I printed a model of an elephant's foot and enabled "elephant foot compensation". My printer disappeared and now I have a mini black hole in my print room. Thanks, Angus.......
On the upside, a quantum singularity is an easy pet to maintain, and it'll eat anything. You have a great way to get rid of your leftovers and trash, and if you can train it to spin, you can generate power for the house.
Go to any automotive shop and get an inexpensive set of "feeler gauges". These are a bundle of metal slivers of calibrated thicknesses used for setting spark plug gaps. Choose the sliver that corresponds to your bed offset and slide it under the nozzle until it barely fits. Attempt same with the next size sliver down, if it will not go, then your know your exact nozzle offset is somewhere between the two. Much more accurate and repeatable than the "paper pull" method and only costs ~$10.
My 3D printer came with feeler gages so I was sort of shocked when he said people use paper to gage the height for the nozzle. Like people who sell these expensive machines don't give enough of an F to give you the tools to use their product correctly? Why?
Right on, these are old old old machinist tips that only the oldest and probably deadest old timers used to use before numeric control. Get a book from a guy named dave gingery (dead) titled "build your own metal working shop from scrap" it's filled with hundreds of tips like this to create precision machine tools without precision measuring tools. I just got my 1st 3d printer and so much of the advice online is watching people struggle to find these ancient answers
One thing to mention: if you slice and print a 20mm cube, no matter how well calibrated your printer is it will *always* measure slightly less than 20mm in the end unless you compensate for material shrinkage by resizing the part slightly.
Exactly, if you home all the axis and level the bed with apiece of paper then you will have a .1/.2 mm gap at z 0 position so when the first layer is printing it will lift up another .2mm depending on the layer height resulting in a .4mm gap. So when i level my bed i lift up the z axis .1 mm and level there with a .1mm piece of paper resulting in a thiny little gap between the noozle and the bed at 0 position because the paper has a little room to play. I was looking for this technique online but didn't found anybody who used it and i didn't know if it was right but it works for me!
I used to do this too before I got an ABL system. I kind of still do it, but now it's done in firmware and for a different reason. In Marlin I have LEVEL_CORNERS_HEIGHT set to 0.4 and I use a 0.4 mm feeler gauge. Heat the bed, level the corners with the feeler gauge, run a G29 and that guarantees my leveling springs stay at the same level of compression essentially forever (having it set to 0 would compress the springs by 0.4 mm every time I ran a corner leveling sequence, having to at some point loosen the screws when one of them eventually bottoms out).
Elephants foot compensation is “horizontal expansion” in Cura. I’ve used it in a pinch to give models a bit more play or to tighten up gaps for a better fit. Super helpful!!!
Dude your like the best guy on youtube who does not focus on just one printer model. I like how you make most your videos more general so those with a different printers can follow along anyways. Thanks angus! 😄😀😃
I think it's one of those things people learn to live with because it means your first layer adhesion is really good. Most people don't need their parts to be dimensionally accurate to where it matters.
Ex Queue yeah totally agree! My 3D printers create my living currently, so in the past I’d just take a fine rotary tool and take the flair off for closer to accurate dimensions, but still not perfect. Eliminating it all together is the best thing to do but requires a decent amount of tuning. On both design and hardware.
Thanks for this, it validated something I've been doing for months. I needed to print some structural pieces on a resin printer without elephant foot and figured a 0.5mm chamfer would compensate for the 0.5mm elephant foot I've been getting. The nice thing is with resing elephant foot only occurs on the bottom layers that have been set to a longer exposure, so it's very easy to figure out how high to make the chamfer and the part is still very dimensionally accurate.
Found this channel earlier today. As a current biomedical engineering major, accurate and precise 3d modeling and printing is a necessary skill to master. Thanks for your outstanding videos, subbed for sure👍
Hello Angus, this issue has been killing me this week with a filament that I put a lot of trust in. I've only been at modeling and printing for about 2 months now, and your videos have been hugely helpful. This one takes the cake though. Thank you so much!!!
Thank you very much for all the time you spend teaching everyone all these tips and tricks to get by on our printers, it's much appreciated! Please, don't ever stop!
I love how the elephant foot compensation and whatever makes support material easily detachable are both things that are activated by default in the Prusa Slicer. I just switched over from using Cura slicer with a Creality CR10 Mini to Prusa Slicer with a Prusa Mk3S, and I'm very happy with these two features, in addition to the magnetic PEI bed and the generally excellent functionality of the Prusa Mk3S printer.
Thanks so much for talking about rafts. I've always thought they seem super useful but haven't seen many youtubers using them often so I was a little nervous. but this makes me feel better about them!
I don't get why people hate rafts, for some prints they're necessary, especially things like rings where there's not much contact with the bed. I find even a 1 layer raft can make a huge difference, and it really doesn't take that much longer, or use that much filament. I probably waste more changing filaments to make sure the colour has changed!
@@havenview It totally depends, one of my printers absolutely hates printing circles smaller than 20mm on the first layer (probably down to the fake easy peelzy, but I'm too cheap to replace it because it works fine with everything else!) so a brim really doesn't help. My other printer on the other hand can get away with it just fine.
If rafts didn't have a use, they'd not be in the slicer settings :) and people who argue that you don't need rafts simply haven't had the situation where they're needed or they've found an alternative solution... simple as that. They are wasteful of material and they do add some print time but they've really good in certain situations.
I already use most of these things to stop elephants foot on my prints. What i suggest is putting a little bit of glue from a glue stick on the bed before the print and making the first laver height the same of higher than all the others. Really good tutorial. You did a really good job at explaining the fine details for people that are new to 3d printing. A++
This tutorial couldn't have come at a better time :V I had a whole plethora of issues that I only had myself to blame for - I'm a rookie and made the silly decision of doing my first ever nozzle and bed upgrades (despite the fact that my printer was dialled in to near perfection with my old nozzle & bed), right in the middle of a bunch of commissions for some pals. Looks like I was way too confident in my ability to troubleshoot because it was one problem after another, (bed levelling and the nozzle wasn't fitted properly, ontop of speed settings etc.) I kept my head on and solved everything pretty quick until this damn elephants foot which I've been at for days with little improvement. I was just about ready to call it quits and refund my friends till you uploaded this and now I'm happily on my way to finishing each commission. I think this calls for some more celebratory upgrades. :^)
All your videos are amazingly useful even if we know about what you are talking about already but the fact you emphasize it make the information stuck and now I am finally convinced to use Rafts for the parts that I intend them to be precise.
And in Cura it's called "Initial Layer Horizontal Expansion", but instead of a positive number, you have to enter a negative, i.e., -0.3mm or whatever. There's also the command "Horizontal Expansion" which is used to compensate for holes that print out too small (or large). Again, enter a negative number to compensate for holes that print too small and a positive for holes that are too big. Both of these work pretty well for me.
what a good video! Due to other issues with my printer, I've had to make some tweaks that drastically increase the elephant's foot issue for me. It's interesting that this video focuses not on "fixing" the issue, but more on how to compensate for it. Excellent video that just got me re-inspired to print again!
I tend to just eyeball my corner leveling. I find its suprisingly more accurite than basing the height on how well the nozzle grips the paper. Just use babystepping to grt the hight perfect while you print your skirt. Also when leveling corners make sure there isnt any plastic on the tip of the nozzle. Made that mistake plenty.
