How much WEAKER 3D Prints become at Higher FLOW RATES / SPEEDS? (DATA from 1000+ samples)
HTML-код
- Опубликовано: 5 июл 2024
- Get CNC machining, PCB manufacturing, 3D printing, and other professional services for your projects from today's video sponsor - PCBWay (5$ OFF for your first order) - pcbway.com/g/584K9t
👉🏻 In this video I am exploring how much layer adhesion strength we are losing by 3D printing PLA at faster speeds and flow rates. I broke over 1000 samples for this video and the data is surprisingly interesting.
🖤 Support my work at PATREON - / diyperspective
❤️ Support my work at RUclips - www.youtube.com/@DIYPERSPECTI...
📋 RELATED ITEMS TO THE VIDEO (Affiliate):
3D printers:
- Bambu Lab X1C shrsl.com/480se
- Bambu Lab P1S shrsl.com/480sa
Filaments:
- Bambu Lab filaments shrsl.com/480r3
- Creality Hyper PLA shrsl.com/4cj3s
- AnkerMake PLA+ shrsl.com/4gesl
Testing equipment:
- T3 PRO IR Camera - US: amzn.to/469wcGA EU: www.amazon.de/dp/B0BNKH57Z6?f...
- Digital Calipers s.click.aliexpress.com/e/_Dlt...
- Digital Dial indicator s.click.aliexpress.com/e/_Dkw...
- Digital Micrometer s.click.aliexpress.com/e/_Deb...
📷 VIDEO SHOT ON:
- Fuji X-S10 amzn.to/3Zar9BW
- Raynox DCR-250 for close-up amzn.to/43h04AB
📢 OTHER MENTIONED THINGS:
Tensile Strength Testing Machine www.printables.com/model/8051...
🕗 TIMESTAMPS:
00:00 - Main topic
00:25 - Testing machine
01:49 - Testing methodology
02:24 - First results
02:54 - Matte PLA... LOL
03:11 - Final design
03.45 - Adhesion drop-off results
06:31 - Adhesion strength in MPa
07:07 - Conclusion from all the results
🔗 YOU CAN FOLLOW ME:
Twitter: / diyperspective
#3dprinting #3dprinter #layeradhesion 
Наука
From some comments, I see that people misunderstood what I was trying to say or I didn't explain well enough. So here is the TLDR:
- Printing speeds alone most likely almost have no impact on the adhesion.
- Higher printing speeds are directly related to a higher flow rate.
- Higher speed and flow rate means that the filament spends less time in the hotend.
- As plastics have significantly lower thermal conductivity than metals, it doesn't heat up enough and that is why we start losing adhesion.
But what if we increase in the hotend temp and how fast we can cooldown the PLA after it's been laid down?
Thanks for the clarification.
However printing faster is not the only way to increase flow rate.
What your data shows is the flow rate correlation to layer adhesion.
Sure, printing faster pushes the flow rate, but so does layer height and bigger nozzles.
It would be interesting if you could repeat this test using bigger layers (0.3?) as well as larger nozzles (0.6, 0.8?) to see if the layer strength is truly speed limited or purely a flow rate issue (and not speed related)
Thats what Bambu Lab does. More Heat than usual.@@greyvlad
@@Shadow27374 Marlin has flow rate dependent temperature correction too.
I'm glad you did something like this! Reinforces my experience too where anything up to around 60% of max flow rate of your printer is acceptable strength. Knowing you can go to around 80% and still have decent strength is good to know for prototyping. Great content as always.
Yeah, rapid prototyping with filaments that don't lose adhesion at higher flow rates and you can test parts without worrying that they will immediately break is really nice. :)
...print temperature is probably the most important part toxinvestigate with print speed, not material type.
Temp will make a world of difference, I've had total crap, brittle filament that I left in the Sunlu dryer at 60c while printing it at 225c , it was PLA.
That allowed me to get useful prints out of it and surprisingly still handled overhangs, I didn't attempt any bridges.
If it sucks, crank it till you burn it. What do u got to lose. Just don't clog your extruder by leaving it running while not extruding. Worst case, you lose some detail.
