People critics a lot, but what I like the most in your videos is you show tons of possibilities, and while your focus is production on large scale, you bring us into creativity along the way. We tend to forget that you just don't throw your answer in our face "go accept it" but you offer all the thought process that brought to this last design. And nobody is obliged to apply to their build, so personnally I feel grateful and a time well wasted to watch your videos.
You could literally see how the fins DID NOT align the two blocks together.... At least a good 1 or 2mm out from each other 😂 Steel dowel pins are very cheap, accessible, and are very accurate. They also don't have the risk of shear like printed pins do like you shown in the video. All you need to do is make one part with holes that have an interference fit, and the other to have holes that have a close clearance fit. You don't need to reinvent the wheel. There's a reason that method is used in traditional injection moulding equipment.... BECAUSE IT WORKS AND IS REPEATABLE!!! I get the drive to think differently, but you need to also consider the fact that if something isn't broken, it likely doesn't need fixing... Also, 3D printing is really only suitable for at most low quantity batch production parts. If you're doing 10's of thousands of parts, you'd be best to just go down the injection moulding route... You'd get parts far quicker, more robust, more accurate, nicer finish, and probably very similar in cost even with mould tooling design and manufacture included.
this part could obviously be best suited for dowel pins and injection molding. You need to look 1mm below the surface that OTHER applications can not be injection molded and still need alignment pins.
Making two fins with these huge chamfers doesn't really constrain them vertically though. They will just slide out if pushed from the top or bottom. You might want to angle them 45 deg to each other which then works perfectly with printing on the chamfered corner like you showed in a previous video. edit: also changes the layer direction to a better angle so you don't risk breaking them right in plane
That is only feasible if there is the area to support them. Generally features like these would be made smaller and be on the outer edge of an electrical enclosure. It is also ideal to chamfer the inner edge of the slot as well so that part is constrained in both Axis.
@@OArchivesX Not trying to start drama, I'm genuinely curious since I've been 3D printing and always enjoy learning new things for my hobby. Are there better ways to do this than pins and holes? I would like to learn more if you have any resources or videos.
teaching tech has a great video on shadow lines which is my preferred method for this. great to get people thinking about more then just using pins and holes (A method common in furniture kits because dowels and drilling holes is cheap in that context) and thinking more critically about what they need from their registration features
If the rib is thinner than the straight pin, it is weaker in that direction. The rib also has a lot more surface area, so more opportunity for error, or over-constraining the locating feature, causing improper fit (as seen in the video).
You could decrease print time and improve manufacturability by using just holes on both sides. And then using a separate pin. I like to make my location holes the diameter of my filament and then I can use some pieces of filament as pins. If I have just a short piece of filament that doesn't fit on a roll I will cut that up for pins. I only use that method for parts that I'm going to be gluing together. If I were making a lid or something like that I would go with a lip all the way around the perimeter.
I usually just put holes on both sides, and use pieces of cut filament as the pins. It's much stronger than printed pins and while I'm not sure how it compares to fins, it's probably good enough for most applications and pretty fast to do.
That is an option. But it adds extra assembly time when the features could be integrated into the print to make it more scalable. Imagine manually adding pins 10,000 times.
The textbook way to do it without over constraining things would be to have 2 pins in one part, one hole and one slot on the other. The pin in hole gives you xyz and then the pin in slot gives you the clocking, anything more is overconstraining it.
You actually had me at cones. I think thats a fantastic change thay ill actually implement going forward. But then you lost me at this thin flimsy fins, perhaps an x shape would be best for slots, but i think you hit the nail on the head with the cone. Can Print in both directions and has equal strenth in all directions. Circles are your friend!
Pins work fine if your using them as an alignment for walls or something that is going to be superglued together. This also helps make said model stronger when gluing together as there is now more surface area for the glue to bind to and less likely the model will come apart. I usually use notches or some sort of means to align my designs together, usually corners of a wall of a model building. This makes thing a little stronger and easier to put together, while making the model itself easier and faster to print, since its all laid out flat on the bed. Due to printing inaccuracies that happen, I always error on the larger side for any holes, usually about .5mm bigger in X/Y then the notch/pin/whatever is. This makes sure that the designed alignment piece WILL fit in the hole, even if it has a little bit of wiggle room, since prints can vary from print to print even on the same machine.
A couple of variations that would increase 2 axis alignment would be to make the fin a slight s shaped arc or to break it into to smaller fins that are slightly angled opposite of each other.
I like to print my pins separately and use an octagonal shape along the long axis. Both plates have octagonal holes and the pins are printed on their sides.
All modern CAD programs have libraries of fasteners and openings to use them. Now we just need libraries of predesigned features like those shown here for mating parts, as well as simple parameters to size them and apply both in a single operation. Think of it as an extension of the Hole function, where you create the hole and attendant countersink all in a single step. One click should let you choose a mating style from every reasonably strong possibility and be able to set in place on both sides of the mating as features added to the design. This should even include splitting a model into two parts for assembly afterward by simply positioning the mating type and locations desired as part of the split process, such as locating them within a plane being used for the split. A simple see through view can let you see and position the desired new features before committing the split to render the then separate parts.
I've been binge watching the last few weeks, you have NO idea how much this channel has helped my design phase. Off topic, I'm hoping to learn more about your fulfilment side of things sooner than later. As someone that is currently stuck at home but not great at 3d printing, that would be huge to me.
I personally just print square pins/ jigs separately. Your solution introduces alignment issues. If you still want to print the fins together with the block, use those same fins with tapered ends, cross them and rotate them at 45degrees and print vertically. It should look like a square with an X in the center. You should get decent alignment with that.
By the way, on the slot where the X goes in, I would chamfer the inner corners of the X or make a round cavity to avoid problems slotting the X fins in.
