Це відео не доступне.
Перепрошуємо.
3d printing multiple objects at once vs single object 3D printing (one-by-one) - same strength?
Вставка
- Опубліковано 7 сер 2024
- I was curious, if there is any difference in strength (layer adhesion) between parts printed one at the time and those printed with other parts at the same time. Advantage of multi part 3D printing is that we have start printing only once, and printer will do the rest. There are 3 disadvantages to this method: risk, same color, stringing.
My checklist and risky factors mentioned in the video (edited according to suggestions in comment):
- small contact area
- too big overhang
- too big bridging
- tall, thin objects
- are supports added where needed
- check printing parameters
If you like my work, you could support me:
/ mytechfun
www.buymeacoffee.com/mytechfun
www.mytechfun.com/donation (PayPal link)
Test object used in the video can be downloaded from:
www.mytechfun.com/video/125
Contents:
0:00 Introduction
0:27 The theory
2:42 About layer adhesion
3:22 Preparing test specimens
4:32 Layer adhesion test
5:48 Results
5:59 Conclusions
#3dprinting
OK, closer to solving the mistery (thx to comments), I have to figure out how to check this. Possible reasons:
- with multiple 3D printing extrusion stops (even retract 0.8 mm back to nozzle) and filament melts more in nozzle during the travel. And it stick to next layer better
- in prusa slicer "Slow down if layer printing time is below: 15 seconds", nut sure how can this make layer weaker, but this also may resulted the weaker adhesion with single printing.
That starts making a bit more sense. I really don't get it otherwise
Well in this case you could simply test two parts printed at various distances or print all 5 pieces as close together as possible. I'd expect faster travel speed/acceleration to also have a measurable effect.
You are really testing all aspects of 3D printing no one else touched. Really well done!
Between him and Stefan at CNC Kitchen, I've learned a lot about the properties of 3D prints and different filaments that led to improved prints. Some of the videos by Angus at MakersMuse have also been pretty useful for me regarding part design and implementation methods.
@@davidg5898 he's now taking the lead about material testing over CNC Kitchen, with tests which are less common and cover less known aspects.
I hope he'll do a "same material, increasing flow rate" tests to evaluate whether printing fast (= fast flow rates not fast linear travel) impacts strength, but I'm not sure his Ender can print above 10 mm³/s, while he would need up to 30 for a good test (5-10-15-20-25-30 mm³/s)
Very cool that you are doing these experiments! Not the result I was expecting either. Thank you for your efforts and sharing.
I did not expect that. We need a specialist on FDM materials who would know about layer adhesion, cristalisation point (T°), cooling, time spent on the bed... and other factors I am not aware of.
Thank you for your time and effort !
Wow, not what I would have expected either
Thanks for sharing your experiences with all of us :-)
Thanks for sharing your knowledge and all these essays. They are very useful for people who are entering the world of 3d printing and obtain good results.
Gracias por compartir tu conocimiento y todos estos ensayos. son muy utiles para las personas que estamos entrando al mundo de la impresion 3d y obtener buenos resultados.
Saludos desde Argentina!
Another thing that stands out is one of your multi had the same breaking point as the singles. Only one of your multi was printed in the center of the bed, like all the singles. Could the placement on the bed be causing this? EDIT: the multi that broke at 80 does look like the center one, you put a slightly longer 1st stroke on the M.
Hypothetically bed could be higher at the corners giving a better squish to each layer. Could be due to heat differences from center to corner or retraction etc etc. Could rule this out by marking the multi parts with a seperate number (0-4) and single parts with s in future test. Or by printing singles in the same points as multi (m0-m4) (s0-s4). Perhaps print s3 and s4 after the multi part print.
I'm happy both times that I didn't bet on the last two tests what the results would be! :-)
The outcome you are seeing has to do with two different effects. One is surface tension and the other is plasticize flow. When the lower layer is still plasticize its surface tension will cause it to move out of the way of the new layer being placed on top rather than mixing with the top layer. In other words the lower layer will just squish down because of its semi-solid state rather than mixing with the upper layer because of the surface tension. Once the lower layer has cooled to a more solid state it will re-melt but because of its solid state it can't as readily move out of the way of the new layer material and will be forced to mix with the new layer being added on top.