Very good work. Have you tried repeating a test with an increase in heater temperature along with the increase in speed? I routinely print at high speeds but with elevated temperatures to compensate for the reduced residence time in the heat zone. I have not seen a significant drop in strength and my PLA material has a gloss appearance not matte.
No, I did not. It depends a lot on the material itself. Some remain more shiny some quickly turn to a slightly matte finish. But maybe I will investigate that sometime.
@@PrintingPerspective I would also recommend increasing the print temperature along with the print speed. Basically repeat testing but all at 230 for example.
I second this, try a higher temperature to see if it make layer adhesion better at high speeds
If you're printing fast enough that the material doesn't sit in the hotend for minutes at a time cooking, there's no reason to use temperatures below 250. 250-260 is my daily driver PLA temperature on my fast printer; 270-280 on my Ender with short melt zone.
Would these results be affected by the extruder? Layman here, but I would imagine that longer melt zones, more thoroughly melted filament and evenly consistent performance would affect layer adhesion. Do they? Do CHT type nozzles change results? This is very interesting, and thank you for making such a clear video about the topic!
I feel like weighing the print could shed substantial light on what i think is the reason for the shape of his graph
@astrumrocket6556 But longer melt zone hotends have a max flow rate limit too, so most likely it will still be half of the max flow rate. But until anyone tests it in practice we will never truly know, same with the CHT nozzles. :)
@@PrintingPerspectiveMy line of reasoning was that weak layer adhesion happens because the plastic didn't have time to melt properly so it doesn't bond as well. Longer melt zone would mean more time for plastic to melt. No idea if that would actually affect anything.
@@PrintingPerspective please test the CHT style nozzles! (especially the Chinese CHT clones which are actually better) They are such a massive easy upgrade for BBL printers especially, I would love some real world data on it for high strength parts.
These graphs look similar to the maximum flow graphs from Stefan of CNC Kitchen. It might be interesting to also measure the thickness of the samples to get a reading on the flow rate.
Yeah, that is the preset I like the most in Excel :D I was measuring the extrusion width in my previous video with a micrometer. But that is a great idea if I will be testing filaments I could determine how they flow extruded by the width and maybe even find if there is any correlation to the adhesion strength.
After working with PPS I believe you are right. I believe at that higher speeds the extruder can’t properly melt the plastic evenly and throughly enough. As I used a custom high flow hotend I was able to print faster and at a lower temp with a high flow nozzle.
Good point. I think a lot of the drop off may be attributable to reduced perimeter width at very high flow rates. Less cross sectional area = lower strength measurement. If we can control for width, it would be interesting to see if you would still see a drop off in strength.
Great video. I think it would be interesting if you could show images of how the samples looked at the different colours and speeds. If there was some visual correlation (shine, layer quality, etc) with the strength of the parts, it would be inmensily useful for other people, as factors such as hotend, extruder, cooling will be 3d printer dependant.
I'm thinking to get the same layer adhesion at higher speeds you need to increase the heat at a proportional rate, what that rate is will have to be determined.
My guess is the faster speed is allowing the filament to cool faster and not get the same bond.
Good job as usual. :)
Thx for doing all this work and sharing your results with us!
Very interesting!
This is a VERY IMPORTANT investigation. Thanks!
Great video, thanks for your effort in this subject!
Fantastic video! I'm also into material testing and currently building my own motorized tensile tester. Your video motivated me keep working on mine because this is such an interesting topic!
Thanks! Yeah, it is a way deeper rabbit hole than I expected, and I expected it to be quite deep! :D
would love to see the same test on ABS /ASA and other materials
Your videos are always interesting.
I think extrusion temperature has a big role too but it is already very useful to know these data so thanks for the effort 👍
Extruded plastic temperature is the main thing, but it naturally heats up less because it spends less time in the hotend and plastics have low thermal conductivity. That's the unavoidable reality.
@@PrintingPerspectiveso.. bump up the temp?