Great troubleshooting video! You can also set your first layer height in the slicer to 0mm or whatever with a piece of paper or even pause the printer at its first layer/temperature to level it then. Anyways, I just eyeball the level and either adjust screws or babystep/set home offset until things are level while printing. I assume a tolerance of .2-.3mm with a .4mm nozzle for projects so there is no need for a raft. Another tip is lowering the bed temperature. I was printing PETG at 85 on the bed and just reducing the temperature to 80 reduced elephant foot almost completely.
I wish I would have had this info 3 years ago when I first started 3d printing! I knew nothing about 3d printing besides I was getting a computerized machine with a mini hot glue gun on it. The first 6 months or so of 3D printing was figuring out how to get the perfect first layer with no "squeeze out" was what I called it then. In the beginning I thought maybe my filament was too hot, or the heated bed was too hot and my prints were melting inside themselves! I just have to say you have done a glorious job of explaining what elephants foot is and how to compensate for it. Thank you thank you!
Oh! That explains that. I'm very new to 3D printing, and have so far printed one thing (on a library printer) and it definitely had elephant's foot. I was wondering why it was a little wider at the base. Nice to know that my plan of adding a chamfer round the base edge is actually a recommended solution.
This was the video I needed. My prints come out pretty well and I'm happy with the dimensional accuracy, already better than some stuff I've seen at Uni. Except for the first 3-5 layers.. I was looking on how to fix this. Trying it tomorrow! Next step: the remaining tiny bit of Z inconsistency and the infill structure slightly popping through
After having issues, I bought your tolerance test. First print? .3 required force, the others broke. Second? They spinnin'! ALL of them! I was having a major issue with the infill causing expansion on the perimeters/walls, and after reading who knows how many forums (Prusa, reddit, etc) it seems that more walls would help. Sure enough, if you're having issues like that: up the wall count! So, it seems there's something unique about that middle one...
Root causes of Elephant's foot, extruded width is wider than you spec'd due to wrong height, perimeter printed last gets nudged out by too wide fill, first layer stays too hot gets crushed. Fix these and get a perfect first layer: 1. Robust tuning to handle bed variation. First layer needs to be the thickest layer so changes in bed level are a small percentage of the first layer height. A 0.1 mm variation in bed level is 50% of a 0.2 mm layer, and results will always be crap. Use a thick first layer like 0.32 and pick a layer that is divisible by 4 full step motor steps for most Z accuracy. 2. Print first layer, stop, peel up and measure on all sides with micrometer. Adjust nozzle height and bed level manually until what you peel up and measure is as close as possible to the first layer specified in slicer. Measure near each level screw if possible. 3. Leave room for extra material due to variation. Set first layer extrusion width to the widest to get great adhesion, like 120%. This both increases flow rate and spaces the traces farther apart. Then set first layer flow rate to 90%. This will allow for variation in bed level before the print gets too big. 4. Print perimeter first on first layer. This avoids overly thick fill traces from bumping out the perimeter. 5. Slow down enough so first layer is cool before second layer goes down. Second layer with the fan blasting, to avoid crushing and expanding the first layer. Those setting at 1:30 are the opposite of what you want. Use First layer height 0.32 and First layer width 0.48, first layer flow 90 %, adjust as needed. Enjoy the perfect first layer.
Ultimaker Cura has a tick box for printing rafts. I don't recognize your slicer. Cura is pretty good. Besides, I'm new. Okay, not quite as new after the previous line has aged a while, but I just want to let you know that I paid for the Clearance and Tolerance 3D Printer Gauge to test my profiles. I sliced it using .08, 0.10, & 0.12 layer height, and as expected, when printed using the .12 layer height, both 0.2 and .15 are stuck tight, and I had to break .3 free. But I surprised myself when I sliced it with the 0.10 layer height, then printed it. Imagine how awesome I feel that on an Ender 3, I have it tuned well enough that all 5 are free and rotate easily. I didn't have to break them apart either. And, I didn't know that the middle piece comes out when all 5 outer discs are lined up with the cutout over the center. I didn't bother slicing and printing the .08 version. Why go that low? Thank you for creating that gauge, so noobs like me have a target to shoot for. If anyone thinks a Creality Ender 3 can't print well enough to get all 5 discs to break free, without breaking free, they just don't know how to tune the printer correctly. For y'all to understand what I did, I first got rid of the elephants foot by using first layer horizontal expansion limits, then a couple other similar settings. I'm not going to go into any more detail. I'm willing to share my profiles, but if you really want to be good, you need to struggle to figure things out on your own first. Oh, look for hints in online sources, but remember, how you set up your printer in the first place will make a huge difference. The Ender 3 is a kit, so treat it like a challenge to get everything perfect, e.g., perfect 90 degree angles where they belong. Once you have that done, then tune it for printing. And welcome to a new drug. A very addictive drug. But, the highs mean you are serving someone in your life by creating things to make their lives easier. Or more enjoyable. I make stuff my grandkids love. Yeah, I'm an old FART (Fathers Against Raising Teenagers - an odd club I thought I as in when my kids weren't married and all grown up and moved out.)
This was the biggest problem I've had to deal with lately. Thank you so much for putting this out there! Found the "Initial Layer Horizontal Expansion" setting in Cura and I'm going to try it tomorrow :)
Wow, I have been fighting with this for a week. Im really new to 3D printing and its been a bit of a nightmare to get it all going. CR10 but the glass was apparently warped out the gate. I bought 12x12 panes from lowes and it took me all day to dial that in. I just ended up using rafts like you said as I still must be off a touch or Z is somehow or being told to increase z height at the start. Thanks for the video!
A few words about crazy and complicated 3D designs: procedural generation and scripting helps A LOT. F.e. - using openSCAD in conjunction with Python and SolidPython library. It made possible create 3D models literally impossible to hand-draw in regular CADs, yet going to need at least elementary programming skills and intent to dig into it.
Just at 11:30, I love rafts. I don't need them anymore, but on my first days working on a makerspace machine it wouldn't have come along without a raft. This bit of extra material is often worth compared to failed prints.
Our 2 stratasys printers default to using rafts otherwise they would fuse, the markforged prints directly to the bed but it always looks fantastic. My guess is a combination of accurate leveling and software correction. My printer at home uses foam blocks with HIPS so that's fun.
I fiddled with the settings and tightened my belts to try to get the perfect C cube and I got it bang on 20.00 mm in the X, Y but the Z-axis was 19.96 mm, I knew it was Elephants foot that was causing the loss in the Z-axis height. I am now going to slice my cube with compensation to see if I can get that perfect C cube.
Elephant foot compensation probably won't help with the Z axis dimensions. It will print less plastic in the X and Y axis on the first layer, but the Z axis will still be squished. A raft might help though. Or you could increase the size of the cube in the z axis in your slicer by like 0.04mm
If I remember right Cura has an adjustable number of layers to apply elephant's foot compensation; I use it on almost anything that isn't printed on a raft. And yeah, when I really need that perfect accuracy, I do swear by using a raft. One of my prints required several copies of a part being printed on a hexagonal raft about 10mm wide, and since they pulled away so cleanly I made the resulting rafts into coasters for beverages! No wasted material, they look cool, and they actually work quite well!
cool video dude, just recieved my ender 3 v2, not had chance to build yet, so been watching as many tuition vids to get up to speed!! great info thankyou
Recently all of my prints have been a little off on the base, like it was squished, and i just couldnt figure out what was going on, thank you for this!
