My biggest issue is fitment. I swear no one leaves tolerance in their models that have parts to fit together. I'm over here sanding sanding sanding, hammering hammering hammering, pushing pushing pushing.
I can totally relate! I recently printed a storage for my camera batteries. It took three different remixes and then time scaling until it fitted. The measuring tool in the Slicer is almost mandatory at this point.
That helps. Different materials and printers make the behavior impossible to predict unless you tested it before so more planed tolerances make for universally useable models.
I totally agree. I have spent days printing two different sets of drawers for my Ender 3 Pro, both from different contributors, and both were too tall, hitting a screw in the bottom of the build plate. So frustrating!
I have a friend whom I helped out with her thesis project. She had a team partner who was in charge of designing all the parts for the robot they did, and I swear it was a nightmare every time she asked me to print something because the guy had literally zero notion about additive manufacturing, let alone 3D printing (he never got his Ender 3 running, for some reason). And I don't know if it's a coincidence or what, but you mentioned basically all the design mistakes he made and I had to correct or account for in the slicer, so the parts were actually printable. This video is gold and needs way more exposure.
In a case where there is that guy who doesn't know s..t about 3D printing, still sends you parts to be printed and never learns, I have a special treatment. I just remove the Z support gap and laugh maniacally while the guy is trying to remove those damn supports! I am probably a bad person to do that to someone, but if that someone keeps repeating exactly the same mistakes, what can I do?
That's the spirit! I once copied the design of a bike wall holder since there was no source file and uploaded everything. People thanked me so many times for the source since you could fit it to any wheel size.
I only upload STP files. And because of this fact, I asked one designer to also offer STP files, because I wanted to update his model to a new version of the object inside of it.
How to print a sharp inside corner. The trick is a "mouse bite". Intersect a small cylinder with the corner so the STL shows a cutout up the corner like a small ice cream scoop was run up it. When the nozzle goes around that small cut, the squeeze out flows and closes it in, leaving a sharp corner. The first time I used this trick I was modeling a replacement part that had a rectangle hole which had to fit onto a metal tab and a square hole into which snapped an injection molded plastic part. I did a TLAR (That Looks About Right) on the size of the corner cuts, sliced off those parts of the model to do a test fit and they were both perfect the first time. The concept comes from fonts designed for printing on cheap paper where ink bleeding is a problem. The Bell Centennial font was designed specifically for printing phone books. To improve its legibility at small sizes, all inside corners of the letters have notches so when the ink bleeds into the paper it flows into the notches and makes a sharp corner.
Way to go! When I was a total beginner I always thought the bad print is my fault. It is printable with very fine layers but that's not really a good solution.
@@SmallBatchFactorycura has a great feature called "make overhangs printable." This actually alters the geometry of the model so nothing is more than a 45 degree overhang (or possibly whatever you have your overhang angle set to in support settings, not sure). If the only problem you have is badly designed rounded chamfers, that should fix it.
Or you could do what I do: make a curved bottom edge that starts at a 40°-45° slope and curves upward to vertical. This prints well and looks better than a simple chamfer, but it's a bit more work to figure out how much to lower the curve center.
I like that approach as well. It gives me peace of mind. I also often run a screws perpendicular to the layer lines to compress the print. Makes them virtually unbreakable.
Everyone needs to see this. I already incorporate a lot of these in my designs. One more thing, on the bottom of my models touching the build plate I will usually add a 30 degree chamfer typically 0.4mm high (or two layers for most slicer profiles) to prevent elephants foot. This way people who download my model won't have to compensate for it in their slicers, as a lot of inexperienced people might not know how to compensate for that. 30 degrees keeps the stepover small enough so there's no chance of sagging lines even if people don't squish their first layer.
Good! With elephants foot there's always the issue that it's highly dependant on your printer. A chamfer is a good start. My Prusa default profile for example is not enough for how close I print my first layer.
I print a lot of miniatures and my biggest issue is details where there doesn't need to be any ... Like a tongue inside a closed mouth !!! Looking at you HeroForge 😤
I guess they design the model once for different poses and don't really care how you're supposed to print it. Only a few models like Adalinda the dragon are made with printing in mind and work without any supports.
@@SmallBatchFactory yeah! It's certainly not an over exposed topic, I don't think anybody is tired of watching people use mesh mixer. So many uses. I saw somebody scan their face and clean up the model in meshmixer to print a mask.
I saw the thumbnail in my "recommended" and thought "I wonder what this is all about". I was surprised to see that key hanger. It's my design! I was both excited and humbled to be featured. For some context, that was one of my first designs in OnShape that I made to attempt a 3D print for my new printer. I know it's not great but I liked it and so I uploaded it. As it is, I printed it without supports but, admittedly, it didn't print real well. I wasn't bothered so I kept it. As for strength, it seems fine as it's only holding car keys. I've been using that particular print hanging on my refrigerator for about a year with no problems. I did update the design with the holes at the back to insert magnets. There were just holes for hooks before. That was about it. All that said, I'm happy to be used as a "what not to do" example. All of us are students and no one of us is as smart as all of us! Input, feedback, knowledge sharing, experience, etc. are all valuable for anyone learning something new or expanding their knowledge of a given subject. Thanks for the lesson!
I'm glad you liked it! Such a small world after all. In many cases "good enough" totally gets the job done anyway. I would lie if I said my early designs weren't a mess either, figured the part about fillets on the build plate the hard way. Somebody even one uped my suggestions with a better one I featured in a community post. We're all just students :-)
Brilliant video. As an engineer, I gave up on using functional prints from any online source due to the reasons you put forward in your video. I mainly print sculpts or miniatures now and if I ever need a functional part, I quickly design it myself. :D
Definitely! I view Thingiverse as "serving suggestion" and a source on inspiration. It usually even takes less time to do it yourself than finding something fitting...
With the hook design, fixing the curvature to make it an angle cut is a good start. But if you want to increase the strength without printing in parts, you want to add a large fillet at the bottom edge of the hook and a somewhat large smaller fillet on the top edge. From my experience, shear forces are never the reason for failure. When you load the hook downward, the failure is caused by the deformation of the hook, this causes the hook to bend forwards, so now the angle of the load vector changes and there's a force component exerting force perpendicular to the print layers. This PULLS the layers out. So a large fillet below bottom edge would work to reduce deformation as a fillet in the loading position acts like an arch and arches resist compression. This technique has worked really well for me so far.
You're absolutely right. That's under intended use. What I expect is somebody getting caught with a shoulder or something while passing the holder and snap of a whole hook.
Good advice! Steep angles on the first layers also grinds my gears. A note in terms of self-presentation: I find myself distracted by your head constantly bobbing to the left and then to the right.
You didn't move your head to much. You were being you and the video was perfect. Whoever was paying attention to your head moving was not here for the actual content of the video.
@patrickcarpenter6258 thank you for your kind words! I guess there is a middle ground I still have to find. After all it's only my 8th video ever. Besides, my wife also said it's too much and you know what that means :-)
@@patrickcarpenter6258 "Whoever was paying attention to your head moving was not here for the actual content of the video." - what a false assumption. I'm here for the content but this head banging is so distracting like someone poking me with a stick every few seconds. Hard to focus on the content. I understand it's my problem that something distracts me but looks like I'm not the only one. Still, good content anyway!
Good selection of painful design choices. Guilty of a few of these myself sometimes when in a hurry and I just need it now, but always regret it later. Another one that would be next on my list is using 3d printed things when a widely available non 3d printed thing is readily available. I was doing this myself with clips to attach a crutch to my wheelchair. I had clips for the crutch and for the chair frame. It was always annoying when the whole clip assembly would come off instead of just the crutch. I had to facepalm and realize that 3d printed clips to attach it to the wheelchair was dumb. Just adding a few slots instead for zipties to go through made it far more functional and strong.
Thanks! We're all guilty, I was too. Combining standard hardware is a real game changer. Be it screws, zip ties or something else. The Voron team for example combines cheap small idlers with printed parts to make bigger ones that would cost over 10 bucks each.
You could make the key rack modular. Each section is one hook and back plate. Each piece prints on its side with the hook on the build plate and has provisions to snap the next piece in next to it. Then you can make it as long as your want, the hooks are integral and as strong as possible, you can add to it or replace sections any time, you could customize each section like by adding labels, and you just have to print end caps for the right and left sides to finish it off with no holes.
@@SmallBatchFactory Yeah I just like sharing additional ideas in the comments. Always looking for ways to improve a design for printing, customize it for individual desires, and possibly add functionality if useful. Fun thought exercises. :) Keeping it simple is often the best approach for real usage though.
@@logicalfundy Definitely true. I like to try to think of how it would work both ways and go with what makes sense. Learning how to design for 3D printing is still an evolving art for most.
I came across one, where I noticed a major design flaw after the print had failed because of a combination of that flaw and poor bed adhesion (and the poor adhesion was not something I normally had problems with.) The item was a flat molle-patch with a cutout logo. But that logo contained the letters A and R which don't work for a cutout. The tiny, unconnected pieces within the letters should have been omitted or designed differently. It was obvious to me, that the user had never even printed that model.
Those are common problems. Prusa has a printed plate with their logo so it does work, but the letters need to be large enough You can add a box volume in the Slicer to fill those voids though. An easy way to get rid of unwanted markings in those models.
Nice video. Suggestion for the last one: instead of making the hooks press-fit, flare the base of the hooks, add a recess on the back of the base plate for the hooks to sit in and pass the hooks through the holes from the rear before installation. Then, when you mount the base plate to the wall, the hooks are held in place mechanically between the wall and the plate. Since you want a flat side on the hooks anyway, you can make the recess match to key orientation.
Saying fuck unrelated sponsors is the second most appealing reason to subscribe. Number one is the top notch facts and content delivered in short and precise manner! Good job.
Fillet on the base thing is something everyone should learn when they print something with it once or twice and realise "oh that does not print well" or you spot it when you slide the layers from the bottom to the top and notice very steep overhangs.
Yes, they should. When we printed face shields people handed it prints that already broke when looking at them so I suppose the fillet is the least of some people's problems.
I am at a university and we have a bunch of 3D printers for public use, my biggest pet peeve is by far wasted material. Not only do I see people needlessly printing with high infills, the parts being designed seem to have no notion towards how they will be produced! I constantly see huge boxes of plastic being printed that would function identically with 95% of the material removed in the CAD model. Whenever I design parts to be printed for my FSAE team, my number one priority is to use as little material as possible, and most of the time I can get away with 0 support material by introducing snap fit or glue together parts like you mentioned at the end of the video, plus it is a great way to start introducing the viability of AM to the senior team members if they get to see and handle practically designed parts.