How fins are worse than pins you can clearly see with the misalignment of the two pieces in the end of the video. Moreover these thin fins can be easily broken
I like your approach to trying to get people to focus on more approaches. With this, there are many approaches. Majority of the time pins are not the approach as they snap easy on 3d prints. Usually for me, slide (dovetail), or snap fits like cutting the edge and extruding the others but with a tight tolerance works best. 0 post processing and if you put a ridge on the edge then the two parts connect and break away in a very satisfying way. It does take design practice though. Being able to just follow your edges does in some cases improve the chances that you have a proper fit.
@Slant 3D however, assembly is unavoidable in some designs. Especially when the printer is not big enough or if a limited part count makes assembly harder instead of easier. Often the easier is better than fewer parts. Also pushing pin a into hole b is easier than screws. But screws are better for securing mated parts where there is no room to put a pin in hole connector.
This kind of reminds me of something I did for a cube I was making where I wanted something stronger than just pegs, so I used chamfered square posts (which are more trapezoidal now)
So you started the video recommending to print items like this horizontally on the bed whenever possible (which contradicts your other videos) and then your whole suggestion is to print this vertically?
My personal favourite is to create holes on both sides (on one side press-fit) and I insert a metallic pin or set screw. If necessary gluing on one side.
I usually just print the pins separately laying flat, then brush the mating surfaces with acetone before assembling. Otherwise I split the parts like a puzzle and add keys to lock in place, depending on how i need to assemble or what it will be used for
the chapfer idea is great... the other ones are not practical for parts that are reusable, i mean the parts that use alignment, the chapfer is strong, provide alignment and is strong and wont catch.
So the alternatives to pins were: - Funny-shaped (cone) pins - Fewer pins - Funny-shaped (fin) pins It's a shame, because this video was really well made. You should keep making videos, but consider hiring a writer or someone to help with ideas.
Interesting alternatives to typical metal dowel pin alignment 👍. Actually the pins need not be round to start with, if to print hexagonal pins laying down it should print fine without support. You can also make the length a little undersized. Chamfer on end faces both on holes and pins should help as well. Then it is just inserting these pins with a drop of cyanoacrylate(aka superglue), is all that is needed. Should makeup for a much stronger and durable part with least amount of overhead interms of effort/time. The approach, does not sacrifice extra space, or need to accomodate new design requiremts for things such as slots. It is also Poka-Yoke. Just some thoughts ....
You can also print them lying down. What may increase the print time there is that the normal slicer settings will often require multiple layers on the bottom to be 100% filled. But this setting can be overridden, even if it's just for the bottom. Also keep in mind layers are usually 0..2 mm and nozzles are 0.4, so it will take two layers to replace one wall.
IME cones and domes work best for 3D print mould registration print to print on a parting line. But for gluing finished parts together on the parting line I use tongue and groove through cuts to the end. Offsets on the male side of the tongue and the through cut allow epoxy to squeeze out during final mating. The adhesive squeeze-out is very important and the groove system creates more surface area.
To be honest, all of those options are vastly inferior to designing in a proper shadow line feature. If you want to teach people REAL mass production techniques that they can integrate into their 3D designs then shadow lines beat everyone of the options in this video on pretty much every point that matters, well constrained, repeatable, quick and easy to locate / assemble / disassemble, quick and easy to 3d print, strong, versatile.
Another pro tip. Is key your prints. If you have a single joint just identify a side to match. But if you have multiple sides. Use circles And squares and triangles to determine direction and attachment sides.
Would it be a good idea to make slots on both sides, and just use coins or washers of some kind to join them? More parts, but I feel like this would make for a less brittle fixture.
Fins could be made perpendicular like |- which could make it fit better with much less sliding possibility, basically if you try to slide it in one direction, the perpendicular fin will stop that from happening.
Super insightful! I don't do mass production 3D printing, but I do design and print things that involve interlocking parts. I'll keep this in mind next time I design something similar!
I'd still go with pins but cut a flat side on both the top and bottom (essentially a rectangle with two rounded sides) so it can print horizontally and flat on the build plate. Also adjust the holes to match the pin shape.
tl;dr reduce complexity (material) and thus points of failure. fins over pins, if need be. This is the bread and butter of mechanical design and engineering textbooks. Mechanical behavior: statics, dynamics, strength of materials, vibrations, reliability, and fatigue. Could be a made into an introduction series of a few YT videos, going over these basics through one example.
Do a tol stack. Know your tool and size your hole and pin accordingly. Perhaps hole and slot of the like. Very common method for proper component fits.
Just print both sides with holes, then print the 4 pins separately - solves all your issues. And stop making your pins cylindrical ! Make your holes and pins _square_ - the holes can sit at 45° for problem free printing in all orientations, and the pins can be printed flat for maximum strength.
Great tutorial, but if it's just for alignment, those fins only need to be 2mm tall, and maybe 4-5mm wide. Use 4 if it doesn't matter which way the lid is assembled. Use 3 if you want to constrain it to only fit one way. Making it possible to assemble 4 different ways might slow down manufacturing a tiny bit, but it speeds assembly. This is all if something else is holding them in position, and the features are just for alignment.
I still have plenty to learn about fusion 360-what is that move where you drew two circles to “undo” two of the pin extrusions without going back in the timeline? Didn’t look like you used the cut feature either
I think it should be noted that fins are nice, but they provide a very large interface surface between fin and hole, which reduces robustness as slight variations in any part of the fin hole make the entire part no longer fit. By definition robustness is the tolerance of something to variation, and in a manufacturing engineering context where large runs are expected, features like this would be avoided due to how much variation would impact final work. A good solution to this problem is to remove some more material from the fin holes so that the fins only contact the holes at a set number of places, like 2 or 3 max
I realized that solution is a bit confusing, the important part is that rather than making a full fin surface interface with a surface of equal size, change the interface to only physically touch the fin at a couple of places and have small air gaps elsewhere. Constraints remain the same, while tolerance to variation is much more kind
Instead of printing pins make holes on both side and insert steel pins on one side and just chamfer the steel pin end which will remain outside so that pins will go smoothly in holes... I did the same to make printhead holder magnetically detachable so that if I need to change from printhead to laser I just need ro pull out the printhead holder and unplug the connector and plug the connector of laser module which is attached on same kind of holder and just pop it in place... No screws and wiring problem..😊 its a simple design once you got the basic working of the thing in mind... And my holder was same like yours just a flat plate but with hotend holder and other with holes to screw the laser module and back plate was exactly same... and I did this around 5 years ago on my custom made printer which I built in 2017.. recently fusion ended up my free license which I got as education license so I am not into 3d printing right now as I love designing and printing my own designs not someone's designed models..