Great video ! I would love to see a comparison between different types of auxiliary fans, it seems that the cooling is a key factor aswell
Would be interesting to see with a higher flow longer meltzone hotend. I know one that would be interesting to see. 😊 Very nicely made video mate! Thanks for sharing
Yeah, there's so many things I would love to test, but there's always not enough time ;D Thanks
@@PrintingPerspective oh man ...ikr! I say that because with a higher flow hotend /longer meltzone, it tends to overcome that issue and bonding is strong even at high speed :)
Keep the research coming, you're a godsend for the 3d printing community!
I do wonder if staying further below the max flow rate of a hotend helps since it heats the filament more thoroughly. If possible i'd like to see this test with a super high flow hotend like the goliath (with cht nozzle). Or even test melt zone length.
Great video, whats next on the to test list?
Great job. Would like to see something like that for petg. I got some great results with high temps above 150mm/s. It seems many are giving up on petg these days but I think there is a lot of potential.
Would be great if you could test:
- high flow-rate hotends vs normal hotends
- normal vs CHT vs hardened nozzles
- "high speed" PLA, like Polysonic from Polymaker. They specifically say that it retains layer adhesion at high flow rates.
Excellent test, thanks for doing it. I have been using Elegoo Rapid PLA+ and their Rapid PETG as well with great results on my Qidi Xmax3. The PLA+ is noticeably stronger than the Hyper PLA and it is quite a bit cheaper as well. I use the PETG for functional parts and the PLA+ for everything else.
Very cool. I always knew that feed rate and temperature depends on color but man... That greeeeen :D
What about rising temp when speeding up?
It's gotta cool a bit before the next layer. It's a balance.
Thanks for sharing your test setup and the thorough analysis. What are your thoughts on the main cause of the drop in adhesion when raising movement speed? From what I understand, it could be (1) lower temperature of the extruded material caused by the higher flow rate, (2) actual underextrusion from an overburdened extruder/hotend, or rather (3) less precise movement of the print head at high speeds.
I think mainly because plastics has low thermal conductivity and the less time it spend in the hotend the lower temperature it will reach when extruded and won't properly bond to the previous layer.
Hello, I am very appreciative of your number of samples. I think it makes your results much more representative.
I was wondering what program you used to make the graphs. They look very nice.
Thanks, that's Excel, but I also do more photo editing after to add other required info.
Good tests. Thx.
6:38 In the future it would be nice when graphing these sorts of tests if you could include the range of the results instead of just the average. even if you are only losing X% strength on average it would still be useful to compare consistency.
I was thinking how to represent results the best as graphs get cluttered very quickly. But because it's a print not in a vase mode, there were a tendency for the results to fluctuate in a similar pattern at every flow rate. But you are not wrong, that would be good info to add.
@@PrintingPerspective Thank you for the consideration. IDK what program you use for your graphs, but it probably has an option to add error bars to each point (the little I shapes). I just thought of this because CNCKitchen does it for his bar graphs, and its very useful to see when a specific sample is more or less consistent (or if they are all the same). I dont think it would clutter your graphs too much, but they wouldn't overlay so cleanly with them on. You could show each filament sample on the graph by itself with the error bar, and just keep the line when comparing multiple filaments.
This is a great video, and very helpful. So I figured it was only fair to try and give some (hopefully) helpful feedback. Looking forward to more like this!
The probe temperature and the actual temperature will be different values because the probe temperature is updated every X second/s (or moves).
So you’re probably just measuring layer adhesion versus temperature. This is why layer adhesion doesn’t change below a certain speed - which is slow enough for the temperature probe and heater to react.
That theory could definitely be tested by printing at higher speeds and higher temperatures
Great info. Now you need to test different kind of printers, different hotends or direct vs bowden.
Would be nice to see another test, where you take samples at 250m/s where it's degraded, and test how higher temps or changing the volumetric extrusion rate, can affect the resulting strength. The purpose is to test all combinations, and to exclude or determine their effect on strength, so that you can isolate exactly how much just the speed itself is affecting the strength, and not just the lack of temperature or non-calibrated extrusion settings.
Found your channels a few days ago. Already binge watched 90% of of your video. Great contents!
I have built a tensile testing machine before that record stress-strain curve data into a computer. I am actually interested to know your electronics design (I guess many others do too). Especially the "control tower" thing with touch screen and the data meter. Much appreciated if you can add that into your printable link / a write up/ or even a video.