So much wasted material... I'm a huge advocate of efficiency and optimization. I even print a lot of models with no infill at of I don't need to. There are so many great ways to optimize the design like screws that are flat in top and bottom for example. Maybe I should do a part 2 and show things like this.
I'm not a video producer, but these peeves have been making me want to make a video similar to this. Thanks for doing it for me. Now I don't have to! I see too many designs in which the layer bonds must bear most of the load. As an alternative to chamfers at the bottom, in my designs I use a round curve with its center lowered so that the curve starts out at a 40° angle from the build plate, and then curves up to vertical. It looks good and prints well. Any printer can safely do a 40° overhang if your layer height is no more than half the nozzle diameter.
That's true, there's a lot more nuance to the chamfer thing. A Fillet also works as long as the radius isn't too large. That alone could be a while video of its own
I tend to use 3d models with my cnc machine where curved corners look superior (even if they take longer). Different use case. Regardless, make a step file available and anyone can adjust as needed.
Great video, but as someone who design 3D prints from time to time, I wanted to say that when I make walls thicker than 2 mm it is usually for aesthetic reasons. Sometimes thicker walls look better. But there is no reason to go beyond 2 mm if you can only see one side of the wall.
Of course it always depends on the use case. If you have walls that have a considerable distance to each other, like 5mm or more it also increases torsion strength. It's not a one size fits all approach :-)
another option for the hooks would be to add a taper and design the hook so that you could insert them from behind and allow the taper to lock them in place. there's no real need for them to be flush with the back so that allows for a tight fit without needing high precision.
4x0.4=1.6, arachne perimeter generation may extrude more plastic and make up the extra .4mm, and prusa slicer extrudes slightly wider than the nozzle orifice by default but if you are using legacy/classic slicing and a .4 line width it may not work properly (4 perimeters may not make a solid 2mm wall) depending on slicer settings or the slicer used.
Finally someone notices! I swear I wrote 5 in my script... At least for the Prusa 2mm always made 5 perimeters since extrusion width is set to 0.42 by default. The animation you see in the video is with Arachne so it doesn't matter as much as it used to anyway.
Generally. Don't forget that every perimeter also has a few percent of overlap with the others. With Arachne that doesn't matter as much anymore. Formerly I would've checked if all perimeters are yellow (as opposed to white, which means gap fill) and used that value. 2mm turned out to work well for me.
It's very interesting how some models print with no defects but similar shapes from other makers have issues. I realise there are settings but I have a big appreciation for the designs that print perfect first go.
Expert designers like the Voron team add a lot of little hacks to make printing easy. Like supporting a free floating hole with a 0.2mm "half layer" that helps the Slicer not printing in mid air for example.
As for the hooks, superglue is fine but my favourite way of connecting PLA parts is ethyl acetate: it dissolves PLA and evaporates, leaving a clean and "native" bound (fused) connection. It's also dirt cheap. Basically same what acetone is for ABS/ASA.
Is it save though? I've heard of PLA fusing with things like chloroform, which is obviously not save. I don't know about ethyl acetate though. Other viewers pointed out flaring out the end of the hook is an even better solution.
@@SmallBatchFactory ethyl acetate is a simple organic ester (of acetic acid and ethanol) present in most fruit and contributing to the aromas we consider "fruity". Of course it can be irritating and should not be ingested in large concentrations, it's also flammable -- but it's still considered safe to handle.
Another thing to consider is what TYPE of printing that model was intended for. Many of the issues you outlined are non issue for a resin printer (like rounded edges). Sometimes whoever posted the file online don't include what printer was used in the description, and sometimes not even a picture of the make itself. This leads to having to figure out if it will work for you or not depending on the printer you have. I, myself, have both kinds and I prefer FDM printing, but I have a Mars 3 for the odd thing that the Voron can't do.
This is among the top reason I am so hesitant to buy a model file sight unseen. So many of these models are designed by people who are great at 3d cad, but have zero experience with how that model would work in a production environment. As such you end up with many really cool looking models that aren't going to print ever. I am far more willing to support a 3d artist if it is evident they understand the 3d print world and if t hey have some free models that I can print myself and see in person that it works.
Yes definitely! The Voron project is one of the greatest example of people understanding 3D printing extremely in depth. Not only is the printer great the parts are all printable without any extra supports, everything is already accounted for.
I don't regard myself as being particularly 'great' at 3d modelling, but with a distant past in metal fabrication I find it much easier to have a look at thingiverse to see what other people have done and then I'll just model from scratch. It's generally quicker and a whole lot less agravating than dealing with other peoples mistakes (and if I get it wrong it's my fault, which is OK). One thing I do use which I don't see other people doing though is using a 3d printing pen to 'weld' together the parts of multi-part prints. You can get decent results from it with a bit of practice.
That's true, incorporating some inspiration in your own model usually yields the best and quickest result. The clothes hanger shown had an f3d file but it was pretty hard to work with. I've only welded to mitigate issues in a print. I've been using a soldering iron and filament, like a TIG welder. Haven't had much luck with the pen alone
In one of my CAD courses we had to remake some parts for the original RepRap printers. Some parts had to be printed with horizontal holes in mind. This required a half circle with a right triangle on top in order to print effectively
For a radius at the build plate I start with a 45° chamfer then radius/fillet the upper edge of the chamfer the full height of the chamfer. It can take a while to understand and incorporate “printability” into your modeling, but it’s also kinda fun and I wish more people would take the time before sharing their models.
Digga, that’s what I’ve been struggling with for years, I just take inspiration from designs online and always make my own. That is the only way I get it exactly right to my needs 😂
How I'd do the hooks: similar to your last version but can add a "lip" on the back of the hooks (only up/down to still lie flat on bed) and a recess in the back of the plate, then the plate captures the hooks when installed on a wall.
As others have said its painful to use many models from thingiverse etc. So much hassle could be avoided with just some though into the design, you have some good examples. As an engineer i find the challenge fun to design with printing in mind right from the start, to improve strength while minimising print time and supports.
Yes indeed, it's way more fun to do it properly. I usually think about models for days before I even open Fusion 360. After that it's often just a few minutes to create the model itself.
9 місяців тому+1
I had to come and say thanks for all the videos you've been making. It has been a a treasure trove of information for a beginner 3D printer like me. Precise information and no fuss, I love it! Keep doing this and you will get a following,
Thank you! It seemed to me a lot of channels today only do "Review of $RANDOM_PRINTER". Knowing how to make that printer work properly on the other hand is a rare thing...
Thanks! Only downside it was a tad bit small after adding the visor and compressed the guys nose while wearing. He still was very happy though, I guess cosplayers are used to endure such things.
I was guilty of adding round fillets to the base of models and quickly realised that chamfers are the way to go. Top notch video. Keep up the great content. Cheers 👍😎🇦🇺
I think a two stage hybrid thing works ok where you can have the bottom most part as a chamfer and then the joint between chamfer and the vertical part being executed as a round fillet. You can even make the chamfer a little steeper than 45 if you like for more of the joint to be rounded.
I work on these aspects all the time, but I also push the limits of what is possible with my 3D printed speaker drivers. Single walled TPU parts for the surround and spider for example
Yes! Omg! These things have been annoying me for years. People either seem to try to recreate metal (or injection molded) designs that are too delicate for plastic or not suitable for 3d printing, or the go the total opposite route and make foot thick walls.
@@SmallBatchFactory seems like a lot of people get hung up on making prints stronger rather than strong enough. I still have a box I made as a first vase mode test years ago that I’ve used to drop hot nozzles into during nozzle changes (with a wet sponge at the bottom and sometimes some water) and it both doesn’t leak and has been plenty strong enough ever since.
I'm a bit late but what a masterpiece i learned a lot and wil try to improve i also find it a good decision to keep out the sponsors but if you really need it it's your own choise. please keep going you are one of the best 3d printing youtubers ever
One that you missed that always gets me is sharp corners (especially after long straight extrusions) on the intended build plane. What constitutes "sharp" depends on the intended printing material, but with some large PETG prints even 90° is too much. because as it contracts slightly in two directions even a well cleaned build plate, with excellent z-offset and adhesion may break contact in the corner leading to severe warping. The fix is SUPER easy too: just fillet the corner to a 2-3mm radius and it changes the forces just enough that it stays stuck. This is part of why a brim can help the corner naturally rounds off. If the vertical corner must be as square as possible you can blend the fillet into the corner in the first vertical mm or so. Or add a 1 layer thick "mouse ear" which is easier to clean than an entire brim. This is particularly useful for more technical filaments that are known warp monsters like PC and PA-11 (and in extreme cases you can use the mouse ears + small neodymium magnets to help hold things down assuming the build plate is ferrous)
That's another bummer for sure. You can always add mouse ears in the Slicer easily. Adding a magnet it something I never thought about but is a very intriguing idea. Would need to be high temperature resistant though. I remember Prusa searching for a while until they found capable magnets for their bed.
@@SmallBatchFactory not really. Any Amazon small button neodymium magnet works in this case. Nd magnets have a curie temp above 300C. They'll hold to the steel sheet at 110C just fine. (yes Coercivity goes down above ~100C, but it's still plenty strong to hold down a .2mm thick piece of plastic)
I have to print models for mechanical engineering students - who also have lectures in additive manufacturing - every single one of them I "send home" telling them to fix the exact issues mentioned in this video!
Fair points made apart from the radiator bracket. Don’t use a chamfer the parts as suggested. ❌ Less material in contact with the bed. Doesn’t matter how good your first layer is, minimal contact with the bed more chance for the print to fail. Flatten the one side as a feature ✅ However I do agree that parts should be designed using its manufacturing method in mind with considerations to the inherent weakness of that manufacturing method. Although people that share these models aren’t designing them with a view of selling them, in which designing correctly with research & R&D who would expect
I've printed stuff with very little surface area with no problems. For examle my hinges on the printer enclosure needed to be printed standing up for layer orientation with not much more than a millimeter in width on the build plate. It's usually fine, result may vary of course. I think it's a matter of experience. It doesn't take longer to do it properly, but people lack the experience to know that. It's totally fine, may this video help some people to get better at it in the future.
When I looked at your subscriber count I was shocked, there are channels with hundreds of thousands of subscribers with lower quality videos and less informative scripts. Keep it up!
I have encountered all of these issues and many more, and frequently end up redesigning parts I find online to make them printable (often after wasting a lot more time trying to fix/patch a broken model when it would have been faster to just redesign it in the first place). The one that really gets me annoyed is when people upload models that are simply entirely broken from a real object perspective. Not watertight, non-manifold surfaces, impossibly thin walls or intersecting corners... it's as if they designed the model to go into a 3D game engine and not actually be used to print a real physical object. This wouldn't be so bad if they posted the design files, but your options are quite limited if they only upload the STLs. There are a variety of automated tools that can attempt to fix some of these issues, but if they're bad enough I just use the design as inspiration and make my own with proper real dimensions.