My first reaction to the chamfered pins was "he created nipples." If its purely for allignment and you don't need it to prevent movement hemispherical bumps and holes would be a simpler solution. If there's directionality to the alignment then an asymetric pattern is useful.
The time it would take to print that part horizontal then print that part vertical is drastic. Printing that part vertical would drastically increase the print time of that part.
when mass printing, you can stack a lot more on the plate in a vertical orientation, requiring much less human time to set up and clean off the machine (with the added benefit of layer line orientation.) attending to the machine once per shift vs 7 times per shift.
Print time is an irrelevant metric. That is point of print farms. Large parallelization of automated production so that the project takes as long with printing as with any other process.
@@slant3d print time correlates with rent and machine amortization costs. True for any machine. Just because a farm has many of them doesn't mean it doesn't matter. Sure, you can charge 1000$ per print on one machine even if it takes 1 hour, but then you're in luxury territory and no longer commodity manufacturing and no longer makes sense with the rest of marketing messaging.
@@slant3d how is print time an irrelevant metric? Especially for print farms lol. When you have hundreds of printers operating all making the same “mass production” design, 1 hour saved per print stacks. Your yield is drastically increased in the same amount of time, your utility costs go down bc you saved 1 hour per print per machine (less power needed to run the printers for less time) you also reduce the wear on the machines thus saving accumulative maintenance/repair costs. You also make more money by finishing that production wave sooner allowing you to make money sooner on the next designs. Why do you think old school factories and assembly lines focused on the literal amount of foot steps workers had to take, cutting out seconds and fractions of seconds anywhere they could? Print times are irrelevant to hobbyists printing a couple parts a month. I’m not even mad at the video. It’s a fair idea in specific use cases, but when your parts didn’t even align properly I had to come check the comments. If you are going to talk about quality and accuracy at least reprint the demo to align correctly for the video lmao. Not saying you don’t have the skill. Idk what it was, design error, printer error, printer calibration, etc. but to hype up the mass production time saved, stability, accuracy, print focused design choices, etc. and then make comments like print time is an irrelevant metric while using the scenario where print time matters most as your arguing point for its irrelevance is smooth brain and ultimately null and void if your outcome doesn’t even align properly 😆 I’ll check out some more of your videos and give you a fair shake (this is my first) but just one man’s opinion, do better man. Don’t argue with commenters on YOUR videos giving you play time (regardless if they’re right or wrong) about petty shit and then not even have the parts align dog. Post sand that shit, hide the misalignment with a better camera angle, something. You clearly have more skill than that outcome. I hope to see it in your other videos
His last point, have one fin. I like mirror parts, meaning 1 part design that fits into its mirror. So two bins of parts with 100 each. If one side is poor, then you need to reprint, meaning better yields. 100 with 10 bad parts on 1 side means 90 assemblies. Mirrored results in 94 competed. If this is stamped or molded, its one dye not 2.
Fusion 360 and Blender are two different types of software so you could not use one instead of the other once you start doing more advanced modeling. I'm really over simplifying here but it's a bit like comparing making a technical drawing with a ruler or by freehand. F360 is a CAD software and it is actually free for personal use. So is OnShape which is a major contender. FreeCAD is a completly free alternative but it has a different workflow and can be difficult to get used to. I've given up in frustration several times 😅 Blender is more used for "artistic" purposes and some comparable softwares are Unity and Maya.
Great content, but if you want accuracy, you dont use FDM for mass production. It's weak and takes forever. For a one time prints, you could use piece of wire or even pirchase small guide pins that come in useful.
Fin slant all the way to the surface allows for misalignment. You can actually see the misalignment whe the part is being held. Plus as you stated earlier in the video, dont print unnecessary features. You should only print 1 fin because it will fully constrain the part. Finally you gan fin rigidity by printing vertically. But you loose surgace finish and stength for the rest of the part. Always a balance, cant get something for nothing.
Good alternatives depending on design constraints and usability. I want to explore the use of bendable prints, you know a thin piece printed that connects the lid to the box and is flexible, I wanna see which design is the strongest and can be folded back and forth indefinitely.
Ok, here is my amateur tip of the day. Make just holes, and use the filament of the 3d printer as the mating pin. Also, the video is showing that 4 pins is bad, and then the second solution is to drop it to 2 pin, doesn't make any sense.
These pin chamfers are a construction mistake and don't make sense, because now you have something called a "Doppelpassung" in german (I couldn't find an english word for it). Basically you have two faces that act as a fit, which is impossible if you have any type of tolerance. Plastics are flexible so it might work in this case but you should at least mention it.
What is the actual point of this? If you are assembling parts you need them to stay together which means fasteners or snap features. If using fasteners, a tight fit clearance hole and a tapped hole or heat set insert should provide as much aligbment accuracy as printed pins. If not using fasteners, the snaps should be able to be designed such that they drive alignment. I cant think of a real use for this. Even your use case of a box would be better suited to a simple lip under the lid that engages the walls of the box.
@@bigcheese82 I use pins plenty. That isnt relevant. The whole point is that this was posted as a novel use of pins, but its novel because its one of the worst ways to accomplish the goal.
To say what’s the point smacks to me of not having tried to mass assemble stuff. Having location features so you can smack parts together is really helpful - as even if you were then to fire in a plastite screw as a fastener - these features ensure the holes for the plastite screw are aligned . Otherwise you have a much slower and tiring activity to match the two parts up to then put the screw in. Also - you can reduce the number of fasteners - as the fasteners keep the surfaces together - but you don’t need multiple screws to align the parts- hence saving assembly time screwing stuff in - and part count.