Thanks, I am glad you like them.
I do not record the curve, only the the ultimate tensile strength. Parts that I used was:
- 500kg load cell www.aliexpress.com/item/32866914213.html
- Load cell data display/meter www.aliexpress.com/item/32766229850.html
- Touch screen and motherboard are just parts from old printer that control the movement.
It is pretty basic setup that I feel is more than enough for me. The remixed CAD design - www.printables.com/model/805135-universal-tensile-testing-machine-v2-remix
Thank you for testing this! When I first got my bambu lab printer, my PLA prints had a matte finish and I was concerned that the strength might be reduced. I did some very basic tests of my own and found no real difference, but it's nice to get some science on it!
It also concerns me that most of the filaments you tested had significant drops after 200-250mm/s, and my default 0.2mm layer profile exceeds this for everything except the outer perimeter and top surface...
If you use Orca or Bambu slicer there's flow rate limit set in the material section so you are never reaching those speeds. But default Bambu hotend's flow rate should be even lower than Volcano that I tested on, so you might need to lower speeds even more for strength.
@@PrintingPerspective that makes sense! It's not obvious what the actual print speed is, but nice to know that it's just an upper limit
i think as long as you keep your minimum layer time high enough, you are good, gives time for everything to cool down correctly
This was super informative. I'd love to see a test with the swinging hammer. PETG seems a little more brittle than normal when I print fast.
PETG is a different beast from my experience. It varies heavily from brand to brand. Some brands PETG needs to be printed at very low flow rates.
Great data!
This is interesting but I wonder how a cht nozzle affects it because one reason for poor layer adhesion is the platic isnt being melted completely, but it can still be pushed through so it makes poor layer adhesion. The Cht nozzle midigates this.
Recommend only using volumetric flow rates. Cannot go by linear speed since there are various layer heights and line widths.
Predictable results.
You have to increase hotend temperature for high flow/speed printing. Filament is simply cold as it going to fast through hotend. Test 240-250C.
So PLA with Volcano 12 mm^3/s outer, 15 mm^3/s inner perimeters and 18 mm^3/s infill. Simple enough and works for most filaments, apparently.
Is there any reason you couldn't attach a sort of heat gun to the tool head that would reheat the previous layer right before laying down the next? Might be a good way to increase adhesion, and maybe a good video idea to find out
It is possible with high chamber temperature and high fan pressure
I don't think that helps for PLA, CNC Kitchen have a video about it.
This sounds like a gratuitously complex way to avoid doing the right thing: cranking up the nozzle temperature. For most materials there's no reason you can't go 50-90 degrees over the recommended temperature as long as you keep moving and don't let it sit and cook.
what if there's a way better hotend, for example VzBot Goliath? I can do over 40mm3/s with my modded hotend, but I feel like 30 is safe maximum. however, calculated into speed it would be 375mm/s at 0.4 width and 0.2 height. I do not have Goliath, but that can do over 1000mm/s for sure since it flows over 100mm3/s.
Probably if your calculation is correct. But there are so many variables that it is impossible to say for sure. That's what I learned from doing more and more testing.
man would love to see a similar testing methodology applied to filled nylon filaments. When I need strength + durability + abrasion and temp resistance, I use filled nylons
Great videos as always
Thanks!
How do CHT nozzles change the picture? they increase flow rate at a given under-extrusion % by 30% (as tested in various other channels), but do they also expand the speed range where the layer adhesion stays as good as at slow speed?
I know you took a huge number of samples, but for this new test a smaller number of samples, and only one printer, would be sufficient.
I just noticed that on AliE they sell nozzles "plated copper" or "brass CHT" (copper CHT insert on a brass nozzle). I wonder which one will perform the best (ignoring behaviour with abrasives).
I'm at the beginning, but I've always thought that there has to be some sort of trade off for the speed machines
I regularly print at 500mm/s and never had noticed poor adhesion. I think if you up your temps while printing fast it will give you a different result. I get you are trying to set a control, but it's a variable that also moves with print speed. Run 250mm/s with 230-240c.