Inspiration is usually my approach as well. I even do that with furniture... Usually you can see it's not bad intention, it's a lack of skill. The clothes hanger bracket for example came with a source file and was very botched together. Changing this was a challenge.
3:00 dont use chamfers over fillets. instead, combine them! do a chamfer on the corner and a fillet on the upper corner of the chamfer. this gives a nice curve while still keeping good overhangs
I love that mold makers are trying out 3D printing! But yeah they probably need to keep working on learning how 3D printing is different. I've ran across my share of designs that look like they would work out better in an injection mold than a 3D print.
It's mostly a matter of what people have learned to adapt to. 3D printing is so young compared to injection molding designers don't have much experience. I've even seen the Prusa video how Volkswagen uses 3D printing and they wasted so much support material...
The curved overhand thing is a huge issue I have, I mostly model and print figures and the backs of them always turn out rough, wish there was a way to fix it that didn't involve changing the design It is nice to know that this is a common issue and not just my 3D printer
Only way to mitigate it a bit is finer resolutions, maybe variable layer height. Resin printers don't have as many issues with it since layers cure at once.
@TheTurtleyOne oh yes. I bought one and I hate it. So much waste produced and a lot of fumes. I have a profesional mask but still... I don't even keep it in the house.
I think because its fairly easy to learn how to make an STL leads to so many not so great designs. I find that I don't really understand the best way to orient prints so I really appreciated you explaining how to make prints better. I shall have a look at your course, I want to lean Fusion 360 but I find it rather intimidating.
That's probably it. Almost no other hobby enables you to upload all your designs. I imagine there are also a lot of bodged woodworking projects out there, but you can't upload them. If you have any questions about the course you can drop me an email. I'm always grateful for feedback too.
@@SmallBatchFactory we also have things like makerworld (Bambu Labs) they give us points when we upload projects and then they can be used to get discounts when you buy their products, there is an explosion of people uploading to get those points, I understand that Bambu Labs want to populate their platform with lots of content, and there is heaps of great stuff on there, but there is also a lot of sub-par and sometimes just bad designs on there.
There is an advantage to thick hollow walls. They have increased stiffness, since for a given bend radius, the outer walls that are further apart would need to compress or stretch more, which they resist. Many of these prints can even be done single wall just fine with a lot of void and you don't need a lot of infill just to tie the walls together and help it hold shape.
That's correct. I specifically referred to parts like the power strip holder "covering space". Stiffness is for the "solid blocks" I mentioned. The prints you're talking about get their strength mostly from the top and bottom layers, like a square pipe or an I beam
I think I initially designed a part with arc overhangs in mind at one point. def helps for 90-degree overhangs but at the time, and last I checked was still the case, I found an edge case in the script that requires quite a bit of design consideration
While I appreciate the engineering aspect of things, as a Dane I do desire better designs too (for those not in the know, I am referencing Scandinavian Minimalism. Minimalism itself started in Germany (the BAUHAUS movement) and it is focused entirely on the functional aspect of a certain product, removing all unnecessary features. Scandinavian Minimalism started in Scandinavia (as the name implies) and is focused around retaining design features by integrating them in the functionality of the product). Jokes aside, layer line orientation, or rather disregard for the manufacturing method used, is also my arc nemesis.
This is why I tend to just make my own objects. If its super high in detail, thats one thing. But a simple clip or holder of some sort? Eh, design something in 20 minutes, spend another 20 adjusting tolerances, adding chamfers/fillets, and call it good
It's a major thing I hate: designers really need to make sure multi-part models have the same start layer. I've had a ton of camera parts that have failed because 3/4 parts start at 0mm but the last piece starts at 0.2mm.
That's annoying. You could try to split the parts and put everything on the plate. A good advice is to always check the resulting print in the Slicer at least briefly. Floating layers are usually easy to spot.
Liked just for the "solid infill" comment about it... not to say it wasn’t a good video but outside my normal modi operandi. As a person new to 3D printing and modeling myself I find these types of videos very helpful. This is the first video of yours I have found but look forward to more in the future. It is why I enjoy watching @slant3d videos on the same topic so much. Thanks for putting this out there.
I don’t know why some people hate in designs that aren’t one part prints. I made a cable holder for my Valve index so the cable can be mounted straight down and added a wave spring to reduce stress on the cable. Had to print it in two parts or else one of the parts orientation would result in weakened sections that would break near instantly
@@SmallBatchFactory Personally i find it way easier. You can split up you big part and assemble it later with screws and glue or just slot parts together. If you have single print parts you need to adjust so much about your print to get equal results. Working out the right orientation or needing more support because some parts hang in the air. What about print fails? Now you have a failed print that takes 12 hours fail at hour 8. Instead you could have split it up and could have had a total print time of only 10 hours and split it up in into a 5 hours, two 2 hours and one 1 hour prints. One fails? Not that big of a problem. Only real drawback i see whit this is, that you need slightly more work after the print. So less useful for scaled up production that needs an assembled product. Or go the Ikea rout and deliver an assembly kit.
My most popular design on printables/thingiverse has a fillet facing the print bed ... sorry about that. Now that I know about the "poor man's fillet" (ie, a chamfer half way and a fillet the rest of the way), I need to upload a v2 of the models, but I just haven't gotten around to it yet.
Thin walls should be related to the extrusion width and layer height, not just multiples of nozzle width. In Prusa Slicer, you can see the recommended thin wall thickness for various number of vertical shells at different layer heights. E.g., 4 lines at 0.2 mm should be 1.67 mm, not 2 mm
Gosh, I really appreciate the advice on cylinder orientation and edits for stronger prints! One of my personal bugbears is electronics boxes modeled with broad, flat sides and too-thick walls. These are horrifically weak, especially against torsional forces.
I understand the thick walls, as they dont really bear any load they can be much thinner than usual. But I disagree about the other thing you mention. What do you base that conclusion on? Closed box sections are actually incredibly torsionally stiff. Even without a lid a square section is fairly torsionally rigid, only when you remove one of the walls the stiffness drops dramatically. See also chapter 3 of Gere's Mechanics of Materials. Of course ribs can be added to increase the torionsonal stiffness even further, but for an enclosure is that really necessary? When is that ever subjected to torsional loading?
Good advice, but for power strip, especially schuko, the forces are not insignificant. I ended breaking a few power strip casings just because the plugs were stubborn. Now I have replaced most of them with aliminium ones.
My own wall mounts are shown briefly which I tortured a bit before the video just to be sure (they're Schuko) . Haven't had any issue with them so far, being printed with PETG.
Makerworld at least requires you to upload an actual picture of a print. People try to get around it and the community usually reports it when they see it. I avoid models that have no real prints or prints that look terrible because it likely means the file is unprintable.
While your example about not wasting filament for the plug holder is right, its quite a simplification to say that a thin solid wall is always stronger, when there are in fact many scenarios where a wide part even with very low infill will in fact have much better mechanical properties.
It very much depends in the use case of course. That's why I said "geometry covering empty space". If you need something resistant to torsion forces, spanning over a bit of distance you need thicker walls for sure
You are 100% correct. Often the inside does not carry much load, most of the shear loading is carried on the outer skin. This is exactly why people prefer to use hollow tubes instead of solid bars. For the same amount of material a hollow tube (of much larger diameter) is much stiffer and stronger in any DOF compared to a rigid rod of comparable mass.
In such cases it's worth to try solid infill every X layers instead of conventional infill. Eventually use concentric infill at very high percent. Sometimes it works as another type of fast wall printing
@@SmallBatchFactory It's just a widespread issue I see in 3dprinting videos talking about "strength" in very simplified terms. An actual proper video about different mechanical stresses in 3d printed parts would be wonderful, because the vast majority of "strength" videos are mediocre at best, and outright misinformation at worst.
Literally did this yesterday. the part was a battery holder. It was too tight and required 2 different lengths of screws unless being screwed into a thick board or a blind space. Tried to modify it but fusion 360 wasn't having any of it.
There’s a TON of design consideration that goes into every manufacturing method, except 3D printing. There, the best most people do is try to avoid supports (even though they’re not THAT big of a deal these days) and don’t really consider a lot of the other advantages/disadvantages of 3d printing.
I guess it's a lack of experience. Thingiverse is mostly populated with hobbyists. I downloaded a model from a coffee mashine company once for example and it was awesomely designed.
@@SmallBatchFactory seems like there’s a lot of people who picked up “tips and tricks” and just kept doing things the same way and not a lot of new people learning to design stuff themselves now that so much is available so easily. But I think videos exactly like this are what’s needed to both get people interested in at least modifying designs for their own use, if not learning to design things specifically for 3D printing themselves. Guess it’ll just take a while for people’s imagination to not be based in traditional manufacturing techniques/designs.
@heyspookyboogie644 yes definitely. Most people can only work in small iterations. That's why radically different products usually flop until we'll arrive at the same product years later. I think there are too many printer reviews and too few "fundamental education" channels in the niche. People seem to agree.
Oh yes they do and I'm certainly guilty myself. If you keep it to a millimeter or two they still work well. Many overdo it and that's when they fail miserably.
My favourite example of a great design ruined by a design flaw is a case for an esp01 relay box i saw once. Everything about it was perfect, except the clip thats meant to hold the wires down had both screw holes placed too close to the edge causing them to actually clip through the sides of the clip causing a whole mess of random polygons in the negative space. No idea how they managed to upload it without noticing the error but the designer just dropped off the face of the planet Theres also no way to fix it without completely remodeling that clip from scratch, which is doable but tedious
I spent 43 years in computer hardware support. After reading the comments it sounds just like the issues I ran into. The people/ engineers designing things never actually put it together, use it or even tested what they came up with, hence the problem. I have had many classes learning brand new products and instructors literally throughout the manual. Because there was such a poor match between the manual and real world equipment.
@WhiteG60 this or sometimes even gross incompetence. At least with a lot of early Chinese printers there was a lot of "everybody does it" implementation without grasping the reasoning for why to do it that way or not.
Turning off part cooling at specific layers and slowing down to about >50mm/s (@0.2 layer thickness) will create near mold injection like strength. Depending on your printers hotend of course, you may need to be under 30mm/s. Especially if it's a open air bed slinger. The bed moving back and forth cools the part as well.
Not always possible my man. You also have differences in printers. No two are the same. The printers own accuracy comes into play. I try my best to avoid certain shapes, provide tolerance, etc., but it doesn't always work 100%. You don't want to fillet a bottom edge, use a chamfer. You can fillet top edges and vertical wall edges. You can reinforce prints using ribs and rafters. Just incorporate them into the look if they are visible. You want to fillet where edges meet between connecting parts. It will increase strength where the part connects. Arcs distribute stress well. Print parts solid infill that should be solid.