@@slant3d If you need 25,000 of something, you would surely just injection mould it.. I can't see how FDM can compete with the price, speed, dimensional accuracy, and robustness of an injection moulded part. Yes there's tooling cost, but at 25,000 parts that's likely to be totally negligible in the grand scheme of things.
I would never understand why the 3D printing community don't like screws... just leave the holes and use metric set screws. It would be cheaper, faster and way more reliable.
@@slant3d plastic parts are the largest point of failure... I assure you that any metal screw would last way more than any 3D printed part... and also a 3D printed hole is more reliable than any 3D printed pin or fin. So, just put a freaking 3D printed screw hole on one side and regular hole on the other side... install a simple set screw and you're done.
@@slant3d and again... a plastic 3D printed part is the biggest point of failure... It's pretty obvious that you don't work with other materials or process. Seriously, I have years in the manufacturing industry. I work with CNC machines, laser cutters, and 3D printers and the less reliable part are the 3D-printed parts. The only reason we use them, it's because it's cheap and fast, not because it's the best option. Using 3D printed pins and fins is just stupid. Those break with almost no effort. A hole is way more resilient than any pin/fin and adding metal inside (like a set screw) is pretty simple and add way more structural resistance. Unless you're stupid and want to use 3D printed screws, you're actually adding structural resistance by adding metal parts. You're designing for 3D printing, not for CNC metal parts where you actually use a single piece of metal, and it would be better to have pins.
Screws are for fastenings and pins are for aligning. They are not interchangeable (in most cases). The 3D printing community often uses screws when they make the most sense. But it is always a good thing if you can make a design that requires no hardware. That way you are completely self sufficient with your printer.
@@slant3d To hammer in 4 dowels doesnt take a lot of time. Especially considering printing the parts takes a lot longer. Even with a huge farm hammering in those dowels wouldnt take any longer than removing the part from the build plate. At industrial levels firms are likely to be using injection moulding anyway. I see your point though it is quicker to just print these parts but adding dowels will be more accurate and increase the strength of the connection. I just cant see why you wouldnt do that. Just because you can build the solution into the 3d printed part it doesnt mean you should.
@@slant3d Imagine 3d printing 250,000 parts... The volumes you're talking about is just unrealistic. You wouldnt 3D print that many, you'd use more appropriate methods of manufacture. At that scale you'd be using robots anyway.
Even with those 2 pins it's a bad design because you can just rotate the part 180 degrees and the parts will be aligned. Different size pins or no pins at all. Why not tapered triangles. It can be literally any shape as long as it's simple enough to produce so it doesn't cause any problems. Triangle instead of pin is great, because you can have different triangle shapes and those can even point a direction. Pins come from drilling and inserting steel pins. Why are people using it in fdm manufacturing? :D
no one ever said pins need to be round.... dont you remember Geometry class? There is such a thing as a diamond which is much more logical and common in the 3d printing industry yet you never mentioned that... I work in the manufacturing field and see diamond holes way more often on 3d printed parts than round holes
"They slot quickly and easily" Has to force both plates together leaving a gap and a visible offset.
People critics a lot, but what I like the most in your videos is you show tons of possibilities, and while your focus is production on large scale, you bring us into creativity along the way. We tend to forget that you just don't throw your answer in our face "go accept it" but you offer all the thought process that brought to this last design. And nobody is obliged to apply to their build, so personnally I feel grateful and a time well wasted to watch your videos.
You could literally see how the fins DID NOT align the two blocks together.... At least a good 1 or 2mm out from each other 😂
Steel dowel pins are very cheap, accessible, and are very accurate. They also don't have the risk of shear like printed pins do like you shown in the video. All you need to do is make one part with holes that have an interference fit, and the other to have holes that have a close clearance fit. You don't need to reinvent the wheel. There's a reason that method is used in traditional injection moulding equipment.... BECAUSE IT WORKS AND IS REPEATABLE!!!
I get the drive to think differently, but you need to also consider the fact that if something isn't broken, it likely doesn't need fixing...
Also, 3D printing is really only suitable for at most low quantity batch production parts. If you're doing 10's of thousands of parts, you'd be best to just go down the injection moulding route... You'd get parts far quicker, more robust, more accurate, nicer finish, and probably very similar in cost even with mould tooling design and manufacture included.
Adjust dimensions to your application.
this part could obviously be best suited for dowel pins and injection molding. You need to look 1mm below the surface that OTHER applications can not be injection molded and still need alignment pins.
Making two fins with these huge chamfers doesn't really constrain them vertically though. They will just slide out if pushed from the top or bottom. You might want to angle them 45 deg to each other which then works perfectly with printing on the chamfered corner like you showed in a previous video. edit: also changes the layer direction to a better angle so you don't risk breaking them right in plane
That is only feasible if there is the area to support them. Generally features like these would be made smaller and be on the outer edge of an electrical enclosure. It is also ideal to chamfer the inner edge of the slot as well so that part is constrained in both Axis.
Yeah, you can literally see the two demo pieces not even align correctly on the other axis
I agree and will try out a V or Y configuration
Fr reinventing the wheel for no reason and the replacement is worse😂
@@OArchivesX Not trying to start drama, I'm genuinely curious since I've been 3D printing and always enjoy learning new things for my hobby. Are there better ways to do this than pins and holes? I would like to learn more if you have any resources or videos.
teaching tech has a great video on shadow lines which is my preferred method for this.
great to get people thinking about more then just using pins and holes (A method common in furniture kits because dowels and drilling holes is cheap in that context) and thinking more critically about what they need from their registration features
If the rib is thinner than the straight pin, it is weaker in that direction. The rib also has a lot more surface area, so more opportunity for error, or over-constraining the locating feature, causing improper fit (as seen in the video).