I haven't had elephant's foot on my prints ever since i got and started using Simplify3D on my paper leveled CR-10. It's nearly all in the slicer as far as i'm concerned. My paper method consists of me pushing the bed down out of the way of the head by .3 and starting a random print that puts the start in the center. I then stop the print (but keep the heat on). Using the paper i raise the bed on all four sides evenly until i can do the slip fit (on mine, because the fan motor got hit at some point, there's a bit of harmonics, not enough to hurt prints, but i can instantly feel when the tip is transmitting vibrations through the paper on the bed (and my hand by extension). Once that's accomplished, i do a half way square up, recent to make sure i didn't lose the center height, and finally do a (on bolts) square up. Obviously all this is done with a fully covered glass bed (classic paper tape, flattened down with a spatula). The paper i use is just a on old business card i have tons of, i think it's twice the thickness of regular paper. If i observe any issues, i raise the starting of the print via Simplify3D. Now, why do i say that it's all in the slicer? Because before, when i used Cura, i had inconsistency after inconsistency. Not enough to make me level the bed a lot, but enough for me to note it. With S3D i just start the thing, make sure i don't have any drip, and leave it be. I level the bed only once every 60 prints or so, because i like to clean the tracks and wheels and lubricate the bearings.
To mitigate the gap created by paper thickness when touching off, try using something other than paper, which is .003" in thickness. I suggest getting some steel shim stock, available in .001" thick so it takes 2/3 of the problem away. Another choice might be the cellophane wrapper on many cigarette boxes, if you know someone who smokes. It is most often .001" also, but a bit more flimsy and tears easily. I have used shim stock to touch off tools on a cnc mill.
*Doesn't like to use support material and has an entire video on how to avoid needing them* *Solution to elephant's foot: use a raft* All jokes aside, thanks for yet another useful video, my prints would suck without your help!
I use normal zikzak support instead of a raft. Usually I print a model 3mm in to the air. This works very well. Especially with ABS to protect against warping. Enable „brim for support“ to get more bed adhesion. Using support means less material, far less print time and basically the same result.
I dunno if it's just me, but I haven't suffered this problem at all in the past year or so. Slicers have been coming with very good settings to avoid it, and model quality has improved a lot since people are learning better CAD techniques. I suspect your newer videos have something to do with that trend :D
Yeah now that you mention it I haven't had any issues with the latest batch of machines. Mesh bed levelling and print surfaces mean we don't have to drive the first layer super close anymore as well, which is nice 😅
dont forget, you can also mitigate some with a well adjusted and leveled bed AND proper z axis comp. (provided that bed isnt warped) PLUS having a bl touch can help for beds that are not prefect.
Concerning the "Elephant Foot": in Cura, look for "Initial Layer Horizontal Expansion". Putting a slight negative value there will do the trick. This would be a way more elegant solution than modifying the design of the model... Then if you have issues with tolerances and dimension accuracy, have a look on how to calibrate the "Horizontal Expansion" value. This value differs a lot between different materials and a little between different settings for the same material. (I get values like -0.07mm for PETG on an Ender-5)
I use the following technique: use a feeler gauge, not paper, to set the bed height to the required "first layer height", then configure the slicer to pint the first layer at 0 mm. this seems to work fine
Thanks for this video. I have a persistent EF problem on my Ender 3 but all the other videos I have seen only talk about the EF being on the first layer. Mine is visible up to as many as 10 layers and all protruding outwards, not concave like some you showed. I have tried many adjustments but I will now revisit the problem.
Dude it happens to me on my ender 3 pro as well. I'm thinking the stock bed surface is garbage, but when I'm leveling the bed, I make it that the gap is slightly tighter than a sheet of paper and that helps.
Use a proper set of feeler gauges and set the Z to the size of the gauge you choose to use before starting to level. I find a 0.4mm gauge has enough ridgidity but still allows enough flex to get under the nozzle assembly. Feeler gauges also allow you to level when the bed and nozzle are hot which also increases accuracy.
Col_Weab976 YT you should get one. They are a lot of fun to make things with. My grandson asked me what I was going to do with one. I told I didn't know. But since then I have made quite a few things that wood or steel was just not a good medium to use. Plus now you have a good excuse to learn a 3d CAD software to keep your brain from turning to mush (it may fry it though). At 66 I am still learning and having a blast making things.
There are two different elephants foot effects. One is the first layer squish, which you described. The other is where the layers above taper inwards and then come out again. I think it is because close to the bed the plastic takes time to solidify and is not fully hardened allowing the next layer to contract inwards. I mitigate that by dropping the bed temperature after the first layer. It also seems to be worse with low infil percentages, so I often use 95%.
Very interesting and informative. I think it is also worth mentioning the *temperature of the bed:* the squeeze of the first layer is possible because the bed heats plastic enoguh to be able to deform. By reducing the temperature of the bed after the first layer from, say, 60 to 50 degrees Celsius (PrusaSlicer has a specific option to set this), also reduces the squeeze of the first layer. I don't like using chamfers to correct the printed object, as it is not actually a part of the object, just a trick to overcome the squeeze of the first layer. But it works ^^', just as rafts do; sometimes you have to do what you have to do in order to have the printer make the object well
I always thought this happens due to too high bed temperature so that the bottom piece of the print softens too much and collapses under the weight of the printed object. Because to be honest, this effect remains for the first few layers, not only the bottom layer.
As far as leveling the bed: When I use my printer's leveling function, the Z height is at 0.2mm, so all you have to do is adjust the bed to be 0.2mm (~0.008") away from the nozzle. If you fold standard A4 print paper so it has double the thickness, it's around 0.007", which gets you pretty close. You just want to make sure the machine is at the coordinates that it thinks it's at.
Hi Angus, found an issue with printing and figured you were just the guy to help My printer, for whatever reason, prints any inner circle, such as a hole through a block, as a square-ish circle, like the cylinder was squashed on the sides. Any idea what could be going wrong? Many thanks
THANK YOU! This was a very well done, very useful video covering the entire spectrum of the elephant foot issue and giving valuable advice and great solution options on how to remedy the issue! 👍👍👍👍🙏🙏🙏👏👏👏👏
One aspect to consider is bed texture. My Artillery Sidewinder X1 v4 has about a 1mm waffle pattern in the glass bed. It was driving me crazy, and no amount of bed leveling, slicing, or model compensation was giving me perfect parts. One "trick" was to cover the bed with glue-stick, let it dry, then scrape off as much as I could using a very sharp wood chisel. But what about when I wanted the bed texture to appear on the print? The brute-force method is just let the print do what it does, then do a quick pass with a deburring tool. Which works great where the first layer edges are straight or gently curving, but fails badly for tighter curves and sharp corners.