Even a cheap bowden CR-10 of the first generation gives you really accurate prints. The rest is tuning, given the machine is has no defects. The rest I totally agree with.
I don’t use screws or threaded rods for alignment issues, I make the weak area hollow but then I fill it with epoxy or print a properly oriented piece i can glue in I did this with a Vesa mount adapter, I wanted something that would last forever so I just put in big structural voids I could fill up with resin
@@SmallBatchFactory I just haven’t had time to play around with threads and get used to it, so for now I’m paranoid about it eventually delaminating and I’d still end up using glue or resin to fill the crack (especially with the hook example where the hook could twist to delaminate)
@Pentross that's understandable. Layers compressed with metal are virtually indestructible. Me studio lights hang in such printed parts so I pretty much trust that concept. I personally have more aversion to Resin than to metal hardware :-)
This video couldn’t have come at a better time! I just started designing my first models. I’ve already even made the first two sins from this video, now I feel like I’m truly part of the 3d printing community! 🎉 I started your fusion360 course a few days ago, and I’m really happy I did (and also terrified of the process of actually learning). Although my current needs for models haven’t really required anything more than Tinkercad, I’m going to slowly learn the material so I can feel comfortable enough to tackle any problem in Fusion360, and your course is an excellent resource to have on hand for the journey.
Welcome to the family, glad I could help! I think it's the ideal compliment to the course, helping improve Thingiverse & Co. A bit. Makes me really proud to hear that! Let me know in case you get stuck anywhere. Getting that beginner view really helps improving explanations where necessary.
@@SmallBatchFactory I’m sort of “stuck” in a sense as I’m trying to motivate myself to finish the last videos of the “creating your first model’ section. It can feel like a bit of an information dump to a newbie like myself. I wanted to remix a fan bracket and figured I’d try your course as an intro to CAD/Fusion360. I kept trying to use what I had learned in the first chapters of your course to get started, but I couldn’t really figure out how to sketch out the dimensions of what I wanted drawn on paper yet. I’m sure it’s more comprehensively laid out in further lessons, and I just had super unrealistic expectations about quickly modeling the design (I was so eager to design the part and I am very inept at viewing the world in a “problem solving” manner - so I was really operating in unfamiliar territory.) That’s when I tried out TinkerCAD and found “hey this works just like a slicer, I can manage that” and a day or so later I had the part printed. Poorly, and without the ability to make quick adjustments to dimensions, but it was made and I really needed that dopamine hit. Iterating on that design I can do things that took me 2 hours in 2 seconds now. When I get some more time in the next week, I plan on sitting down and taking a proper crack at the course. I feel like the lessons I’ve learned playing in TinkerCAD will only benefit me, since I literally had no idea what I was doing before, and now I’ve come a long way (with still soooo long go!) So if I had suggestions for the course, I’d say add in some ways to give my ADHD ass some dopamine hits early on and frequently. Really shove those dopamine hits down my throat, learning is painful without them. I’m possibly a unique case though in that sense 😅
Oh yes, doing everything properly surely is a lot of information. My first attempts also were mostly "I want to get things done". I did my best to get to the first practical model as soon as possible and that also means there are still a lot of things to learn to create more complex shapes. I guess that's what you experienced, you have something on mind that's actually easy to build out of 3D shapes while it takes some getting used to doing it with sketches. Maybe we can use your project and add it as additional practical example to the bonus section.
You're spot on! You said it better than I could've. I wanted to create a more complex shape, without really knowing that's what I wanted to make. I have a couple more things on my "wishlist" of learning. A replacement latch for my freezer door, some cool accessories for my cargo bike (cupholder for my daughter perhaps?). Silly stuff like that I guess. I think I'll be able to do it, maybe even quite easily - especially once I have a bit more time to watch the rest of the course.
I'm sure you'll get used to "thinking in sketched" with a bit of practice! Something mounted to a bike frame or some other tube would actually be a good example to add to the course. I already did a few bike accessories screwed onto the frame and this might be not too obvious to some. I'll definitely add that to the ToDo list!
Every single 3D file I download for printing needs repairs in one way or another. Sometimes there's non-manifold or detached geometry, sometimes the tolerances weren't spec'd properly, sometimes flat faces aren't actually flat... Open source is the wild west, you get what you get :/
We can't really complain if we don't pay for it. But to be fair I have the same experience with furniture. I usually already have an upgrade plan when leaving the shop or even build it from scratch.
Small things i disagree with. If you are good with 3d printing you can get ~80% of the strengts in both ways on layers. Most people have a issue because they either use way to much cooling or way to low temperature. Sometimes trying to go to fast. For something meant to be aesthetic i wouldn't care about layer orientation. 3d printing curves at either the base or top of a 3dprint should be done with incremental steps/variable layer height. If i got a 3d print that got a big radius on the bottom that for some reason have to have it. I would start with a 0.02mm layer after the initial layer height. And increment it higher. Also tend to let the bed be 5°c colder than usual to avoid the layer flexing down. Problem with 45° is they may be to sharp in some instances.
I've never seen any test showing 80% strength across layers with all samples having the same section size. Variable layer height works of course. In many cases it's a bit of a crutch IMO
I really don't think that this powerstrip holder was bad. It should be plenty strong!! My mother used to rip the sockets straight out of the wall in our place. trust me some people do not treat them gently.
It's just an example anyway. It's more about the core message "do walls appropriately". I can tell you power plugs in Germany hold really Strom in their sockets and even with the I can't rip mine from the wall. I tested is specifically because of comments.
8:14 I personally like to plan for zero or near zero tolerance in Fusion 360 and use the “Exclusive” slicing method in the Cura experimental tab. However I do not know if there exists another similar option in other slicers.
By far the biggest issue is people just giving an STL. It's so prevalent that people thank me when I also upload the CAD file. The websites should REALLY incentivize people to also upload whatever CAD file they used to create the STL I also find it super problematic when people design for a certain nozzle size when it's not needed. I routinely print with 0.6mm nozzle (and frankly almost everyone should!), but some models needlessly claim that they are optimized for 0.4mm. If you do that, at least give the CAD file!
I'd say to fix the key holder you should shift all hooks down and make them square, at that point they will touch the build plate and you can print it rotated to get that additional strength, it does make the mounting plate a bit weak but for a small load if you don't torque the screws too much it will be fine, that removes a lot of hassle for a less experienced printer, adding threads that size is not that useful in my opinion and uploading something with such threads will just be problematic to people that have bad settings I'd say you better thread it yourself with the screw and leave a hole, also you should have added a recess for the screw at the back cause that is going to be against the wall
You're totally right, it's just a quick inspiration of how you could do things. The rest is up to the designer. The original was designed for magnets. Adding a recess for screws would also work. Really depends on the individual use case.
Learn Fusion 360 the easy way: smallbatchfactory.com/fusion-360-fundamentals
You know it's going to be a good time, when a video starts with a german accent talking about precision
This is zhe way!
That's a good one, deutsche die Englisch sprechen lieben wir 😂
This is the most German video spoken in English ever! And I love it!
Sign zee paperz olt maan!
😂
My biggest issue is fitment. I swear no one leaves tolerance in their models that have parts to fit together. I'm over here sanding sanding sanding, hammering hammering hammering, pushing pushing pushing.
I can totally relate! I recently printed a storage for my camera batteries. It took three different remixes and then time scaling until it fitted. The measuring tool in the Slicer is almost mandatory at this point.
I always leave a 0.2mm tolerance for easy but snug fit
That helps. Different materials and printers make the behavior impossible to predict unless you tested it before so more planed tolerances make for universally useable models.
I totally agree. I have spent days printing two different sets of drawers for my Ender 3 Pro, both from different contributors, and both were too tall, hitting a screw in the bottom of the build plate. So frustrating!
The measuring tool in the Slicer is so helpful. Prevented me from printing stuff that wouldn't fit anyway a few times already.
Inserting a screw from behind to strengthen the hooks is pretty genius. Great idea.
Thank you! It's always delighting so see people excited about something I took for granted.
I have a friend whom I helped out with her thesis project. She had a team partner who was in charge of designing all the parts for the robot they did, and I swear it was a nightmare every time she asked me to print something because the guy had literally zero notion about additive manufacturing, let alone 3D printing (he never got his Ender 3 running, for some reason). And I don't know if it's a coincidence or what, but you mentioned basically all the design mistakes he made and I had to correct or account for in the slicer, so the parts were actually printable. This video is gold and needs way more exposure.
Thanks! And those are only the most ubiquitous and easiest to fix mistakes...
Feel free to share the videos anywhere :-)
Yeah I used to make lots of mistakes in designing parts but now a smooth it out
In a case where there is that guy who doesn't know s..t about 3D printing, still sends you parts to be printed and never learns, I have a special treatment. I just remove the Z support gap and laugh maniacally while the guy is trying to remove those damn supports!
I am probably a bad person to do that to someone, but if that someone keeps repeating exactly the same mistakes, what can I do?
@yuriysukhorukov391 this is eeeeeeevil. It like it!
This is why every model i upload i also provide a STP to make it easier to reference or modify/remix if need be.
That's the spirit! I once copied the design of a bike wall holder since there was no source file and uploaded everything. People thanked me so many times for the source since you could fit it to any wheel size.
This is the way
I only upload STP files. And because of this fact, I asked one designer to also offer STP files, because I wanted to update his model to a new version of the object inside of it.
Sharing step files is commie shit
I wonder why STEP isn't the default anyway. Why did STL become the standard?
How to print a sharp inside corner. The trick is a "mouse bite". Intersect a small cylinder with the corner so the STL shows a cutout up the corner like a small ice cream scoop was run up it.
When the nozzle goes around that small cut, the squeeze out flows and closes it in, leaving a sharp corner.
The first time I used this trick I was modeling a replacement part that had a rectangle hole which had to fit onto a metal tab and a square hole into which snapped an injection molded plastic part.
I did a TLAR (That Looks About Right) on the size of the corner cuts, sliced off those parts of the model to do a test fit and they were both perfect the first time.
The concept comes from fonts designed for printing on cheap paper where ink bleeding is a problem. The Bell Centennial font was designed specifically for printing phone books. To improve its legibility at small sizes, all inside corners of the letters have notches so when the ink bleeds into the paper it flows into the notches and makes a sharp corner.
That's a really handy trick! Just like angle tools have a circle where the edge goes in to make it fit in any case
I must agree, curved geometry near the bed is making me cry inside.I’m all for joining a dedicated chamfer lobbying organization.
Way to go! When I was a total beginner I always thought the bad print is my fault. It is printable with very fine layers but that's not really a good solution.
