely going to keep this video in mind for future projects!
I was wondering about the pullout strength of bolts heated to 210 degrees C then screwed directly into PLA. It seems like this would give higher pullout strength because the PLA can flow more fully into the thread form. Another brilliant video Stefan.
Thank you for the video, honestly only ever used the Injection molding inserts and just figured they were all about the same. One difference between how I use them and how you tested could make for a video idea. When I design holes, I make the octagonal rather than circular..... they seem to be printed more accurately in size, and since they are printed so small, there's little difference in the viewed shape made. Too that, I usually design object to accept the insert in from the rear or the part, the chase the threads with a tap. IDK if it helps with strength, but my thought was that I'm essentially trying to pull the inset through the plastic when tightened, rather than pulling against the direction of insertion.
In some cases you might put the cheap inserts on the other side of the piece in order to get better performance
Great testing, I was curious of the quality I should get and now I know. Best part is that you sell em on your website, so I can support the creator & get quality inserts in one transaction!
awesome video once again. I finally supported you on patreon, I'm happy to help support your fantastic videos!
Thank you very much for the support! Highly appreciated.
I just stumbled on your channel recently, and this was a great video. Thanks for your work on it.
Something worth noting about screwing directly into the PLA: since you're effectively cutting threads out of plastic using your screw, those threads likely won't stand up very well to the wear and tear of removing and re-inserting a screw multiple times. If you're looking at a situation where you need to add and remove screws from a part multiple times, you'll almost certainly want to use an insert.
When I need, I insert square M3/M4 nut into print
For me your channel is the best for testing, I also used parts off your knife sharpener.
damn that 0.1 difference is great. most printers I see make it 0.5~ smaller. mine used to do 0.3mm i think? i'll check again
EDIT: huh mine can do 0.1 too with ease. I was wrong this hole time....
Yeah 0.1-0.2mm is normal. More is usually holy overextrusion batman.
Mine was on point, though I need to put it back together with the new stepper drivers.
.... that pun...
you are a certain kind of hole. have my upvote.
I use hot inserts similar to the ones you prefer for 1/4"-20 UNC tripod mounts. For other applications I print holes and hand tap as it works better for me than forcing a machine screw. For holes of 2mm or less, I use wood screws as machine screws are not intended to be self tapping into soft materials.
Half an hour ago I looked on Aliexpress to order some M3 Insert Nuts XD
Well, contextual content production is definitely better then contextual advertising.
Very interesting results, I expected the cheap inserts to be much worse, but surprisingly not bad in comparison. The raw plastic results were the most surprising, I didn't realize pla was that strong.
I’m wondering, what if you twist the insert while inserting? Would that be good or bad for the cheap inserts?
Thank you for this very informative comparison, it undoubtedly has save many persons the grief of using cheap inserts.
I never thread the plastic or use threaded inserts, I put a nut on the backside instead. This is far stronger than any threaded insert.
I bought some of the injection moulding inserts a couple weeks ago when the original threaded inserts video appeared in my feed. I was going to use them for a Redox keyboard, and I still am. However, I now know that they're best suited for threads where there isn't much of an expected load. I've had to get the bottom case parts reprinted because I was trying to force an M3 threaded insert into a 3mm inner diameter (I didn't RTFM). I've been able to get the 3mm inner diameter increased by 0.45mm, which makes the inner diameter of the hole only 0.05mm smaller than an M2 threaded insert outer diameter. Obviously that's only on paper, and in the real world, it'll be more like 0.08mm smaller. But at least it'll insert without gunking up too much
SO basically expensive insert perfectly solve a problem that doesn't exist.
no ... an insert is way better, when using the screw very much. here he tested only first use, not long time use ;)
@@Ce3pyo good point Mr. Hartmann, I however always screw directly into plastic (or to a nut behind it). Inserts are just not ever absolutely required, really, so I've never needed them..
I am grateful for your work! There are also many 3d printing engineers in Russia and your videos are a very useful tool, especially tests and comparisons.
Выражаю благодарность за ваш труд! В России тоже много инженеров, занимающихся 3д печатью, и ваши видео являются очень полезным пособием, в особенности тесты и сравнения.
The cheapy inserts need plastic to fill the recessed section between the knurls. Good for plastic injection, no good for printing.