I heard that you're supposed to make the first layer thicker not thinner, for adhesion. No need to over smush the material if there enough being extruded. The excess material can also help it get into small crevices of a textured build plate. Although looking at one of my prints it seems more like "elephants layer" thane "elephants foot". (the first layer kind of sharply sticks out somewhat)
I used to have horrendous frustration with first layer adhesion problems, so much so that using the printer was as much a pain as it was a pleasure. Using the paper method for getting that optimal first layer thickness to allow good adhesion was an absolute nightmare, particularly as my bed is not the flattest. I was thinking of getting some automotive feeler guages to help with setting accuracy but I decided this was getting ridiculous and there simply had to be a better way, most likely involving some sort of glue. After some experimentation, I settled on a diluted solution of builders PVA glue and water, around 10:1 so it has the same consistency and colour as milk. I thinly wash the bed with this solution using a brush, leaving a translucent layer. By the time the bed has heated up its all dry, leaving a slightly shiny and vaguely tacky surface. You don't have to wash the whole bed, only the area that the print will make contact. Since using this method I have never had a single failure related to bed adhesion. Removing the print is easy too, it just cracks right off when the bed has cooled off, far easier than my experiences with other print beds. After a number of prints you may need to clean the bed of residual PVA that might have built up but this is easy as PVA is mildly soluble in water even after having dried and isopropyl alcohol just whips it off. Having a 100% repeatable solution to bed adhesion allows far more lee-way with your initial layer height and is amazingly forgivable with a non flat bed. I rarely use the paper adjustment method anymore, I use it once to get a rough setting, set the printer off and manually fine adjust once only when its printing a Brim layer. On subsequent prints I need make no adjustments, I don't even need a Brim for adhesion which is especially handy if you're printing gears 😁.
Great video! I never really notice elephants foot in my prints, but maybe I'm just trimming it off along with the brim or something - I love my deburring tool for print cleanup and rarely do print in place things. I'll have to buy your nice clearance test one of these days and see just how good I've gotten it. I do feel like I do a better job setting bed height using a feeler gauge instead of paper, but maybe that's because I have practice using a feeler gauge elsewhere and have a sense of how it feels at the right spacing. (I think it also helps to have the next size smaller and larger to check against, helps me get more precise, though given I have a thick aluminum bed that loves to warp, it's all relative) Kind of surprising you see more than one layer of elephants foot (not sure the mechanism for how that would happen, is the first layer also ending up a little taller than expected due to squish?) but prusaslicer only compensates one layer. Seems like that would be an easy improvement to make, though of course it wouldn't help with things like the part of your headphone holder that you intentionally didn't chamfer to keep the contact area large enough. (Could probably add features to the slicer to restrict compensation to areas with a sufficiently large and wide contact area, but that's probably a lot harder to add.)
Could the multiple layers of elephants foot be caused by the part melting into the heated bed? It seems to be most visible on really long prints since these have the most weight pushing down for a long time.
As a person who does not own a 3d printer but occasionally watches videos because I think they're interesting, I'd have liked to have understood how "rafts" are removed from the workpiece. It was the very first thing I thought when you started talking about first layer problems; "Why not just make your model 2 layers higher and then remove them?" I was hoping you'd go through how those layers are removed without damaging the piece. Sanding? Razor knife? Some kind of milling machine?
The printer puts a slight gap between the two layers and in principle, they peel away cleanly. Often though, rafts can "weld" to the part with incorrect settings and yeah, then you're stuck with having to mechanically remove them with sanding or destructive methods.
the "paper" way is to set your nozzle to be JUST touching the paper, a single sheet, not folded or anything just a normal piece of paper. Basically adjust, till you feel the nozzle "tug" on the paper, then back it off lightly till you get no or VERY little tug. This fixes the foot issue, and doesn't put it too far away or too close, its a nice sweet spot for it.
1:40 First layer extrusion width is the Prusas secret weapon to eliminating elephants foot. That's .42mm vs I bet .6mm or .8mm regular extrusion width. That allows you to be a little too close to the bed for a great first layer stick and still actually end up with a tiny chamfer on the bottom instead of an elephant's foot!
I used to go through hell getting my first layer to stick properly. Once I started using Aqua Net hairspray I've said goodbye to my issues, it prints perfect every time and it's still easy to pop off the bed when it's done. I swear by this hairspray and it's also super cheap. Try this for an easy solution. Just spray it on while the bed is heating and it won't be all wet when you start printing. I've even had the first layer stick perfectly with the Z-axis a little bit too high. This works amazingly well and you should definitely give it a try if you're having issues getting your first layer to stick.
The Cura equivalent of this is called "Initial Layer Horizontal Expansion, and needs a negative value to mitigate the elephant foot. I use 0.04mm which seems to work well for me though I haven't tried other values to try and dial it in better. The only downside is that it is only for the first layer. However, with Cura being open source enough suggestions might have the devs add an additional value to gradually reduce the entered value over a set number of layers. That would be perfect. I should add that there is another setting, "Hole horizontal Expansion" that will increase the size of holes by the set amount, though this doesn't seem to be limited to a single first layer.
I printed a model of an elephant's foot and enabled "elephant foot compensation". My printer disappeared and now I have a mini black hole in my print room. Thanks, Angus.......
Hate when that happens
I print the elephant foot now my house is full of nuclear radiation and melt my printer :/
😅
I’m shocked at how much I can relate.
On the upside, a quantum singularity is an easy pet to maintain, and it'll eat anything. You have a great way to get rid of your leftovers and trash, and if you can train it to spin, you can generate power for the house.
Go to any automotive shop and get an inexpensive set of "feeler gauges". These are a bundle of metal slivers of calibrated thicknesses used for setting spark plug gaps. Choose the sliver that corresponds to your bed offset and slide it under the nozzle until it barely fits. Attempt same with the next size sliver down, if it will not go, then your know your exact nozzle offset is somewhere between the two. Much more accurate and repeatable than the "paper pull" method and only costs ~$10.
Dude! I never would have connected the dots.. Thanks for this suggestion,,,I also do car repair so I have these too,,,genius
I preach the Feeler Gospel daily! It makes such a huge difference.
Feeler gages are awesome. Not only does it tell you exactly what the gap is, it will also not change thickness or get damage as you use it.
My 3D printer came with feeler gages so I was sort of shocked when he said people use paper to gage the height for the nozzle. Like people who sell these expensive machines don't give enough of an F to give you the tools to use their product correctly? Why?
Right on, these are old old old machinist tips that only the oldest and probably deadest old timers used to use before numeric control.
Get a book from a guy named dave gingery (dead) titled "build your own metal working shop from scrap" it's filled with hundreds of tips like this to create precision machine tools without precision measuring tools.
I just got my 1st 3d printer and so much of the advice online is watching people struggle to find these ancient answers
One thing to mention: if you slice and print a 20mm cube, no matter how well calibrated your printer is it will *always* measure slightly less than 20mm in the end unless you compensate for material shrinkage by resizing the part slightly.
In Cura it's called Initial Layer Horizontal Expansion. Works like a charm.
Rafts produce an ugly finish on the bottom layer.
thank you for directing me there.
And I find rafts impossible to remove from PETG prints.
@@dragonthc yes they are in PETG. I need to use electrical tools and sand it loosing accuracy gained with it. I hate rafts lol
So what would you set the "Initial Layer Horizontal Expansion" at? Specifically
@@jazzmeltsface -0.1 (YMMV) to slightly shrink the size of the first layer and eliminate the elephant foot.
You’re so engaging that I somehow watched all of this when I don’t have any interests in 3D printing.. it was quite interesting, I gotta say
Exactly, if you home all the axis and level the bed with apiece of paper then you will have a .1/.2 mm gap at z 0 position so when the first layer is printing it will lift up another .2mm depending on the layer height resulting in a .4mm gap. So when i level my bed i lift up the z axis .1 mm and level there with a .1mm piece of paper resulting in a thiny little gap between the noozle and the bed at 0 position because the paper has a little room to play. I was looking for this technique online but didn't found anybody who used it and i didn't know if it was right but it works for me!