@@SmallBatchFactorycura has a great feature called "make overhangs printable." This actually alters the geometry of the model so nothing is more than a 45 degree overhang (or possibly whatever you have your overhang angle set to in support settings, not sure). If the only problem you have is badly designed rounded chamfers, that should fix it.
Or you could do what I do: make a curved bottom edge that starts at a 40°-45° slope and curves upward to vertical. This prints well and looks better than a simple chamfer, but it's a bit more work to figure out how much to lower the curve center.
I usually make parts that can be joined with screws so that if something breaks, I have to print only that specific part.
I like that approach as well. It gives me peace of mind. I also often run a screws perpendicular to the layer lines to compress the print. Makes them virtually unbreakable.
Everyone needs to see this. I already incorporate a lot of these in my designs. One more thing, on the bottom of my models touching the build plate I will usually add a 30 degree chamfer typically 0.4mm high (or two layers for most slicer profiles) to prevent elephants foot. This way people who download my model won't have to compensate for it in their slicers, as a lot of inexperienced people might not know how to compensate for that. 30 degrees keeps the stepover small enough so there's no chance of sagging lines even if people don't squish their first layer.
Good! With elephants foot there's always the issue that it's highly dependant on your printer. A chamfer is a good start. My Prusa default profile for example is not enough for how close I print my first layer.
I print a lot of miniatures and my biggest issue is details where there doesn't need to be any ... Like a tongue inside a closed mouth !!! Looking at you HeroForge 😤
I guess they design the model once for different poses and don't really care how you're supposed to print it. Only a few models like Adalinda the dragon are made with printing in mind and work without any supports.
I think people are scared of mesh mixer but if they give it a chance these issues would a 2 minute fix
@knifeyonline definitely and I'm guilty myself. Seems like a good video idea
@@SmallBatchFactory yeah! It's certainly not an over exposed topic, I don't think anybody is tired of watching people use mesh mixer. So many uses. I saw somebody scan their face and clean up the model in meshmixer to print a mask.
@knifeyonline that's right. If it were I probably would've used mesh mixer already.
I saw the thumbnail in my "recommended" and thought "I wonder what this is all about". I was surprised to see that key hanger. It's my design! I was both excited and humbled to be featured.
For some context, that was one of my first designs in OnShape that I made to attempt a 3D print for my new printer. I know it's not great but I liked it and so I uploaded it.
As it is, I printed it without supports but, admittedly, it didn't print real well. I wasn't bothered so I kept it. As for strength, it seems fine as it's only holding car keys. I've been using that particular print hanging on my refrigerator for about a year with no problems.
I did update the design with the holes at the back to insert magnets. There were just holes for hooks before. That was about it.
All that said, I'm happy to be used as a "what not to do" example. All of us are students and no one of us is as smart as all of us! Input, feedback, knowledge sharing, experience, etc. are all valuable for anyone learning something new or expanding their knowledge of a given subject.
Thanks for the lesson!
nerd
My first designs also totally sucked. Learning takes time!
I'm glad you liked it! Such a small world after all. In many cases "good enough" totally gets the job done anyway.
I would lie if I said my early designs weren't a mess either, figured the part about fillets on the build plate the hard way.
Somebody even one uped my suggestions with a better one I featured in a community post. We're all just students :-)
It's also depends a lot on material. PLA is a trash material when it comes to functional prints. I only print functional stuff in PETG.
That's a way too generic take on PLA
The idea of a screw from behind to reduce the impact of the bending moment is a good one I wouldn't have thought about, thank you.
Brilliant video. As an engineer, I gave up on using functional prints from any online source due to the reasons you put forward in your video. I mainly print sculpts or miniatures now and if I ever need a functional part, I quickly design it myself. :D
Definitely! I view Thingiverse as "serving suggestion" and a source on inspiration. It usually even takes less time to do it yourself than finding something fitting...
With the hook design, fixing the curvature to make it an angle cut is a good start. But if you want to increase the strength without printing in parts, you want to add a large fillet at the bottom edge of the hook and a somewhat large smaller fillet on the top edge. From my experience, shear forces are never the reason for failure. When you load the hook downward, the failure is caused by the deformation of the hook, this causes the hook to bend forwards, so now the angle of the load vector changes and there's a force component exerting force perpendicular to the print layers. This PULLS the layers out. So a large fillet below bottom edge would work to reduce deformation as a fillet in the loading position acts like an arch and arches resist compression. This technique has worked really well for me so far.
You're absolutely right. That's under intended use.
What I expect is somebody getting caught with a shoulder or something while passing the holder and snap of a whole hook.
Good advice! Steep angles on the first layers also grinds my gears.
A note in terms of self-presentation: I find myself distracted by your head constantly bobbing to the left and then to the right.
Thanks, I need to work on that. Compared to the last video I managed to wave my hands less so there's that :-)
You didn't move your head to much. You were being you and the video was perfect. Whoever was paying attention to your head moving was not here for the actual content of the video.
@patrickcarpenter6258 thank you for your kind words! I guess there is a middle ground I still have to find. After all it's only my 8th video ever. Besides, my wife also said it's too much and you know what that means :-)
@@patrickcarpenter6258 Just because you do not notice things, does not mean they don't happen.
@@patrickcarpenter6258 "Whoever was paying attention to your head moving was not here for the actual content of the video." - what a false assumption. I'm here for the content but this head banging is so distracting like someone poking me with a stick every few seconds. Hard to focus on the content. I understand it's my problem that something distracts me but looks like I'm not the only one. Still, good content anyway!
Good selection of painful design choices. Guilty of a few of these myself sometimes when in a hurry and I just need it now, but always regret it later.
Another one that would be next on my list is using 3d printed things when a widely available non 3d printed thing is readily available. I was doing this myself with clips to attach a crutch to my wheelchair. I had clips for the crutch and for the chair frame. It was always annoying when the whole clip assembly would come off instead of just the crutch. I had to facepalm and realize that 3d printed clips to attach it to the wheelchair was dumb. Just adding a few slots instead for zipties to go through made it far more functional and strong.
Thanks! We're all guilty, I was too.
Combining standard hardware is a real game changer. Be it screws, zip ties or something else. The Voron team for example combines cheap small idlers with printed parts to make bigger ones that would cost over 10 bucks each.
Yeah it's very easy to fall into the trap of 'if all you have is a hammer...', especially when said hammer is a hobby unto itself.
A lot of people have 3D printing as a hobby. I see 3D printing as a tool for my hobbies. Just like laser cutting, welding, electronics and woodworking
You could make the key rack modular. Each section is one hook and back plate. Each piece prints on its side with the hook on the build plate and has provisions to snap the next piece in next to it. Then you can make it as long as your want, the hooks are integral and as strong as possible, you can add to it or replace sections any time, you could customize each section like by adding labels, and you just have to print end caps for the right and left sides to finish it off with no holes.
That would also work! Your imagination is the limit in 3D printing.
@@SmallBatchFactory Yeah I just like sharing additional ideas in the comments. Always looking for ways to improve a design for printing, customize it for individual desires, and possibly add functionality if useful. Fun thought exercises. :) Keeping it simple is often the best approach for real usage though.
I think a lot of people tend to try to model things as a single piece, even when a design would really benefit from multiple pieces.
@@logicalfundy Definitely true. I like to try to think of how it would work both ways and go with what makes sense. Learning how to design for 3D printing is still an evolving art for most.
I came across one, where I noticed a major design flaw after the print had failed because of a combination of that flaw and poor bed adhesion (and the poor adhesion was not something I normally had problems with.)
The item was a flat molle-patch with a cutout logo. But that logo contained the letters A and R which don't work for a cutout. The tiny, unconnected pieces within the letters should have been omitted or designed differently. It was obvious to me, that the user had never even printed that model.
Those are common problems. Prusa has a printed plate with their logo so it does work, but the letters need to be large enough
You can add a box volume in the Slicer to fill those voids though. An easy way to get rid of unwanted markings in those models.
I swear its like people dont consider the manufacturing process theyre using when they design. Bothers me so much.
Nice video. Suggestion for the last one: instead of making the hooks press-fit, flare the base of the hooks, add a recess on the back of the base plate for the hooks to sit in and pass the hooks through the holes from the rear before installation. Then, when you mount the base plate to the wall, the hooks are held in place mechanically between the wall and the plate. Since you want a flat side on the hooks anyway, you can make the recess match to key orientation.
That's actually even better than my example. There's always room for improvement :-)
Saying fuck unrelated sponsors is the second most appealing reason to subscribe. Number one is the top notch facts and content delivered in short and precise manner! Good job.
Thank you! Got a request to advertise a desk today. Despite all their praise for my content they didn't bother to watch it I guess...
Fillet on the base thing is something everyone should learn when they print something with it once or twice and realise "oh that does not print well" or you spot it when you slide the layers from the bottom to the top and notice very steep overhangs.
Yes, they should. When we printed face shields people handed it prints that already broke when looking at them so I suppose the fillet is the least of some people's problems.
I am at a university and we have a bunch of 3D printers for public use, my biggest pet peeve is by far wasted material. Not only do I see people needlessly printing with high infills, the parts being designed seem to have no notion towards how they will be produced! I constantly see huge boxes of plastic being printed that would function identically with 95% of the material removed in the CAD model. Whenever I design parts to be printed for my FSAE team, my number one priority is to use as little material as possible, and most of the time I can get away with 0 support material by introducing snap fit or glue together parts like you mentioned at the end of the video, plus it is a great way to start introducing the viability of AM to the senior team members if they get to see and handle practically designed parts.
So much wasted material... I'm a huge advocate of efficiency and optimization. I even print a lot of models with no infill at of I don't need to.
There are so many great ways to optimize the design like screws that are flat in top and bottom for example. Maybe I should do a part 2 and show things like this.
I'm not a video producer, but these peeves have been making me want to make a video similar to this. Thanks for doing it for me. Now I don't have to! I see too many designs in which the layer bonds must bear most of the load.
As an alternative to chamfers at the bottom, in my designs I use a round curve with its center lowered so that the curve starts out at a 40° angle from the build plate, and then curves up to vertical. It looks good and prints well. Any printer can safely do a 40° overhang if your layer height is no more than half the nozzle diameter.
That's true, there's a lot more nuance to the chamfer thing. A Fillet also works as long as the radius isn't too large. That alone could be a while video of its own
I’m no 3d modeller but even I can understand your point here. I’ve lost count of the edits I have to make. Thank you for sharing my pain!
Looking at the comments many people share the pain. I decided to do something about.
I tend to use 3d models with my cnc machine where curved corners look superior (even if they take longer). Different use case. Regardless, make a step file available and anyone can adjust as needed.