Definitely a few ideas here that will generate difficult assembly and improper fits
You could decrease print time and improve manufacturability by using just holes on both sides. And then using a separate pin. I like to make my location holes the diameter of my filament and then I can use some pieces of filament as pins. If I have just a short piece of filament that doesn't fit on a roll I will cut that up for pins. I only use that method for parts that I'm going to be gluing together. If I were making a lid or something like that I would go with a lip all the way around the perimeter.
I usually just put holes on both sides, and use pieces of cut filament as the pins. It's much stronger than printed pins and while I'm not sure how it compares to fins, it's probably good enough for most applications and pretty fast to do.
That is an option. But it adds extra assembly time when the features could be integrated into the print to make it more scalable. Imagine manually adding pins 10,000 times.
@@slant3d on that scale though, we're beginning to venture into territories, where 3d-printing is just not economically viable anymore
This is the way
yes, or print some dowels that you can friction-fit in the holes.
@@ichbrauchmehrkaffee5785Where does the scale line start and end?
The textbook way to do it without over constraining things would be to have 2 pins in one part, one hole and one slot on the other. The pin in hole gives you xyz and then the pin in slot gives you the clocking, anything more is overconstraining it.
Hmmm… how come when you showed the two parts put together, they didn‘t align properly then?😂
ok I'm not blind then🤣 content is still informative so i like it
Thank you for showing us your process. Some of the comments here are negative, but for new users this is eye-opening.
You actually had me at cones. I think thats a fantastic change thay ill actually implement going forward. But then you lost me at this thin flimsy fins, perhaps an x shape would be best for slots, but i think you hit the nail on the head with the cone. Can Print in both directions and has equal strenth in all directions. Circles are your friend!
Very True. Thanks for watching
Pins work fine if your using them as an alignment for walls or something that is going to be superglued together. This also helps make said model stronger when gluing together as there is now more surface area for the glue to bind to and less likely the model will come apart. I usually use notches or some sort of means to align my designs together, usually corners of a wall of a model building. This makes thing a little stronger and easier to put together, while making the model itself easier and faster to print, since its all laid out flat on the bed. Due to printing inaccuracies that happen, I always error on the larger side for any holes, usually about .5mm bigger in X/Y then the notch/pin/whatever is. This makes sure that the designed alignment piece WILL fit in the hole, even if it has a little bit of wiggle room, since prints can vary from print to print even on the same machine.
A couple of variations that would increase 2 axis alignment would be to make the fin a slight s shaped arc or to break it into to smaller fins that are slightly angled opposite of each other.
That is alright if you have the space. Alternatively you can chamfer the inner edge of the slot to abutt the finned chamfer
I like to print my pins separately and use an octagonal shape along the long axis. Both plates have octagonal holes and the pins are printed on their sides.
But that is not optimized for mass production because it adds several assembly steps that add cost to the part
All modern CAD programs have libraries of fasteners and openings to use them. Now we just need libraries of predesigned features like those shown here for mating parts, as well as simple parameters to size them and apply both in a single operation. Think of it as an extension of the Hole function, where you create the hole and attendant countersink all in a single step.
One click should let you choose a mating style from every reasonably strong possibility and be able to set in place on both sides of the mating as features added to the design. This should even include splitting a model into two parts for assembly afterward by simply positioning the mating type and locations desired as part of the split process, such as locating them within a plane being used for the split. A simple see through view can let you see and position the desired new features before committing the split to render the then separate parts.
That would be great
@@OArchivesX Not everyone is a trained engineer. Increasingly AI is going to push into CAD to enable untrained users to do more.
I've been binge watching the last few weeks, you have NO idea how much this channel has helped my design phase.
Off topic, I'm hoping to learn more about your fulfilment side of things sooner than later. As someone that is currently stuck at home but not great at 3d printing, that would be huge to me.
Thanks for watching. Glad it is helpful
What an excellent video!
Just brilliant!
Having the skills to design parts for 3D printer manufacturing requires a completely new mindset.
Very True
Of course you would say that. He found an excuse to 3d print nipples.😂
@@ianbelletti6241 Very True
I personally just print square pins/ jigs separately. Your solution introduces alignment issues. If you still want to print the fins together with the block, use those same fins with tapered ends, cross them and rotate them at 45degrees and print vertically. It should look like a square with an X in the center. You should get decent alignment with that.
By the way, on the slot where the X goes in, I would chamfer the inner corners of the X or make a round cavity to avoid problems slotting the X fins in.
How fins are worse than pins you can clearly see with the misalignment of the two pieces in the end of the video. Moreover these thin fins can be easily broken
I like your approach to trying to get people to focus on more approaches. With this, there are many approaches. Majority of the time pins are not the approach as they snap easy on 3d prints. Usually for me, slide (dovetail), or snap fits like cutting the edge and extruding the others but with a tight tolerance works best. 0 post processing and if you put a ridge on the edge then the two parts connect and break away in a very satisfying way. It does take design practice though. Being able to just follow your edges does in some cases improve the chances that you have a proper fit.
One thing people commonly do is have holes only and use scrap filament to lock parts via those holes. This can be done for hinges too.
Yes. But that is not viable for mass production because it is an additional assembly step
@Slant 3D however, assembly is unavoidable in some designs. Especially when the printer is not big enough or if a limited part count makes assembly harder instead of easier. Often the easier is better than fewer parts. Also pushing pin a into hole b is easier than screws. But screws are better for securing mated parts where there is no room to put a pin in hole connector.
I like the three pins. Can’t mess up orientation!
This kind of reminds me of something I did for a cube I was making where I wanted something stronger than just pegs, so I used chamfered square posts (which are more trapezoidal now)
The chamfered pins is just what I need for the part I’m working on. Going to try it out right now!
Glad it was helpful
So you started the video recommending to print items like this horizontally on the bed whenever possible (which contradicts your other videos) and then your whole suggestion is to print this vertically?
My personal favourite is to create holes on both sides (on one side press-fit) and I insert a metallic pin or set screw. If necessary gluing on one side.