Another great video from the kitchen 👍🏻 I am impressed how much detail you put into them and how much that helps us! Keep them coming!
2:40 - chamfer the hole
my thought exactly
Damn, these videos are always to the point and address all the points I'm intested in on a certain issue, definitely the best ressources for learning to construct structurally sound 3d printable designs
i hated these inserts my whole life. now I found out it was cause I used the wrong kind... I'll start using them more in my designs i guess xD
I have these same inserts, and I had came to this conclusion as well, the 18c ones are excellent and are my preference, easy to insert, and hard to pull out.
4:47 I have spotted the deathly hallows
When I'm screwing directly into PLA, I first heat up the bolt a bit with a butane torch (to about 60-80°c). It makes screwing it in a lot easier, and I think it helps the thread form better, since the plastic that's pushed out can fill any gaps made by extending the hole. It also prevents the plastic from splitting or cracking. I'd love to see a comparison between a bolt that was screwed in warm VS one screwed in cold.
My dad failed the pullout test
For my 3D prints I've been using flange-head press in inserts backwards, and heated into the 3D printed part. This is so the flange head is hopefully spreading the force on the flange, and gas to pull through the print to fail rather than away from the side it was inserted on.
“Quality screws” shearing at 3-4nm?! Haha no.
They arent bad but still wouldnt call them quality
Yes, that was very odd.
M3 screws are not that big. A random screw calculator at www.amesweb.info/Screws/Metric_Bolt_Grades_Strength.aspx gives 2.1 Nm minimum for the best 12.9 grade, so these are quality indeed IMHO.
lol, people never understood why i told them to always model with as many sides as you can with a circle. Some cad programs allow you to pick the amount of line segments for circles, IE sketchup. so i make mine with 96 sides. This makes it less severe for dimensional shrinking. 3D printing is great but its not ment to replace certain processes in the machining world. Its ment to compliment and add to the manufacture process. I have been watching these videos of yours for a while and i thought i had subscribed, i guess not, but today i did. :)
Thanks for this one! I just came across this while shopping for some heat set inserts for my prints. I ordered a ruthex assortment and their tips as well. Thanks again!
Great work Stefan ! Perhaps you would consider testing a Helicoil insert thread as well ? Helicoils should have the advantages of a plain tapped thread, but with the additional strength of a larger diameter thread.
I have done some pull out-test myself. Stopped doing them after the birth of my second child. Great job on the videos!
Excellent research, test results and presentation. Video could not be improved, well done. Thanks from Colorado.
Hi,
I use the cheap ones for when needed. More like multiple loosen/tighten threads. Because pla wears out.
But I have another approach.
I make the hole slightly smaller.
Heating up a little and then pull them in with a bold on the other side. Not completely through the hole and a little plastic in front so it's locked up. And before pulling it in a drop of CA glue.
Maybe you can test something like this too.
Yesterday I saw your epoxy / CA glue video. I also thought CA would improve strength between layers but turns out it doesn't.
For expoxy I use Z-poxy but probably much more expensive then HK.
Although it less yellow I think.
After a while it still colors yellow due sunlight.
And try some carbon fiber or glascoat with CA instead of epoxy in your next test.
Keep up the good work.
i've bought those cheap inserts and had ENORMOUS amount of frustration with them, thank you for video, now i now what were the problem.
If you ever do any repeat testing. In the past, I've also added hexagon holes in the bottom of the part for a standard hex nut, also a pocket that the nut slides into.
Splendid work! Your engineering-based experimental approach has greatly informed my own choices: My Sidewinder X1 arrives tomorrow! Thanks to your investigations, I believe I can make truly strong and reliable large parts that will also support repeated assembly. Thanks!
This was great, good to see I’m not the only one who wants to test stuff!.
I wish I had consulted this video before ordering inserts. Well done, and thank you for the detailed testing!
great video, was curious of the process, I was prototyping a handle cover for something with some PLA and with just set screws the plastic of course ended up cracking away at the layers when you over torqued it as you'd assume, so also got these heat inserts to use with the set screws and I got the same style of the amazon ones you showed in this video, and they work much better. Very informative video, great content.
I love your DIY testing machines and the always great data-driven content. Coming from a material lab manager at a big Automotive supplier