Can't argue with that! Step files would be the very best solution.
I appreciate that you don't run 3rd party ads. Ads are annoying enough as it is.
I appreciate the faithfully re-created "What Really Grinds My Gears" graphic
I dont print but I find your talking about optimizing designs for printing very interesting.
Great video, but as someone who design 3D prints from time to time, I wanted to say that when I make walls thicker than 2 mm it is usually for aesthetic reasons. Sometimes thicker walls look better. But there is no reason to go beyond 2 mm if you can only see one side of the wall.
Of course it always depends on the use case. If you have walls that have a considerable distance to each other, like 5mm or more it also increases torsion strength. It's not a one size fits all approach :-)
another option for the hooks would be to add a taper and design the hook so that you could insert them from behind and allow the taper to lock them in place. there's no real need for them to be flush with the back so that allows for a tight fit without needing high precision.
4x0.4=1.6, arachne perimeter generation may extrude more plastic and make up the extra .4mm, and prusa slicer extrudes slightly wider than the nozzle orifice by default but if you are using legacy/classic slicing and a .4 line width it may not work properly (4 perimeters may not make a solid 2mm wall) depending on slicer settings or the slicer used.
Finally someone notices! I swear I wrote 5 in my script... At least for the Prusa 2mm always made 5 perimeters since extrusion width is set to 0.42 by default.
The animation you see in the video is with Arachne so it doesn't matter as much as it used to anyway.
Says 0.45mm width for 0.4mm nozzle in prusaSlicer 2.7.1 as default.
but the idea is extrusion width * whole number = thickness of walls you design?
Generally. Don't forget that every perimeter also has a few percent of overlap with the others. With Arachne that doesn't matter as much anymore. Formerly I would've checked if all perimeters are yellow (as opposed to white, which means gap fill) and used that value. 2mm turned out to work well for me.
It's very interesting how some models print with no defects but similar shapes from other makers have issues. I realise there are settings but I have a big appreciation for the designs that print perfect first go.
Expert designers like the Voron team add a lot of little hacks to make printing easy. Like supporting a free floating hole with a 0.2mm "half layer" that helps the Slicer not printing in mid air for example.
As for the hooks, superglue is fine but my favourite way of connecting PLA parts is ethyl acetate: it dissolves PLA and evaporates, leaving a clean and "native" bound (fused) connection. It's also dirt cheap. Basically same what acetone is for ABS/ASA.
Is it save though? I've heard of PLA fusing with things like chloroform, which is obviously not save. I don't know about ethyl acetate though. Other viewers pointed out flaring out the end of the hook is an even better solution.
@@SmallBatchFactory ethyl acetate is a simple organic ester (of acetic acid and ethanol) present in most fruit and contributing to the aromas we consider "fruity". Of course it can be irritating and should not be ingested in large concentrations, it's also flammable -- but it's still considered safe to handle.
@TomaszStachewicz sounds interesting! I usually use superglue since I have it at home anyway.
Another thing to consider is what TYPE of printing that model was intended for. Many of the issues you outlined are non issue for a resin printer (like rounded edges). Sometimes whoever posted the file online don't include what printer was used in the description, and sometimes not even a picture of the make itself. This leads to having to figure out if it will work for you or not depending on the printer you have. I, myself, have both kinds and I prefer FDM printing, but I have a Mars 3 for the odd thing that the Voron can't do.
You're totally right about the Resin printer. In the case of my examples they were all printed with FDM.
This is among the top reason I am so hesitant to buy a model file sight unseen. So many of these models are designed by people who are great at 3d cad, but have zero experience with how that model would work in a production environment. As such you end up with many really cool looking models that aren't going to print ever. I am far more willing to support a 3d artist if it is evident they understand the 3d print world and if t hey have some free models that I can print myself and see in person that it works.
Yes definitely! The Voron project is one of the greatest example of people understanding 3D printing extremely in depth. Not only is the printer great the parts are all printable without any extra supports, everything is already accounted for.
I don't regard myself as being particularly 'great' at 3d modelling, but with a distant past in metal fabrication I find it much easier to have a look at thingiverse to see what other people have done and then I'll just model from scratch. It's generally quicker and a whole lot less agravating than dealing with other peoples mistakes (and if I get it wrong it's my fault, which is OK). One thing I do use which I don't see other people doing though is using a 3d printing pen to 'weld' together the parts of multi-part prints. You can get decent results from it with a bit of practice.
That's true, incorporating some inspiration in your own model usually yields the best and quickest result. The clothes hanger shown had an f3d file but it was pretty hard to work with.
I've only welded to mitigate issues in a print. I've been using a soldering iron and filament, like a TIG welder. Haven't had much luck with the pen alone
In one of my CAD courses we had to remake some parts for the original RepRap printers. Some parts had to be printed with horizontal holes in mind. This required a half circle with a right triangle on top in order to print effectively
It's one of the lessons in my course, but IMO it's usually not that huge of a deal.
For a radius at the build plate I start with a 45° chamfer then radius/fillet the upper edge of the chamfer the full height of the chamfer.
It can take a while to understand and incorporate “printability” into your modeling, but it’s also kinda fun and I wish more people would take the time before sharing their models.
That works as well. To. Be fair a very minimal Fillet works as well, if you keep it at 1mm or so. It's still an art, not an exact science
Digga, that’s what I’ve been struggling with for years, I just take inspiration from designs online and always make my own. That is the only way I get it exactly right to my needs 😂
Plus it usually only takes a few minutes to design. Less than figuring out if an existing one even fits
How I'd do the hooks: similar to your last version but can add a "lip" on the back of the hooks (only up/down to still lie flat on bed) and a recess in the back of the plate, then the plate captures the hooks when installed on a wall.
That's the one! Somebody sent me in an example of this, I posted it as community post.
As others have said its painful to use many models from thingiverse etc. So much hassle could be avoided with just some though into the design, you have some good examples. As an engineer i find the challenge fun to design with printing in mind right from the start, to improve strength while minimising print time and supports.
Yes indeed, it's way more fun to do it properly. I usually think about models for days before I even open Fusion 360. After that it's often just a few minutes to create the model itself.
I had to come and say thanks for all the videos you've been making. It has been a a treasure trove of information for a beginner 3D printer like me. Precise information and no fuss, I love it! Keep doing this and you will get a following,
Thank you! It seemed to me a lot of channels today only do "Review of $RANDOM_PRINTER". Knowing how to make that printer work properly on the other hand is a rare thing...
Yes, this is absolutely true!@@SmallBatchFactory
1. This is really helpful content, thank you so much!
2. That Great Saiyaman helmet looks REALLY cool!
Thanks!
Only downside it was a tad bit small after adding the visor and compressed the guys nose while wearing. He still was very happy though, I guess cosplayers are used to endure such things.
I was guilty of adding round fillets to the base of models and quickly realised that chamfers are the way to go. Top notch video. Keep up the great content. Cheers 👍😎🇦🇺
Thanks! Only takes one messed up print to reconsider
I think a two stage hybrid thing works ok where you can have the bottom most part as a chamfer and then the joint between chamfer and the vertical part being executed as a round fillet. You can even make the chamfer a little steeper than 45 if you like for more of the joint to be rounded.
I work on these aspects all the time, but I also push the limits of what is possible with my 3D printed speaker drivers. Single walled TPU parts for the surround and spider for example
It's always cool to push the boundaries of what's possible, keep up the innovation!
Yes! Omg! These things have been annoying me for years. People either seem to try to recreate metal (or injection molded) designs that are too delicate for plastic or not suitable for 3d printing, or the go the total opposite route and make foot thick walls.
Foot thick walls are a thing for woodworkers dabbling into 3D printing I guess. It always shows what people are accustomed to
@@SmallBatchFactory seems like a lot of people get hung up on making prints stronger rather than strong enough.
I still have a box I made as a first vase mode test years ago that I’ve used to drop hot nozzles into during nozzle changes (with a wet sponge at the bottom and sometimes some water) and it both doesn’t leak and has been plenty strong enough ever since.
Using a threaded rod or a screw as internal reinforcement is such a great idea!
Thank you!
I'm a bit late but what a masterpiece i learned a lot and wil try to improve i also find it a good decision to keep out the sponsors but if you really need it it's your own choise. please keep going you are one of the best 3d printing youtubers ever
Thank you! Don't make me blush :-)
For now no sponsoring, let's see if I this is a sustainable choice.
One that you missed that always gets me is sharp corners (especially after long straight extrusions) on the intended build plane.
What constitutes "sharp" depends on the intended printing material, but with some large PETG prints even 90° is too much. because as it contracts slightly in two directions even a well cleaned build plate, with excellent z-offset and adhesion may break contact in the corner leading to severe warping.
The fix is SUPER easy too: just fillet the corner to a 2-3mm radius and it changes the forces just enough that it stays stuck. This is part of why a brim can help the corner naturally rounds off. If the vertical corner must be as square as possible you can blend the fillet into the corner in the first vertical mm or so. Or add a 1 layer thick "mouse ear" which is easier to clean than an entire brim.
This is particularly useful for more technical filaments that are known warp monsters like PC and PA-11 (and in extreme cases you can use the mouse ears + small neodymium magnets to help hold things down assuming the build plate is ferrous)
That's another bummer for sure. You can always add mouse ears in the Slicer easily. Adding a magnet it something I never thought about but is a very intriguing idea. Would need to be high temperature resistant though. I remember Prusa searching for a while until they found capable magnets for their bed.
@@SmallBatchFactory not really. Any Amazon small button neodymium magnet works in this case. Nd magnets have a curie temp above 300C. They'll hold to the steel sheet at 110C just fine. (yes Coercivity goes down above ~100C, but it's still plenty strong to hold down a .2mm thick piece of plastic)
I've read they will permanently change their characteristics of they get hot, but I'm no expert in this field.
great, now I have to redesign abunch of my old models. Thanks!
Sorry.... :-D I did that as well for models I printed more than once
I like how your head dances. it's hypnotic
Yeah definitely got another subscriber. 👍
I have to print models for mechanical engineering students - who also have lectures in additive manufacturing - every single one of them I "send home" telling them to fix the exact issues mentioned in this video!
Feel free to show the video in your class :-)
Fair points made apart from the radiator bracket. Don’t use a chamfer the parts as suggested. ❌ Less material in contact with the bed. Doesn’t matter how good your first layer is, minimal contact with the bed more chance for the print to fail. Flatten the one side as a feature ✅ However I do agree that parts should be designed using its manufacturing method in mind with considerations to the inherent weakness of that manufacturing method. Although people that share these models aren’t designing them with a view of selling them, in which designing correctly with research & R&D who would expect
I've printed stuff with very little surface area with no problems. For examle my hinges on the printer enclosure needed to be printed standing up for layer orientation with not much more than a millimeter in width on the build plate. It's usually fine, result may vary of course.