I like the idea, but I feel like this is missing a lot of information. It would be great to relate this to the flow rate graph, from CNC Kitchen, and test it on different printers, to see how it is also impacted by the flow rate of the hot end
Flow rate should be measured too, you might be getting underextrusion there which is being the reason of strength dropping off.
I measure max flow rate, then deduct 20% for safety margin, and print PLA at 240C -- have yet to notice any weakness in the parts.
On Sovol's volcano style on SV07Plus this yielded for PLA 24mm3/s and on Elegoo Neptune 4 Max 21mm3/s
Creality K1 Max i recall defaulting for ABS to 23mm3/s, i should measure what it can do as well.
at 7:35 your max flow rate has been shown, so the drop off starts to happen at about -20% of the max flow rate. There's also measurement confounding factor.
I use the curvy print from orcaslicer to measure and calipers at 0.5mm3/s per 1mm and just pick the first layer of issue, then remove 20% for the final.
So it might be speed doesn't matter, only the flow and cooling rate.
forgot to mention, i use flow rate maximum as the limiter and print typically at 0.2-0.25mm layer heights, so actual movement tends to be quite low with plenty of plastic, and curtain cooling also being used.
So end result is mostly cold prior layer, with a lot of plastic on top at lower speeds of 120-150mm/s, giving it more time to melt the prior layer.
Is volumetric flow taken into account?
There is a problem with the experiment.
The adhesion depends on the thickness of the lower substrate, the wider it is, the worse the adhesion will be due to heat loss in the base.
A line of filament printed on another will melt without problem, printed on 2 it will no longer have as much grip and in the middle of 3 it will reduce it even more but the sum of the whole will be stronger. That is why the vase mode usually achieves adhesion in which the piece breaks by jumping between the layer lines and the reason for turning off the fan the first 3 layers
The cause could simply be that you are pushing the hot end closer to its flow rate limit.
Your graphs look almost *exactly* like underextrusion falloff graphs. If so, you don't really have weaker adhesion at higher temperatures, just weaker adhesion at narrower wall thickness (i.e. seeing result of a correlation not the variable in question).
Wish you had tested a black filament
Black color Hyper PLA had 37.3 MPa layer adhesion strength at 50 mm/s printing speed while natural white Hyper PLA had 35.7 MPa. I have no data for higher flow rates with black color Hyper PLA unfortunately.
Good test but different colors even from the same brand will have different strength properties.
You should really be testing this while ensuring the filament is maintaining the same temperature at different speeds and not the hot-end. The faster you go, the less time the filament spends in the meltzone, and thus it's actual extruded temp is not likely what you've set for your hot end.
This really renders a lot of the data here not applicable as a result.
I test realistic printing scenarios not imaginary ones.
@@PrintingPerspective ey man I'm not tryna come across as rude, I'm just saying that the filaments temp is what's important to consider here, if it's not heating up enough it's not gonna flow well no matter what. this doesn't even account for improper die swell at higher speeds either.
Man, I just left the printer at 200% speed and 135% FR on a 36h job, although I haven't watched this video, I'm going to be huffing copium like mad, I can tell...
very interesting ty
Glad you enjoyed it!
and then there are materials like ABS where it is better to print fast for better layer adhesion and less warping, all a bit complicated
I feel like you might be hitting extrusion limit of your extruder instead of print speed affecting strength.
I updated the title because I got a few similar comments like yours. Higher speeds = higher flow rates, they are directly related. I even explained my testing methodology and all charts show flow rates. I doubt layer adhesion is affected in any meaningful way if we isolate and look only into the speed. Filament has low thermal conductivity and when it spends less time in the hotend it is heated less and we are starting to lose layer adhesion. Maybe it is just obvious to me, it is so easy to forget that, and I need to explain more in the video. I will try to remember to do that in the future. :)
*me still printing 35 mm/s at .08* guess I'm safe
Man really struggling to make out what you're saying at times. I do think that your testing samples arnt typical of actual parts though
WOW! that's all I can say about the time and effort put into this setup, video and testing. I really hope this video pays off. I hate subbing, but I subbed just to help support, and gave a thumbs up.
Magnificent 🤌🏽