That is a fine solution. But the extra labor increases the cost fo the parts in mass production
I usually just print the pins separately laying flat, then brush the mating surfaces with acetone before assembling. Otherwise I split the parts like a puzzle and add keys to lock in place, depending on how i need to assemble or what it will be used for
the chapfer idea is great... the other ones are not practical for parts that are reusable, i mean the parts that use alignment, the chapfer is strong, provide alignment and is strong and wont catch.
So the alternatives to pins were:
- Funny-shaped (cone) pins
- Fewer pins
- Funny-shaped (fin) pins
It's a shame, because this video was really well made. You should keep making videos, but consider hiring a writer or someone to help with ideas.
I don’t know what mass produced item will actually use something like this
2 pins with the chamfer, perfect. No need to go further with these fins. Just as breakable as pins
I use holes on both sides and use leftover lengths of filament for my pins. This doesn't make ease of assembly easier, but it does reduce waste.
I think the alignment nipples is the best idea. Don't know about that fin. Good to know.
I love these videos
Interesting alternatives to typical metal dowel pin alignment 👍. Actually the pins need not be round to start with, if to print hexagonal pins laying down it should print fine without support. You can also make the length a little undersized. Chamfer on end faces both on holes and pins should help as well. Then it is just inserting these pins with a drop of cyanoacrylate(aka superglue), is all that is needed. Should makeup for a much stronger and durable part with least amount of overhead interms of effort/time. The approach, does not sacrifice extra space, or need to accomodate new design requiremts for things such as slots. It is also Poka-Yoke. Just some thoughts ....
I love this channel. All the design question I've had (and mistakes I've made) in one place.
Nice!
I usually just slap in pyramids and their inverted counterparts for the same reason.
You can also print them lying down. What may increase the print time there is that the normal slicer settings will often require multiple layers on the bottom to be 100% filled. But this setting can be overridden, even if it's just for the bottom. Also keep in mind layers are usually 0..2 mm and nozzles are 0.4, so it will take two layers to replace one wall.
IME cones and domes work best for 3D print mould registration print to print on a parting line. But for gluing finished parts together on the parting line I use tongue and groove through cuts to the end. Offsets on the male side of the tongue and the through cut allow epoxy to squeeze out during final mating. The adhesive squeeze-out is very important and the groove system creates more surface area.
Excellent video
To be honest, all of those options are vastly inferior to designing in a proper shadow line feature. If you want to teach people REAL mass production techniques that they can integrate into their 3D designs then shadow lines beat everyone of the options in this video on pretty much every point that matters, well constrained, repeatable, quick and easy to locate / assemble / disassemble, quick and easy to 3d print, strong, versatile.
I usally make a cross in the middle, it can print standing on the side with only support at the bottom.
Another pro tip. Is key your prints. If you have a single joint just identify a side to match. But if you have multiple sides. Use circles And squares and triangles to determine direction and attachment sides.
I like this, Very simple, good job, tanks!
Great video mate, very nice concept 👍🇦🇺😊
Glad you enjoyed
Great tip there
Would it be a good idea to make slots on both sides, and just use coins or washers of some kind to join them? More parts, but I feel like this would make for a less brittle fixture.
You always want to minimize extra parts.
@@slant3d Awwww, okay 😿
Fins could be made perpendicular like |- which could make it fit better with much less sliding possibility, basically if you try to slide it in one direction, the perpendicular fin will stop that from happening.
Super insightful! I don't do mass production 3D printing, but I do design and print things that involve interlocking parts. I'll keep this in mind next time I design something similar!
Thanks for these videos
Glad you like them!
you talk about printing initial pin-containing design either horizontally or vertically only. how about printing the whole thing at an angle?
Was considering taking you up on the Etsy thing, but every time I see your prints the quality is 💩
Draft parts require draft quality. Cosmetic parts require cosmetic quality
This series has to be satire.. at this point I don't see how it isn't.
I'd still go with pins but cut a flat side on both the top and bottom (essentially a rectangle with two rounded sides) so it can print horizontally and flat on the build plate. Also adjust the holes to match the pin shape.
good ideas
Nice video.
Thanks!
tl;dr reduce complexity (material) and thus points of failure. fins over pins, if need be.
This is the bread and butter of mechanical design and engineering textbooks. Mechanical behavior: statics, dynamics, strength of materials, vibrations, reliability, and fatigue.
Could be a made into an introduction series of a few YT videos, going over these basics through one example.
Thanks. Lots more basics videos like these on the channel.
Do a tol stack. Know your tool and size your hole and pin accordingly. Perhaps hole and slot of the like. Very common method for proper component fits.
printing pins separately might be another good alternative.
If you want to glue the piece, you can just put in two roles (one in each piece) and use a wood pin or printed pin in the middle.
Just print both sides with holes, then print the 4 pins separately - solves all your issues.
And stop making your pins cylindrical ! Make your holes and pins _square_ - the holes can sit at 45° for problem free printing in all orientations, and the pins can be printed flat for maximum strength.
Great tutorial, but if it's just for alignment, those fins only need to be 2mm tall, and maybe 4-5mm wide. Use 4 if it doesn't matter which way the lid is assembled. Use 3 if you want to constrain it to only fit one way. Making it possible to assemble 4 different ways might slow down manufacturing a tiny bit, but it speeds assembly. This is all if something else is holding them in position, and the features are just for alignment.
I still have plenty to learn about fusion 360-what is that move where you drew two circles to “undo” two of the pin extrusions without going back in the timeline? Didn’t look like you used the cut feature either
Great video! Has anyone told you that you sound like Kevin Nelon?
I think it should be noted that fins are nice, but they provide a very large interface surface between fin and hole, which reduces robustness as slight variations in any part of the fin hole make the entire part no longer fit. By definition robustness is the tolerance of something to variation, and in a manufacturing engineering context where large runs are expected, features like this would be avoided due to how much variation would impact final work. A good solution to this problem is to remove some more material from the fin holes so that the fins only contact the holes at a set number of places, like 2 or 3 max
I realized that solution is a bit confusing, the important part is that rather than making a full fin surface interface with a surface of equal size, change the interface to only physically touch the fin at a couple of places and have small air gaps elsewhere. Constraints remain the same, while tolerance to variation is much more kind
Nice... I do a square cube dowel... Never considered circular pyramid...