I think it's a matter of experience. It doesn't take longer to do it properly, but people lack the experience to know that. It's totally fine, may this video help some people to get better at it in the future.
When I looked at your subscriber count I was shocked, there are channels with hundreds of thousands of subscribers with lower quality videos and less informative scripts. Keep it up!
Thanks! At least they're growing rapidly. A silver play button would be really nice.
I have encountered all of these issues and many more, and frequently end up redesigning parts I find online to make them printable (often after wasting a lot more time trying to fix/patch a broken model when it would have been faster to just redesign it in the first place).
The one that really gets me annoyed is when people upload models that are simply entirely broken from a real object perspective. Not watertight, non-manifold surfaces, impossibly thin walls or intersecting corners... it's as if they designed the model to go into a 3D game engine and not actually be used to print a real physical object. This wouldn't be so bad if they posted the design files, but your options are quite limited if they only upload the STLs. There are a variety of automated tools that can attempt to fix some of these issues, but if they're bad enough I just use the design as inspiration and make my own with proper real dimensions.
Inspiration is usually my approach as well. I even do that with furniture...
Usually you can see it's not bad intention, it's a lack of skill. The clothes hanger bracket for example came with a source file and was very botched together. Changing this was a challenge.
3:00 dont use chamfers over fillets. instead, combine them! do a chamfer on the corner and a fillet on the upper corner of the chamfer. this gives a nice curve while still keeping good overhangs
That's also a good approach! I've done that in the past but totally forgot about that
I love that mold makers are trying out 3D printing! But yeah they probably need to keep working on learning how 3D printing is different. I've ran across my share of designs that look like they would work out better in an injection mold than a 3D print.
It's mostly a matter of what people have learned to adapt to. 3D printing is so young compared to injection molding designers don't have much experience.
I've even seen the Prusa video how Volkswagen uses 3D printing and they wasted so much support material...
The curved overhand thing is a huge issue I have, I mostly model and print figures and the backs of them always turn out rough, wish there was a way to fix it that didn't involve changing the design
It is nice to know that this is a common issue and not just my 3D printer
Only way to mitigate it a bit is finer resolutions, maybe variable layer height. Resin printers don't have as many issues with it since layers cure at once.
@SmallBatchFactory I looked into getting a resin printer, but the toxicity and post processing scared me off
@TheTurtleyOne oh yes. I bought one and I hate it. So much waste produced and a lot of fumes. I have a profesional mask but still... I don't even keep it in the house.
I think because its fairly easy to learn how to make an STL leads to so many not so great designs. I find that I don't really understand the best way to orient prints so I really appreciated you explaining how to make prints better. I shall have a look at your course, I want to lean Fusion 360 but I find it rather intimidating.
That's probably it. Almost no other hobby enables you to upload all your designs. I imagine there are also a lot of bodged woodworking projects out there, but you can't upload them.
If you have any questions about the course you can drop me an email. I'm always grateful for feedback too.
@@SmallBatchFactory we also have things like makerworld (Bambu Labs) they give us points when we upload projects and then they can be used to get discounts when you buy their products, there is an explosion of people uploading to get those points, I understand that Bambu Labs want to populate their platform with lots of content, and there is heaps of great stuff on there, but there is also a lot of sub-par and sometimes just bad designs on there.
There is an advantage to thick hollow walls. They have increased stiffness, since for a given bend radius, the outer walls that are further apart would need to compress or stretch more, which they resist. Many of these prints can even be done single wall just fine with a lot of void and you don't need a lot of infill just to tie the walls together and help it hold shape.
That's correct. I specifically referred to parts like the power strip holder "covering space". Stiffness is for the "solid blocks" I mentioned.
The prints you're talking about get their strength mostly from the top and bottom layers, like a square pipe or an I beam
New to the 3d printing scene. I love your videos, thank you!
I think I initially designed a part with arc overhangs in mind at one point. def helps for 90-degree overhangs but at the time, and last I checked was still the case, I found an edge case in the script that requires quite a bit of design consideration
While I appreciate the engineering aspect of things, as a Dane I do desire better designs too (for those not in the know, I am referencing Scandinavian Minimalism. Minimalism itself started in Germany (the BAUHAUS movement) and it is focused entirely on the functional aspect of a certain product, removing all unnecessary features. Scandinavian Minimalism started in Scandinavia (as the name implies) and is focused around retaining design features by integrating them in the functionality of the product). Jokes aside, layer line orientation, or rather disregard for the manufacturing method used, is also my arc nemesis.
This is why I tend to just make my own objects. If its super high in detail, thats one thing. But a simple clip or holder of some sort? Eh, design something in 20 minutes, spend another 20 adjusting tolerances, adding chamfers/fillets, and call it good
I usually tend to do that as well. The time searching for something remotely usable can quickly take more time than just doing it yourself.
It's a major thing I hate: designers really need to make sure multi-part models have the same start layer. I've had a ton of camera parts that have failed because 3/4 parts start at 0mm but the last piece starts at 0.2mm.
That's annoying. You could try to split the parts and put everything on the plate. A good advice is to always check the resulting print in the Slicer at least briefly. Floating layers are usually easy to spot.
Liked just for the "solid infill" comment about it... not to say it wasn’t a good video but outside my normal modi operandi. As a person new to 3D printing and modeling myself I find these types of videos very helpful. This is the first video of yours I have found but look forward to more in the future. It is why I enjoy watching @slant3d videos on the same topic so much. Thanks for putting this out there.
Thanks! Slant 3D is awesome. So many printers...
I don’t know why some people hate in designs that aren’t one part prints. I made a cable holder for my Valve index so the cable can be mounted straight down and added a wave spring to reduce stress on the cable. Had to print it in two parts or else one of the parts orientation would result in weakened sections that would break near instantly
I guess it's harder to imagine multi part designs. It takes a bit of engineering mindset to create something like this.
@@SmallBatchFactory Personally i find it way easier. You can split up you big part and assemble it later with screws and glue or just slot parts together. If you have single print parts you need to adjust so much about your print to get equal results.
Working out the right orientation or needing more support because some parts hang in the air. What about print fails? Now you have a failed print that takes 12 hours fail at hour 8. Instead you could have split it up and could have had a total print time of only 10 hours and split it up in into a 5 hours, two 2 hours and one 1 hour prints. One fails? Not that big of a problem. Only real drawback i see whit this is, that you need slightly more work after the print. So less useful for scaled up production that needs an assembled product. Or go the Ikea rout and deliver an assembly kit.
@LickItTM I'm totally with you on this. It can even make things a lot easier. Sadly many people can't wrap their head around it
My most popular design on printables/thingiverse has a fillet facing the print bed ... sorry about that. Now that I know about the "poor man's fillet" (ie, a chamfer half way and a fillet the rest of the way), I need to upload a v2 of the models, but I just haven't gotten around to it yet.
Blessed is he who confesses his sins. I too have sinned in the past.
Thin walls should be related to the extrusion width and layer height, not just multiples of nozzle width. In Prusa Slicer, you can see the recommended thin wall thickness for various number of vertical shells at different layer heights. E.g., 4 lines at 0.2 mm should be 1.67 mm, not 2 mm
I meant to say 5 walls anyway. With Arachne that doesn't matter anymore
Gosh, I really appreciate the advice on cylinder orientation and edits for stronger prints!
One of my personal bugbears is electronics boxes modeled with broad, flat sides and too-thick walls. These are horrifically weak, especially against torsional forces.
In this case you might want to add ribs on the inside, ideally at a 45 degree angle. Those add a lot more strength than simply making walls thicker
I understand the thick walls, as they dont really bear any load they can be much thinner than usual. But I disagree about the other thing you mention.
What do you base that conclusion on? Closed box sections are actually incredibly torsionally stiff. Even without a lid a square section is fairly torsionally rigid, only when you remove one of the walls the stiffness drops dramatically. See also chapter 3 of Gere's Mechanics of Materials.
Of course ribs can be added to increase the torionsonal stiffness even further, but for an enclosure is that really necessary? When is that ever subjected to torsional loading?
Good advice, but for power strip, especially schuko, the forces are not insignificant. I ended breaking a few power strip casings just because the plugs were stubborn. Now I have replaced most of them with aliminium ones.
My own wall mounts are shown briefly which I tortured a bit before the video just to be sure (they're Schuko) . Haven't had any issue with them so far, being printed with PETG.
I design my parts specifically to be reinforced with screws and/or threaded rods. I'm glad to find I'm not one lonely weirdo.
You're not, even the Voron project uses screw compression in many parts :-)
Makerworld at least requires you to upload an actual picture of a print. People try to get around it and the community usually reports it when they see it. I avoid models that have no real prints or prints that look terrible because it likely means the file is unprintable.
That's a good policy to have. Especially Thingiverse is flooded with unusable models.
While your example about not wasting filament for the plug holder is right, its quite a simplification to say that a thin solid wall is always stronger, when there are in fact many scenarios where a wide part even with very low infill will in fact have much better mechanical properties.
It very much depends in the use case of course. That's why I said "geometry covering empty space". If you need something resistant to torsion forces, spanning over a bit of distance you need thicker walls for sure
You are 100% correct. Often the inside does not carry much load, most of the shear loading is carried on the outer skin. This is exactly why people prefer to use hollow tubes instead of solid bars. For the same amount of material a hollow tube (of much larger diameter) is much stiffer and stronger in any DOF compared to a rigid rod of comparable mass.
In such cases it's worth to try solid infill every X layers instead of conventional infill. Eventually use concentric infill at very high percent. Sometimes it works as another type of fast wall printing
@@SmallBatchFactory It's just a widespread issue I see in 3dprinting videos talking about "strength" in very simplified terms. An actual proper video about different mechanical stresses in 3d printed parts would be wonderful, because the vast majority of "strength" videos are mediocre at best, and outright misinformation at worst.
@hypertectonics7009 that sounds like a good idea, noted!
thank you very much for showing examples and how to fix
Literally did this yesterday. the part was a battery holder. It was too tight and required 2 different lengths of screws unless being screwed into a thick board or a blind space. Tried to modify it but fusion 360 wasn't having any of it.
There’s a TON of design consideration that goes into every manufacturing method, except 3D printing. There, the best most people do is try to avoid supports (even though they’re not THAT big of a deal these days) and don’t really consider a lot of the other advantages/disadvantages of 3d printing.
I guess it's a lack of experience. Thingiverse is mostly populated with hobbyists. I downloaded a model from a coffee mashine company once for example and it was awesomely designed.