I used to do this too before I got an ABL system. I kind of still do it, but now it's done in firmware and for a different reason. In Marlin I have LEVEL_CORNERS_HEIGHT set to 0.4 and I use a 0.4 mm feeler gauge. Heat the bed, level the corners with the feeler gauge, run a G29 and that guarantees my leveling springs stay at the same level of compression essentially forever (having it set to 0 would compress the springs by 0.4 mm every time I ran a corner leveling sequence, having to at some point loosen the screws when one of them eventually bottoms out).
Elephants foot compensation is “horizontal expansion” in Cura. I’ve used it in a pinch to give models a bit more play or to tighten up gaps for a better fit. Super helpful!!!
Thanks a lot (:
Dude your like the best guy on youtube who does not focus on just one printer model. I like how you make most your videos more general so those with a different printers can follow along anyways. Thanks angus! 😄😀😃
comment comment engagement. comment for the algorithm.
Miss your face, glad you're keeping up on projects man.
Thanks man ! Getting back into it
Comments for the comment god, and likes for the like throne!
Good video. Have been printing for years and very surprised recently on the “lack” of videos to help with this. Thanks again for everything you do!
I think it's one of those things people learn to live with because it means your first layer adhesion is really good. Most people don't need their parts to be dimensionally accurate to where it matters.
Ex Queue yeah totally agree! My 3D printers create my living currently, so in the past I’d just take a fine rotary tool and take the flair off for closer to accurate dimensions, but still not perfect. Eliminating it all together is the best thing to do but requires a decent amount of tuning. On both design and hardware.
@@exqueue3813 Because most people are printing useless landfill material.
...not to mention that the "Elephants Foot" is the most dangerous object on planet earth. Thanks Chernobyl!!
Help, my printer gives me radiation poisoning and is eating it's way through my floor
@@chrissieblossom if it does just put it in rice smh
Ya thats what I was thinking
you're welcome
Sorry but you're wrong.
My Mother in law is the most dangerous object on Earth.
Thanks for this, it validated something I've been doing for months. I needed to print some structural pieces on a resin printer without elephant foot and figured a 0.5mm chamfer would compensate for the 0.5mm elephant foot I've been getting. The nice thing is with resing elephant foot only occurs on the bottom layers that have been set to a longer exposure, so it's very easy to figure out how high to make the chamfer and the part is still very dimensionally accurate.
Found this channel earlier today. As a current biomedical engineering major, accurate and precise 3d modeling and printing is a necessary skill to master. Thanks for your outstanding videos, subbed for sure👍
Hello Angus, this issue has been killing me this week with a filament that I put a lot of trust in. I've only been at modeling and printing for about 2 months now, and your videos have been hugely helpful. This one takes the cake though. Thank you so much!!!
delete his third letter of name
It's time to talk about the elephant's foot in the room
Lol
😐
i read that in CHEP's voice
@@jimster1111 what
If the elephants foot is in your room you are probably dead
Google it
Thank you very much for all the time you spend teaching everyone all these tips and tricks to get by on our printers, it's much appreciated! Please, don't ever stop!
I love how the elephant foot compensation and whatever makes support material easily detachable are both things that are activated by default in the Prusa Slicer. I just switched over from using Cura slicer with a Creality CR10 Mini to Prusa Slicer with a Prusa Mk3S, and I'm very happy with these two features, in addition to the magnetic PEI bed and the generally excellent functionality of the Prusa Mk3S printer.
Thanks so much for talking about rafts. I've always thought they seem super useful but haven't seen many youtubers using them often so I was a little nervous. but this makes me feel better about them!
I don't get why people hate rafts, for some prints they're necessary, especially things like rings where there's not much contact with the bed.
I find even a 1 layer raft can make a huge difference, and it really doesn't take that much longer, or use that much filament. I probably waste more changing filaments to make sure the colour has changed!
For me, I don't like them because the surface finish of the bottom of the model is not as good. But I do use them if I have to.
A brim would be far more effective with your example
@@havenview It totally depends, one of my printers absolutely hates printing circles smaller than 20mm on the first layer (probably down to the fake easy peelzy, but I'm too cheap to replace it because it works fine with everything else!) so a brim really doesn't help. My other printer on the other hand can get away with it just fine.
I've started embracing rafts more than every in the last couple of months for lightsaber parts
If rafts didn't have a use, they'd not be in the slicer settings :) and people who argue that you don't need rafts simply haven't had the situation where they're needed or they've found an alternative solution... simple as that. They are wasteful of material and they do add some print time but they've really good in certain situations.
I already use most of these things to stop elephants foot on my prints. What i suggest is putting a little bit of glue from a glue stick on the bed before the print and making the first laver height the same of higher than all the others. Really good tutorial. You did a really good job at explaining the fine details for people that are new to 3d printing. A++
This tutorial couldn't have come at a better time :V
I had a whole plethora of issues that I only had myself to blame for - I'm a rookie and made the silly decision of doing my first ever nozzle and bed upgrades (despite the fact that my printer was dialled in to near perfection with my old nozzle & bed), right in the middle of a bunch of commissions for some pals. Looks like I was way too confident in my ability to troubleshoot because it was one problem after another, (bed levelling and the nozzle wasn't fitted properly, ontop of speed settings etc.) I kept my head on and solved everything pretty quick until this damn elephants foot which I've been at for days with little improvement.
I was just about ready to call it quits and refund my friends till you uploaded this and now I'm happily on my way to finishing each commission. I think this calls for some more celebratory upgrades. :^)
All your videos are amazingly useful even if we know about what you are talking about already but the fact you emphasize it make the information stuck and now I am finally convinced to use Rafts for the parts that I intend them to be precise.
Tremendous video, Angus. So much useful information. I get these problems and until a few days ago I didn’t know what a raft was!
And in Cura it's called "Initial Layer Horizontal Expansion", but instead of a positive number, you have to enter a negative, i.e., -0.3mm or whatever. There's also the command "Horizontal Expansion" which is used to compensate for holes that print out too small (or large). Again, enter a negative number to compensate for holes that print too small and a positive for holes that are too big.
Both of these work pretty well for me.
Thanks!
Great advice as always Angus. Those new models look amazing, can't wait to see and print, always love your fiendish puzzles.
Would’ve been nice to see a comparison of a cube printed with and without a raft to see if accuracy is improved.
what a good video! Due to other issues with my printer, I've had to make some tweaks that drastically increase the elephant's foot issue for me. It's interesting that this video focuses not on "fixing" the issue, but more on how to compensate for it. Excellent video that just got me re-inspired to print again!
I tend to just eyeball my corner leveling. I find its suprisingly more accurite than basing the height on how well the nozzle grips the paper. Just use babystepping to grt the hight perfect while you print your skirt. Also when leveling corners make sure there isnt any plastic on the tip of the nozzle. Made that mistake plenty.
Great troubleshooting video!
You can also set your first layer height in the slicer to 0mm or whatever with a piece of paper or even pause the printer at its first layer/temperature to level it then.
Anyways, I just eyeball the level and either adjust screws or babystep/set home offset until things are level while printing. I assume a tolerance of .2-.3mm with a .4mm nozzle for projects so there is no need for a raft.
Another tip is lowering the bed temperature. I was printing PETG at 85 on the bed and just reducing the temperature to 80 reduced elephant foot almost completely.