Take it one step further, a single fin that is a bit thicker and runs diagonally across the part. Just like your logo. ;)
Excellent.
Glad you enjoyed
Instead of printing pins make holes on both side and insert steel pins on one side and just chamfer the steel pin end which will remain outside so that pins will go smoothly in holes... I did the same to make printhead holder magnetically detachable so that if I need to change from printhead to laser I just need ro pull out the printhead holder and unplug the connector and plug the connector of laser module which is attached on same kind of holder and just pop it in place... No screws and wiring problem..😊 its a simple design once you got the basic working of the thing in mind... And my holder was same like yours just a flat plate but with hotend holder and other with holes to screw the laser module and back plate was exactly same... and I did this around 5 years ago on my custom made printer which I built in 2017.. recently fusion ended up my free license which I got as education license so I am not into 3d printing right now as I love designing and printing my own designs not someone's designed models..
4:58 - it turned out not centered
My first reaction to the chamfered pins was "he created nipples." If its purely for allignment and you don't need it to prevent movement hemispherical bumps and holes would be a simpler solution. If there's directionality to the alignment then an asymetric pattern is useful.
Do you ever use snap-together mechanisms? 3D printing can make those shapes
Sure
The time it would take to print that part horizontal then print that part vertical is drastic. Printing that part vertical would drastically increase the print time of that part.
when mass printing, you can stack a lot more on the plate in a vertical orientation, requiring much less human time to set up and clean off the machine (with the added benefit of layer line orientation.) attending to the machine once per shift vs 7 times per shift.
Print time is an irrelevant metric. That is point of print farms. Large parallelization of automated production so that the project takes as long with printing as with any other process.
@@slant3d Tell that to the client that gets charged by the hour hhahah :-D
@@slant3d print time correlates with rent and machine amortization costs. True for any machine. Just because a farm has many of them doesn't mean it doesn't matter. Sure, you can charge 1000$ per print on one machine even if it takes 1 hour, but then you're in luxury territory and no longer commodity manufacturing and no longer makes sense with the rest of marketing messaging.
@@slant3d how is print time an irrelevant metric? Especially for print farms lol. When you have hundreds of printers operating all making the same “mass production” design, 1 hour saved per print stacks. Your yield is drastically increased in the same amount of time, your utility costs go down bc you saved 1 hour per print per machine (less power needed to run the printers for less time) you also reduce the wear on the machines thus saving accumulative maintenance/repair costs. You also make more money by finishing that production wave sooner allowing you to make money sooner on the next designs. Why do you think old school factories and assembly lines focused on the literal amount of foot steps workers had to take, cutting out seconds and fractions of seconds anywhere they could? Print times are irrelevant to hobbyists printing a couple parts a month. I’m not even mad at the video. It’s a fair idea in specific use cases, but when your parts didn’t even align properly I had to come check the comments. If you are going to talk about quality and accuracy at least reprint the demo to align correctly for the video lmao. Not saying you don’t have the skill. Idk what it was, design error, printer error, printer calibration, etc. but to hype up the mass production time saved, stability, accuracy, print focused design choices, etc. and then make comments like print time is an irrelevant metric while using the scenario where print time matters most as your arguing point for its irrelevance is smooth brain and ultimately null and void if your outcome doesn’t even align properly 😆 I’ll check out some more of your videos and give you a fair shake (this is my first) but just one man’s opinion, do better man. Don’t argue with commenters on YOUR videos giving you play time (regardless if they’re right or wrong) about petty shit and then not even have the parts align dog. Post sand that shit, hide the misalignment with a better camera angle, something. You clearly have more skill than that outcome. I hope to see it in your other videos
The easiest thing I've found is to make corresponding 1.9mm holes on each surface, then use filament as a pin.
His last point, have one fin. I like mirror parts, meaning 1 part design that fits into its mirror. So two bins of parts with 100 each. If one side is poor, then you need to reprint, meaning better yields. 100 with 10 bad parts on 1 side means 90 assemblies. Mirrored results in 94 competed. If this is stamped or molded, its one dye not 2.
What are advantages of using expensive software like 360 over something free, like Blender?
Fusion 360 and Blender are two different types of software so you could not use one instead of the other once you start doing more advanced modeling. I'm really over simplifying here but it's a bit like comparing making a technical drawing with a ruler or by freehand.
F360 is a CAD software and it is actually free for personal use. So is OnShape which is a major contender. FreeCAD is a completly free alternative but it has a different workflow and can be difficult to get used to. I've given up in frustration several times 😅
Blender is more used for "artistic" purposes and some comparable softwares are Unity and Maya.
Great content, but if you want accuracy, you dont use FDM for mass production. It's weak and takes forever. For a one time prints, you could use piece of wire or even pirchase small guide pins that come in useful.
FDM mathes any other process in scale and speed
Your "Best option" only constrains in 1 Dimension!
Pins in 2 Dimensions!
The "Best option" in engineering depends on the specific requirements.
great video. have u considered magnets?
Magnets would require an extra manufacturing step of inserting those and would probably increase manufacturing cost by quite a bit.
Correct. If the function is not needed you do not want to add additional pieces.
Fin slant all the way to the surface allows for misalignment. You can actually see the misalignment whe the part is being held. Plus as you stated earlier in the video, dont print unnecessary features. You should only print 1 fin because it will fully constrain the part. Finally you gan fin rigidity by printing vertically. But you loose surgace finish and stength for the rest of the part. Always a balance, cant get something for nothing.
Good alternatives depending on design constraints and usability.
I want to explore the use of bendable prints, you know a thin piece printed that connects the lid to the box and is flexible, I wanna see which design is the strongest and can be folded back and forth indefinitely.