@@SmallBatchFactory seems like there’s a lot of people who picked up “tips and tricks” and just kept doing things the same way and not a lot of new people learning to design stuff themselves now that so much is available so easily.
But I think videos exactly like this are what’s needed to both get people interested in at least modifying designs for their own use, if not learning to design things specifically for 3D printing themselves.
Guess it’ll just take a while for people’s imagination to not be based in traditional manufacturing techniques/designs.
@heyspookyboogie644 yes definitely. Most people can only work in small iterations. That's why radically different products usually flop until we'll arrive at the same product years later.
I think there are too many printer reviews and too few "fundamental education" channels in the niche. People seem to agree.
Good tips!
I've recently been educated to the chamfer/fillet on the bottom edges. Those rounded edges look so good in Fusion360 though :)
Oh yes they do and I'm certainly guilty myself. If you keep it to a millimeter or two they still work well. Many overdo it and that's when they fail miserably.
My favourite example of a great design ruined by a design flaw is a case for an esp01 relay box i saw once. Everything about it was perfect, except the clip thats meant to hold the wires down had both screw holes placed too close to the edge causing them to actually clip through the sides of the clip causing a whole mess of random polygons in the negative space. No idea how they managed to upload it without noticing the error but the designer just dropped off the face of the planet
Theres also no way to fix it without completely remodeling that clip from scratch, which is doable but tedious
Sometimes Slicers manage to fix those issues somehow. I guess that's what happened. Either that or it was never actually printed before.
I spent 43 years in computer hardware support. After reading the comments it sounds just like the issues I ran into. The people/ engineers designing things never actually put it together, use it or even tested what they came up with, hence the problem. I have had many classes learning brand new products and instructors literally throughout the manual. Because there was such a poor match between the manual and real world equipment.
Oh yes, a lot of "meh, good enough" approaches.
@WhiteG60 this or sometimes even gross incompetence. At least with a lot of early Chinese printers there was a lot of "everybody does it" implementation without grasping the reasoning for why to do it that way or not.
For the hooks, my solution would be to have the bottom of the hooks extend to the bottom, with a 45 deg overhang.
That's also possible. There are no wrong answers as long as it prints decently :-)
Turning off part cooling at specific layers and slowing down to about >50mm/s (@0.2 layer thickness) will create near mold injection like strength. Depending on your printers hotend of course, you may need to be under 30mm/s. Especially if it's a open air bed slinger. The bed moving back and forth cools the part as well.
Yes, either go slow or you need cooling. I prefer designs you can print fast with no issues to going slow.
Not always possible my man. You also have differences in printers. No two are the same. The printers own accuracy comes into play. I try my best to avoid certain shapes, provide tolerance, etc., but it doesn't always work 100%. You don't want to fillet a bottom edge, use a chamfer. You can fillet top edges and vertical wall edges. You can reinforce prints using ribs and rafters. Just incorporate them into the look if they are visible. You want to fillet where edges meet between connecting parts. It will increase strength where the part connects. Arcs distribute stress well. Print parts solid infill that should be solid.
Even a cheap bowden CR-10 of the first generation gives you really accurate prints. The rest is tuning, given the machine is has no defects.
The rest I totally agree with.
But... Fillets look so smooth! The allure of the rounded edge. I have started combining chamfers with fillets so the fillet starts at 45º.
That's also a very good approach. They're only good if they still look smooth after printing :-)
I don’t use screws or threaded rods for alignment issues, I make the weak area hollow but then I fill it with epoxy or print a properly oriented piece i can glue in
I did this with a Vesa mount adapter, I wanted something that would last forever so I just put in big structural voids I could fill up with resin
Sounds like a plan, especially with resin. Although threaded rods are less messy in handling.
@@SmallBatchFactory I just haven’t had time to play around with threads and get used to it, so for now I’m paranoid about it eventually delaminating and I’d still end up using glue or resin to fill the crack (especially with the hook example where the hook could twist to delaminate)
@Pentross that's understandable. Layers compressed with metal are virtually indestructible. Me studio lights hang in such printed parts so I pretty much trust that concept. I personally have more aversion to Resin than to metal hardware :-)
Excellent tips and observations. All good things to be aware of. Thanks for sharing
Thank you!
This video couldn’t have come at a better time! I just started designing my first models.
I’ve already even made the first two sins from this video, now I feel like I’m truly part of the 3d printing community! 🎉
I started your fusion360 course a few days ago, and I’m really happy I did (and also terrified of the process of actually learning).
Although my current needs for models haven’t really required anything more than Tinkercad, I’m going to slowly learn the material so I can feel comfortable enough to tackle any problem in Fusion360, and your course is an excellent resource to have on hand for the journey.
Welcome to the family, glad I could help! I think it's the ideal compliment to the course, helping improve Thingiverse & Co. A bit.
Makes me really proud to hear that! Let me know in case you get stuck anywhere. Getting that beginner view really helps improving explanations where necessary.
@@SmallBatchFactory I’m sort of “stuck” in a sense as I’m trying to motivate myself to finish the last videos of the “creating your first model’ section. It can feel like a bit of an information dump to a newbie like myself.
I wanted to remix a fan bracket and figured I’d try your course as an intro to CAD/Fusion360. I kept trying to use what I had learned in the first chapters of your course to get started, but I couldn’t really figure out how to sketch out the dimensions of what I wanted drawn on paper yet.
I’m sure it’s more comprehensively laid out in further lessons, and I just had super unrealistic expectations about quickly modeling the design (I was so eager to design the part and I am very inept at viewing the world in a “problem solving” manner - so I was really operating in unfamiliar territory.)
That’s when I tried out TinkerCAD and found “hey this works just like a slicer, I can manage that” and a day or so later I had the part printed. Poorly, and without the ability to make quick adjustments to dimensions, but it was made and I really needed that dopamine hit. Iterating on that design I can do things that took me 2 hours in 2 seconds now.
When I get some more time in the next week, I plan on sitting down and taking a proper crack at the course. I feel like the lessons I’ve learned playing in TinkerCAD will only benefit me, since I literally had no idea what I was doing before, and now I’ve come a long way (with still soooo long go!)
So if I had suggestions for the course, I’d say add in some ways to give my ADHD ass some dopamine hits early on and frequently. Really shove those dopamine hits down my throat, learning is painful without them. I’m possibly a unique case though in that sense 😅
Oh yes, doing everything properly surely is a lot of information. My first attempts also were mostly "I want to get things done".
I did my best to get to the first practical model as soon as possible and that also means there are still a lot of things to learn to create more complex shapes. I guess that's what you experienced, you have something on mind that's actually easy to build out of 3D shapes while it takes some getting used to doing it with sketches.
Maybe we can use your project and add it as additional practical example to the bonus section.
You're spot on! You said it better than I could've. I wanted to create a more complex shape, without really knowing that's what I wanted to make.
I have a couple more things on my "wishlist" of learning. A replacement latch for my freezer door, some cool accessories for my cargo bike (cupholder for my daughter perhaps?). Silly stuff like that I guess. I think I'll be able to do it, maybe even quite easily - especially once I have a bit more time to watch the rest of the course.
I'm sure you'll get used to "thinking in sketched" with a bit of practice!
Something mounted to a bike frame or some other tube would actually be a good example to add to the course. I already did a few bike accessories screwed onto the frame and this might be not too obvious to some. I'll definitely add that to the ToDo list!
Every single 3D file I download for printing needs repairs in one way or another. Sometimes there's non-manifold or detached geometry, sometimes the tolerances weren't spec'd properly, sometimes flat faces aren't actually flat... Open source is the wild west, you get what you get :/
We can't really complain if we don't pay for it. But to be fair I have the same experience with furniture. I usually already have an upgrade plan when leaving the shop or even build it from scratch.
Those separate hooks might also be heat-welded to the back side of the base plate with a soldering iron.
Definitely works! Others have suggested a tapered back end, would rattle a bit though.
Great video. Lots of great lessons for someone like me that is just learning about designing custom parts!
Small things i disagree with. If you are good with 3d printing you can get ~80% of the strengts in both ways on layers. Most people have a issue because they either use way to much cooling or way to low temperature. Sometimes trying to go to fast.
For something meant to be aesthetic i wouldn't care about layer orientation.
3d printing curves at either the base or top of a 3dprint should be done with incremental steps/variable layer height. If i got a 3d print that got a big radius on the bottom that for some reason have to have it. I would start with a 0.02mm layer after the initial layer height. And increment it higher. Also tend to let the bed be 5°c colder than usual to avoid the layer flexing down.
Problem with 45° is they may be to sharp in some instances.
I've never seen any test showing 80% strength across layers with all samples having the same section size.
Variable layer height works of course. In many cases it's a bit of a crutch IMO
I really don't think that this powerstrip holder was bad. It should be plenty strong!! My mother used to rip the sockets straight out of the wall in our place. trust me some people do not treat them gently.
It's just an example anyway. It's more about the core message "do walls appropriately".
I can tell you power plugs in Germany hold really Strom in their sockets and even with the I can't rip mine from the wall. I tested is specifically because of comments.
8:14 I personally like to plan for zero or near zero tolerance in Fusion 360 and use the “Exclusive” slicing method in the Cura experimental tab. However I do not know if there exists another similar option in other slicers.
Orca Slicer has some contour and hole compensation features I have yet to try. Compensating in the model is a bit of a crutch for now.
By far the biggest issue is people just giving an STL. It's so prevalent that people thank me when I also upload the CAD file.
The websites should REALLY incentivize people to also upload whatever CAD file they used to create the STL
I also find it super problematic when people design for a certain nozzle size when it's not needed. I routinely print with 0.6mm nozzle (and frankly almost everyone should!), but some models needlessly claim that they are optimized for 0.4mm. If you do that, at least give the CAD file!
Indeed, I had people thank me too. Especially for my bike wall holder since you can customize the wheel size to any diameter
I'd say to fix the key holder you should shift all hooks down and make them square, at that point they will touch the build plate and you can print it rotated to get that additional strength, it does make the mounting plate a bit weak but for a small load if you don't torque the screws too much it will be fine, that removes a lot of hassle for a less experienced printer, adding threads that size is not that useful in my opinion and uploading something with such threads will just be problematic to people that have bad settings I'd say you better thread it yourself with the screw and leave a hole, also you should have added a recess for the screw at the back cause that is going to be against the wall
You're totally right, it's just a quick inspiration of how you could do things. The rest is up to the designer.
The original was designed for magnets. Adding a recess for screws would also work. Really depends on the individual use case.
Rounded corners in the first layers help reduce elephant's foot
It's one way to compensate it. IMO this belongs into Slicer settings for the machine. Putting it into the model leads to varying results.
Usually 3D printing is like trying to made pocketwatch using dull and wearout tools. Its possible but must use suitable tolerances.