I wish I would have had this info 3 years ago when I first started 3d printing! I knew nothing about 3d printing besides I was getting a computerized machine with a mini hot glue gun on it. The first 6 months or so of 3D printing was figuring out how to get the perfect first layer with no "squeeze out" was what I called it then. In the beginning I thought maybe my filament was too hot, or the heated bed was too hot and my prints were melting inside themselves!
I just have to say you have done a glorious job of explaining what elephants foot is and how to compensate for it.
Thank you thank you!
For very small parts I find chamfers especially helpful in keeping them dimensionally stable in the X and Y
Thank you so much for your videos I’m currently saving up for a 3D printer and I’m already picking up tricks and techniques thanks to your videos
Seen elephant's foot in title and was wondering how you could stop Chernobyl's elephant foot with a 3d printer.
Oh! That explains that. I'm very new to 3D printing, and have so far printed one thing (on a library printer) and it definitely had elephant's foot. I was wondering why it was a little wider at the base. Nice to know that my plan of adding a chamfer round the base edge is actually a recommended solution.
I would like to say that after 2 years of watching this channel I have finally caved and bought my first 3d printer. Ender 3 Pro.
Congrats, have fun printing!
Wait a minute, this isn't me
This was the video I needed. My prints come out pretty well and I'm happy with the dimensional accuracy, already better than some stuff I've seen at Uni. Except for the first 3-5 layers.. I was looking on how to fix this. Trying it tomorrow!
Next step: the remaining tiny bit of Z inconsistency and the infill structure slightly popping through
After about 5 test prints and too much wasted material I managed to get rid of most of the elephant's foot. So thank you!
You are the channel that inspired me to create my own youtube channel about 3D printing. Keep making great videos!!!
The real issue : There is no official tool for calibration.
I mean to be fair, why make a tool thatll probably cost like barest minimum $5 when you can get a pack of paper that has 500 or so sheets for less?
@@jasonking7570 or a feeler gauge
Freaking exactly
I was waiting for "Let's get started!"
After having issues, I bought your tolerance test. First print? .3 required force, the others broke. Second? They spinnin'! ALL of them! I was having a major issue with the infill causing expansion on the perimeters/walls, and after reading who knows how many forums (Prusa, reddit, etc) it seems that more walls would help. Sure enough, if you're having issues like that: up the wall count!
So, it seems there's something unique about that middle one...
Root causes of Elephant's foot, extruded width is wider than you spec'd due to wrong height, perimeter printed last gets nudged out by too wide fill, first layer stays too hot gets crushed. Fix these and get a perfect first layer: 1. Robust tuning to handle bed variation. First layer needs to be the thickest layer so changes in bed level are a small percentage of the first layer height. A 0.1 mm variation in bed level is 50% of a 0.2 mm layer, and results will always be crap. Use a thick first layer like 0.32 and pick a layer that is divisible by 4 full step motor steps for most Z accuracy. 2. Print first layer, stop, peel up and measure on all sides with micrometer. Adjust nozzle height and bed level manually until what you peel up and measure is as close as possible to the first layer specified in slicer. Measure near each level screw if possible. 3. Leave room for extra material due to variation. Set first layer extrusion width to the widest to get great adhesion, like 120%. This both increases flow rate and spaces the traces farther apart. Then set first layer flow rate to 90%. This will allow for variation in bed level before the print gets too big. 4. Print perimeter first on first layer. This avoids overly thick fill traces from bumping out the perimeter. 5. Slow down enough so first layer is cool before second layer goes down. Second layer with the fan blasting, to avoid crushing and expanding the first layer. Those setting at 1:30 are the opposite of what you want. Use First layer height 0.32 and First layer width 0.48, first layer flow 90 %, adjust as needed. Enjoy the perfect first layer.
Ultimaker Cura has a tick box for printing rafts. I don't recognize your slicer. Cura is pretty good. Besides, I'm new.
Okay, not quite as new after the previous line has aged a while, but I just want to let you know that I paid for the Clearance and Tolerance 3D Printer Gauge to test my profiles. I sliced it using .08, 0.10, & 0.12 layer height, and as expected, when printed using the .12 layer height, both 0.2 and .15 are stuck tight, and I had to break .3 free.
But I surprised myself when I sliced it with the 0.10 layer height, then printed it. Imagine how awesome I feel that on an Ender 3, I have it tuned well enough that all 5 are free and rotate easily. I didn't have to break them apart either. And, I didn't know that the middle piece comes out when all 5 outer discs are lined up with the cutout over the center. I didn't bother slicing and printing the .08 version. Why go that low?
Thank you for creating that gauge, so noobs like me have a target to shoot for.
If anyone thinks a Creality Ender 3 can't print well enough to get all 5 discs to break free, without breaking free, they just don't know how to tune the printer correctly.
For y'all to understand what I did, I first got rid of the elephants foot by using first layer horizontal expansion limits, then a couple other similar settings. I'm not going to go into any more detail. I'm willing to share my profiles, but if you really want to be good, you need to struggle to figure things out on your own first. Oh, look for hints in online sources, but remember, how you set up your printer in the first place will make a huge difference. The Ender 3 is a kit, so treat it like a challenge to get everything perfect, e.g., perfect 90 degree angles where they belong. Once you have that done, then tune it for printing.
And welcome to a new drug. A very addictive drug. But, the highs mean you are serving someone in your life by creating things to make their lives easier. Or more enjoyable. I make stuff my grandkids love. Yeah, I'm an old FART (Fathers Against Raising Teenagers - an odd club I thought I as in when my kids weren't married and all grown up and moved out.)
Nice solution! To fix it in S3D, I just add another process just for the first layer and give it negative Horizontal Size Compensation
This was the biggest problem I've had to deal with lately. Thank you so much for putting this out there! Found the "Initial Layer Horizontal Expansion" setting in Cura and I'm going to try it tomorrow :)
Wow, I have been fighting with this for a week. Im really new to 3D printing and its been a bit of a nightmare to get it all going. CR10 but the glass was apparently warped out the gate. I bought 12x12 panes from lowes and it took me all day to dial that in. I just ended up using rafts like you said as I still must be off a touch or Z is somehow or being told to increase z height at the start. Thanks for the video!
A few words about crazy and complicated 3D designs: procedural generation and scripting helps A LOT. F.e. - using openSCAD in conjunction with Python and SolidPython library. It made possible create 3D models literally impossible to hand-draw in regular CADs, yet going to need at least elementary programming skills and intent to dig into it.
Hey Angus, I’m a bit late for this one, if your puzzle is really difficult please send one to Chris Ramsey.
Great videos, very helpful, thank you.
I just started to try printing with Prusa. This is so helpful. Thank you!
Just at 11:30, I love rafts. I don't need them anymore, but on my first days working on a makerspace machine it wouldn't have come along without a raft. This bit of extra material is often worth compared to failed prints.
You've got the best points on design that are concisely and clearly presented. Thanks!
Our 2 stratasys printers default to using rafts otherwise they would fuse, the markforged prints directly to the bed but it always looks fantastic. My guess is a combination of accurate leveling and software correction. My printer at home uses foam blocks with HIPS so that's fun.
I fiddled with the settings and tightened my belts to try to get the perfect C cube and I got it bang on 20.00 mm in the X, Y but the Z-axis was 19.96 mm, I knew it was Elephants foot that was causing the loss in the Z-axis height.
I am now going to slice my cube with compensation to see if I can get that perfect C cube.