Ok, here is my amateur tip of the day. Make just holes, and use the filament of the 3d printer as the mating pin. Also, the video is showing that 4 pins is bad, and then the second solution is to drop it to 2 pin, doesn't make any sense.
These pin chamfers are a construction mistake and don't make sense, because now you have something called a "Doppelpassung" in german (I couldn't find an english word for it). Basically you have two faces that act as a fit, which is impossible if you have any type of tolerance. Plastics are flexible so it might work in this case but you should at least mention it.
What I'm gonna do is I'm gonna. What you wanna do is you wanna. So tired of these Bob Villa filler phrases in tutorials.
The best option for the lid is missing completely: making a "solid" stepped lid. Or a box with corners where the initial pins are located.
He's not talking about just boxes. He's talking about alignment devices that can be used with many projects.
You never explained in which cases pins (or else) are needed and how does it looks when pins are needed.
you made the fins the same depth as the slot, so you left a gap between the two plates
What is the actual point of this? If you are assembling parts you need them to stay together which means fasteners or snap features. If using fasteners, a tight fit clearance hole and a tapped hole or heat set insert should provide as much aligbment accuracy as printed pins. If not using fasteners, the snaps should be able to be designed such that they drive alignment. I cant think of a real use for this. Even your use case of a box would be better suited to a simple lip under the lid that engages the walls of the box.
It's because people use pins all the time, whether or not you do is irrelevant. This is a video for those people
@@bigcheese82 I use pins plenty. That isnt relevant. The whole point is that this was posted as a novel use of pins, but its novel because its one of the worst ways to accomplish the goal.
Thankyou I have learned so much from watching your videos
I’d probably press in dowels anyway.
To say what’s the point smacks to me of not having tried to mass assemble stuff. Having location features so you can smack parts together is really helpful - as even if you were then to fire in a plastite screw as a fastener - these features ensure the holes for the plastite screw are aligned . Otherwise you have a much slower and tiring activity to match the two parts up to then put the screw in. Also - you can reduce the number of fasteners - as the fasteners keep the surfaces together - but you don’t need multiple screws to align the parts- hence saving assembly time screwing stuff in - and part count.
Is this supposed to be dumb ? Cause it is
'mass production 3d printing' really curious what you mean by this. Needs nomenclature adjustment
25,000+ parts
@@slant3d If you need 25,000 of something, you would surely just injection mould it.. I can't see how FDM can compete with the price, speed, dimensional accuracy, and robustness of an injection moulded part. Yes there's tooling cost, but at 25,000 parts that's likely to be totally negligible in the grand scheme of things.
I would never understand why the 3D printing community don't like screws... just leave the holes and use metric set screws. It would be cheaper, faster and way more reliable.
Screws are one of the single larges points of failure in any machine.
@@slant3d plastic parts are the largest point of failure... I assure you that any metal screw would last way more than any 3D printed part... and also a 3D printed hole is more reliable than any 3D printed pin or fin.
So, just put a freaking 3D printed screw hole on one side and regular hole on the other side... install a simple set screw and you're done.
Screws are one of the single largest points of failure in any machine.
@@slant3d and again... a plastic 3D printed part is the biggest point of failure... It's pretty obvious that you don't work with other materials or process.
Seriously, I have years in the manufacturing industry. I work with CNC machines, laser cutters, and 3D printers and the less reliable part are the 3D-printed parts. The only reason we use them, it's because it's cheap and fast, not because it's the best option.
Using 3D printed pins and fins is just stupid. Those break with almost no effort. A hole is way more resilient than any pin/fin and adding metal inside (like a set screw) is pretty simple and add way more structural resistance.
Unless you're stupid and want to use 3D printed screws, you're actually adding structural resistance by adding metal parts.
You're designing for 3D printing, not for CNC metal parts where you actually use a single piece of metal, and it would be better to have pins.
Screws are for fastenings and pins are for aligning. They are not interchangeable (in most cases).
The 3D printing community often uses screws when they make the most sense. But it is always a good thing if you can make a design that requires no hardware. That way you are completely self sufficient with your printer.
put holes in both side and use external pins (i use cutted nails) , problem solved
Metal dowel pins are cheap as hell and will be stronger and quicker to use.
And they add extra labor and hterefore extra cost to a part
@@slant3d To hammer in 4 dowels doesnt take a lot of time. Especially considering printing the parts takes a lot longer. Even with a huge farm hammering in those dowels wouldnt take any longer than removing the part from the build plate.
At industrial levels firms are likely to be using injection moulding anyway.
I see your point though it is quicker to just print these parts but adding dowels will be more accurate and increase the strength of the connection. I just cant see why you wouldnt do that.
Just because you can build the solution into the 3d printed part it doesnt mean you should.
Imagine hammering in dowel pins 1 million times.
@@slant3d Imagine 3d printing 250,000 parts... The volumes you're talking about is just unrealistic. You wouldnt 3D print that many, you'd use more appropriate methods of manufacture. At that scale you'd be using robots anyway.
They are volumes we do all the time.
Even with those 2 pins it's a bad design because you can just rotate the part 180 degrees and the parts will be aligned.
Different size pins or no pins at all. Why not tapered triangles. It can be literally any shape as long as it's simple enough to produce so it doesn't cause any problems.
Triangle instead of pin is great, because you can have different triangle shapes and those can even point a direction.
Pins come from drilling and inserting steel pins. Why are people using it in fdm manufacturing? :D
no one ever said pins need to be round.... dont you remember Geometry class? There is such a thing as a diamond which is much more logical and common in the 3d printing industry yet you never mentioned that... I work in the manufacturing field and see diamond holes way more often on 3d printed parts than round holes
1:53 the way you cropped them out …. Why LOL just click and press delete
does not work as it only limits in one dimension as seen in the NOT fitting last shot
Those aren’t pins. Pins would be removable. Those are bosses.
Better off putting in a shadow line.