Elephant foot compensation probably won't help with the Z axis dimensions. It will print less plastic in the X and Y axis on the first layer, but the Z axis will still be squished. A raft might help though. Or you could increase the size of the cube in the z axis in your slicer by like 0.04mm
If I remember right Cura has an adjustable number of layers to apply elephant's foot compensation; I use it on almost anything that isn't printed on a raft. And yeah, when I really need that perfect accuracy, I do swear by using a raft. One of my prints required several copies of a part being printed on a hexagonal raft about 10mm wide, and since they pulled away so cleanly I made the resulting rafts into coasters for beverages! No wasted material, they look cool, and they actually work quite well!
Underside chamfer is GENIUS! Thanks!!!
cool video dude, just recieved my ender 3 v2, not had chance to build yet, so been watching as many tuition vids to get up to speed!! great info thankyou
Recently all of my prints have been a little off on the base, like it was squished, and i just couldnt figure out what was going on, thank you for this!
I haven't had elephant's foot on my prints ever since i got and started using Simplify3D on my paper leveled CR-10. It's nearly all in the slicer as far as i'm concerned.
My paper method consists of me pushing the bed down out of the way of the head by .3 and starting a random print that puts the start in the center. I then stop the print (but keep the heat on).
Using the paper i raise the bed on all four sides evenly until i can do the slip fit (on mine, because the fan motor got hit at some point, there's a bit of harmonics, not enough to hurt prints, but i can instantly feel when the tip is transmitting vibrations through the paper on the bed (and my hand by extension).
Once that's accomplished, i do a half way square up, recent to make sure i didn't lose the center height, and finally do a (on bolts) square up. Obviously all this is done with a fully covered glass bed (classic paper tape, flattened down with a spatula).
The paper i use is just a on old business card i have tons of, i think it's twice the thickness of regular paper. If i observe any issues, i raise the starting of the print via Simplify3D.
Now, why do i say that it's all in the slicer? Because before, when i used Cura, i had inconsistency after inconsistency. Not enough to make me level the bed a lot, but enough for me to note it. With S3D i just start the thing, make sure i don't have any drip, and leave it be. I level the bed only once every 60 prints or so, because i like to clean the tracks and wheels and lubricate the bearings.
Great vid.
Superslicer has a setting for elephants foot and also has another setting which lets you select how many layers to compensate for.
To mitigate the gap created by paper thickness when touching off, try using something other than paper, which is .003" in thickness. I suggest getting some steel shim stock, available in .001" thick so it takes 2/3 of the problem away. Another choice might be the cellophane wrapper on many cigarette boxes, if you know someone who smokes. It is most often .001" also, but a bit more flimsy and tears easily. I have used shim stock to touch off tools on a cnc mill.
*Doesn't like to use support material and has an entire video on how to avoid needing them*
*Solution to elephant's foot: use a raft*
All jokes aside, thanks for yet another useful video, my prints would suck without your help!
I use normal zikzak support instead of a raft. Usually I print a model 3mm in to the air. This works very well. Especially with ABS to protect against warping. Enable „brim for support“ to get more bed adhesion. Using support means less material, far less print time and basically the same result.
Thank you, Angus. I'm dealing with this now on a complicated device with low tolerances. I'm sick of using a stanley knife to cut off the elepheet!
I dunno if it's just me, but I haven't suffered this problem at all in the past year or so. Slicers have been coming with very good settings to avoid it, and model quality has improved a lot since people are learning better CAD techniques. I suspect your newer videos have something to do with that trend :D
Yeah now that you mention it I haven't had any issues with the latest batch of machines. Mesh bed levelling and print surfaces mean we don't have to drive the first layer super close anymore as well, which is nice 😅
I'm a newbie and your video took me out of 5 failures on my print, thank you so much. i was like on a time loop. :p
Great to hear!
I printed you clearance castle on my new FLsun SuperRacer at 150mm and no problems everything was prefect!
I use rafts all the time. I have found that having rafts help with my prints staying attached on my bed
What about reducing the flow of material for the first layer. I have it dialed down to 50% and works really good.
this also works. but not, when you have elephant foot affecting first 4 layers, like in my case. None of everything, I have tried, helped....
dont forget, you can also mitigate some with a well adjusted and leveled bed AND proper z axis comp. (provided that bed isnt warped) PLUS having a bl touch can help for beds that are not prefect.
Concerning the "Elephant Foot": in Cura, look for "Initial Layer Horizontal Expansion". Putting a slight negative value there will do the trick. This would be a way more elegant solution than modifying the design of the model...
Then if you have issues with tolerances and dimension accuracy, have a look on how to calibrate the "Horizontal Expansion" value. This value differs a lot between different materials and a little between different settings for the same material. (I get values like -0.07mm for PETG on an Ender-5)
I use the following technique: use a feeler gauge, not paper, to set the bed height to the required "first layer height", then configure the slicer to pint the first layer at 0 mm. this seems to work fine
Thanks for this video. I have a persistent EF problem on my Ender 3 but all the other videos I have seen only talk about the EF being on the first layer. Mine is visible up to as many as 10 layers and all protruding outwards, not concave like some you showed. I have tried many adjustments but I will now revisit the problem.
I use raft every time I print. I don't know why but when I don't use raft, my first layer always becomes a disaster. I'm using Creality CR-20 Pro.
Dude it happens to me on my ender 3 pro as well. I'm thinking the stock bed surface is garbage, but when I'm leveling the bed, I make it that the gap is slightly tighter than a sheet of paper and that helps.
Use a proper set of feeler gauges and set the Z to the size of the gauge you choose to use before starting to level.
I find a 0.4mm gauge has enough ridgidity but still allows enough flex to get under the nozzle assembly.
Feeler gauges also allow you to level when the bed and nozzle are hot which also increases accuracy.
Me: doesn’t even have a 3D printer.
Also me watching this video: Hmm yes, interesting...
Y=same
Col_Weab976 YT you should get one. They are a lot of fun to make things with. My grandson asked me what I was going to do with one. I told I didn't know. But since then I have made quite a few things that wood or steel was just not a good medium to use. Plus now you have a good excuse to learn a 3d CAD software to keep your brain from turning to mush (it may fry it though). At 66 I am still learning and having a blast making things.
@@krisknowlton5935 you know what? This. This comment is what’s pushing me over the edge. Imma get one.
@@thesteelcolonel I bought the Ender 3 Pro. It seems to be a good machine. Mine has made excellent prints.
@@krisknowlton5935 how much does it cost?
There are two different elephants foot effects. One is the first layer squish, which you described. The other is where the layers above taper inwards and then come out again. I think it is because close to the bed the plastic takes time to solidify and is not fully hardened allowing the next layer to contract inwards. I mitigate that by dropping the bed temperature after the first layer. It also seems to be worse with low infil percentages, so I often use 95%.
@3:56 🤯 i always wonder why i have to adjust my z offset by -0.2mm everytime i calibrate my printer. Now i know!
If you're using auto bed leveling some systems will raise the measurement my 0.2mm so you don't scratch your surface.
I used tuning on my ender-3 pro last night to turn the z offset down 0.2• I was getting very frustrated at the bed
Very interesting and informative.
I think it is also worth mentioning the *temperature of the bed:* the squeeze of the first layer is possible because the bed heats plastic enoguh to be able to deform.
By reducing the temperature of the bed after the first layer from, say, 60 to 50 degrees Celsius (PrusaSlicer has a specific option to set this), also reduces the squeeze of the first layer.
I don't like using chamfers to correct the printed object, as it is not actually a part of the object, just a trick to overcome the squeeze of the first layer. But it works ^^', just as rafts do; sometimes you have to do what you have to do in order to have the printer make the object well
I always thought this happens due to too high bed temperature so that the bottom piece of the print softens too much and collapses under the weight of the printed object. Because to be honest, this effect remains for the first few layers, not only the bottom layer.
I started using Prusaslicer today and I'm so happy
Whoa, I was having this issue with a test print yesterday! Perfect timing!
As far as leveling the bed: When I use my printer's leveling function, the Z height is at 0.2mm, so all you have to do is adjust the bed to be 0.2mm (~0.008") away from the nozzle. If you fold standard A4 print paper so it has double the thickness, it's around 0.007", which gets you pretty close. You just want to make sure the machine is at the coordinates that it thinks it's at.
Hi Angus, found an issue with printing and figured you were just the guy to help
My printer, for whatever reason, prints any inner circle, such as a hole through a block, as a square-ish circle, like the cylinder was squashed on the sides.
Any idea what could be going wrong?
Many thanks
THANK YOU! This was a very well done, very useful video covering the entire spectrum of the elephant foot issue and giving valuable advice and great solution options on how to remedy the issue! 👍👍👍👍🙏🙏🙏👏👏👏👏
One aspect to consider is bed texture. My Artillery Sidewinder X1 v4 has about a 1mm waffle pattern in the glass bed. It was driving me crazy, and no amount of bed leveling, slicing, or model compensation was giving me perfect parts. One "trick" was to cover the bed with glue-stick, let it dry, then scrape off as much as I could using a very sharp wood chisel.
But what about when I wanted the bed texture to appear on the print? The brute-force method is just let the print do what it does, then do a quick pass with a deburring tool. Which works great where the first layer edges are straight or gently curving, but fails badly for tighter curves and sharp corners.
Oooh I see you’re working on making your cool spiral thing into a box
Excellent reminder of this and some great tips for mitigating it. Thanks for another great video!
Thank you so much Angus. It's been a big problem for me and now i think my parts might fit.
I dont have a 3d printer but its fun to see people fix the problem
I heard that you're supposed to make the first layer thicker not thinner, for adhesion.
No need to over smush the material if there enough being extruded. The excess material can also help it get into small crevices of a textured build plate.
Although looking at one of my prints it seems more like "elephants layer" thane "elephants foot". (the first layer kind of sharply sticks out somewhat)
I have a similar issue however I commonly print with a brim so it is not a big issue since I have to cut the brim off after print anyway.
"Almost unsolvable" sounds awesome! +1 for that
I used to have horrendous frustration with first layer adhesion problems, so much so that using the printer was as much a pain as it was a pleasure. Using the paper method for getting that optimal first layer thickness to allow good adhesion was an absolute nightmare, particularly as my bed is not the flattest. I was thinking of getting some automotive feeler guages to help with setting accuracy but I decided this was getting ridiculous and there simply had to be a better way, most likely involving some sort of glue. After some experimentation, I settled on a diluted solution of builders PVA glue and water, around 10:1 so it has the same consistency and colour as milk. I thinly wash the bed with this solution using a brush, leaving a translucent layer. By the time the bed has heated up its all dry, leaving a slightly shiny and vaguely tacky surface. You don't have to wash the whole bed, only the area that the print will make contact. Since using this method I have never had a single failure related to bed adhesion. Removing the print is easy too, it just cracks right off when the bed has cooled off, far easier than my experiences with other print beds. After a number of prints you may need to clean the bed of residual PVA that might have built up but this is easy as PVA is mildly soluble in water even after having dried and isopropyl alcohol just whips it off. Having a 100% repeatable solution to bed adhesion allows far more lee-way with your initial layer height and is amazingly forgivable with a non flat bed. I rarely use the paper adjustment method anymore, I use it once to get a rough setting, set the printer off and manually fine adjust once only when its printing a Brim layer. On subsequent prints I need make no adjustments, I don't even need a Brim for adhesion which is especially handy if you're printing gears 😁.
Great video!
I never really notice elephants foot in my prints, but maybe I'm just trimming it off along with the brim or something - I love my deburring tool for print cleanup and rarely do print in place things. I'll have to buy your nice clearance test one of these days and see just how good I've gotten it. I do feel like I do a better job setting bed height using a feeler gauge instead of paper, but maybe that's because I have practice using a feeler gauge elsewhere and have a sense of how it feels at the right spacing. (I think it also helps to have the next size smaller and larger to check against, helps me get more precise, though given I have a thick aluminum bed that loves to warp, it's all relative)
Kind of surprising you see more than one layer of elephants foot (not sure the mechanism for how that would happen, is the first layer also ending up a little taller than expected due to squish?) but prusaslicer only compensates one layer. Seems like that would be an easy improvement to make, though of course it wouldn't help with things like the part of your headphone holder that you intentionally didn't chamfer to keep the contact area large enough. (Could probably add features to the slicer to restrict compensation to areas with a sufficiently large and wide contact area, but that's probably a lot harder to add.)
Could the multiple layers of elephants foot be caused by the part melting into the heated bed? It seems to be most visible on really long prints since these have the most weight pushing down for a long time.
I don't even have a 3d printer. But enjoy your informative content. Thanks!
As a person who does not own a 3d printer but occasionally watches videos because I think they're interesting, I'd have liked to have understood how "rafts" are removed from the workpiece. It was the very first thing I thought when you started talking about first layer problems; "Why not just make your model 2 layers higher and then remove them?" I was hoping you'd go through how those layers are removed without damaging the piece. Sanding? Razor knife? Some kind of milling machine?
The printer puts a slight gap between the two layers and in principle, they peel away cleanly. Often though, rafts can "weld" to the part with incorrect settings and yeah, then you're stuck with having to mechanically remove them with sanding or destructive methods.
the "paper" way is to set your nozzle to be JUST touching the paper, a single sheet, not folded or anything just a normal piece of paper. Basically adjust, till you feel the nozzle "tug" on the paper, then back it off lightly till you get no or VERY little tug. This fixes the foot issue, and doesn't put it too far away or too close, its a nice sweet spot for it.
1:40 First layer extrusion width is the Prusas secret weapon to eliminating elephants foot. That's .42mm vs I bet .6mm or .8mm regular extrusion width. That allows you to be a little too close to the bed for a great first layer stick and still actually end up with a tiny chamfer on the bottom instead of an elephant's foot!
I used to go through hell getting my first layer to stick properly. Once I started using Aqua Net hairspray I've said goodbye to my issues, it prints perfect every time and it's still easy to pop off the bed when it's done. I swear by this hairspray and it's also super cheap. Try this for an easy solution. Just spray it on while the bed is heating and it won't be all wet when you start printing. I've even had the first layer stick perfectly with the Z-axis a little bit too high. This works amazingly well and you should definitely give it a try if you're having issues getting your first layer to stick.
The Cura equivalent of this is called "Initial Layer Horizontal Expansion, and needs a negative value to mitigate the elephant foot. I use 0.04mm which seems to work well for me though I haven't tried other values to try and dial it in better. The only downside is that it is only for the first layer. However, with Cura being open source enough suggestions might have the devs add an additional value to gradually reduce the entered value over a set number of layers. That would be perfect. I should add that there is another setting, "Hole horizontal Expansion" that will increase the size of holes by the set amount, though this doesn't seem to be limited to a single first layer.
Top notch advice as usual, Angus. Thanks again.