This is the reason I was thinking of when watching the video
@@BorisH2000UK After seeing these results I've started printing with full fan after the first layer. Layer adhesion appears to improve. Unfortunately, I have no way to definitely prove this. It would be great if this could be verified in a new video.
Convection? I noticed there was 1 multi print that was similar to the singles and I'm guessing it was in the center of the bed. Also the 84 kg multi may have been next to the control panel and all the others were on the outside where bed convection brought in cool air. It's the same reason why you use walls to keep a steady temperature for materials that don't like having temperature differences. I would print the multi prints again and mark the bed location on each part. You could also print the singles on an outside corner of the bed to see if they are different. For extra credit you could look at the convection flow with a schlieren setup. 😉👍
Excelent test.
My guess is if you look at the gcode preview your fan speed will be higher on the single printed object because the layer times are shorter. This would cause the reduced layer adhesion you're experiencing.
I would like to see your print settings and sliced gcode to confirm.
This seems like the most likely explanation to me.
You can download g-codes from www.mytechfun.com/video/125 (ZIP, bottom of the page)
damn good idea, but G code FAN settings seems fine . This is so confusing, the last two videos with different results as you would normally expect. Umm
@@MyTechFun I stand corrected on the fan settings, but the volumetric flow rate and print speed have much variation between the two files. I suspect this may be responsible for the unexpected results.
I generally print mono color copies as single prints because I get them faster. I never thought about strength. It's interesting though. For multi-color prints I print copies together. Color swaps are expensive both in using a massive amount of purge for the color change, but also you have to wait for the actual swap per layer. I can go from a print where nearly half the time is purge and half the material as well down to a lot less with relative amounts. The color change is essentially free for each additional object.
It would be great to see some desktop FDM manufacturers to implement budget laser technology (eg. LIDAR) for live 3D monitoring and detecting irregular layer paths that leads to issues (bird nest, warping, bad extrusion etc.) and automaticaly excluding part from printing, pausing whole print, or even more advanced - calculating live corrections.
Problem is the center of the bed is usually lower or higher then the edges with the adjustment knobs. So you would have to print the single parts near the knob.
With Octoprint, you can delete an element from the plate when it fails and keep printing the others. You may need a plugin, I can't remember
That’s the Cancelobject plugin. Good idea.
Its useful plugin, but it might be too late for intervention after you wake up in morning...
Please try it with different material, this result is really interesting.
GREAT EXPERIMENT
This appears to be the resting time between layers allowing the previous layer to cure more prior to putting on the next layer, What was the cycle time between layers as it had to print the other 4 objects before the next layer is added on the 5th one. Would you have the same effect if you printed one at a very slow speed or a faster cooling fan speed. We are looking for the scientific answer. Any boffins out there.
So almost like micro-hardening from the second heat cycle?
My first instinct goes in the same direction: If the previous layer is cooled down, the inevitable "pores" (imperfections in the surface) of the layer might be hard enough, such that the viscous next layer gets pushed inside it, effectively "anchoring" the next layer in the previous one. Just like two jigsaw puzzle pieces but very small and all over the contact surface. This is of course highly speculative and can hardly be proven without, e.g., an electron microscope which kind of makes it a bad hypothesis but at least its plausible. 🤷♂️
So would doing another test - single object with pause between layers - have add something to the understanding of the topic? Also what about other types of materials, these could be only PLA properties.
Another possibility is that it's a mild case of annealing and hardening. Layer A has a little more time to cool slowly (annealing) and then experiences a brief reheat (hardening) as layer B is applied on top of it.
Varying the time between layers, with the head moved away so it isn't the fan doing the cooling, would be one way to test that -- though it wouldn't conclusively prove annealing/hardening over other hypotheses.
@@davidg5898 I'm not really convinced by the annealing idea. The break seams are as always between layers so it should rather be an issue of layer adhesion which I would not expect to be strongly impacted by annealing
I would love to chat with you (meet with you) you are a great scientist and your methods are very good.. Thanks for taking the time to make these videos. Peace. Rolf.
Just an idea what may be causing this is that if the plastic is molten and gives better adhesion with layers (single part) then it may create a single homologous solid which means that when the part cracks the crack will propagate throughout the part. This means a singe crack would break the whole part more easily. With the multiple parts the layers still adhere to each other but more like multiple strands of plastic. This would add to strength similar to how rope and carbon fibre work. It would add strength but also prevent a single crack from propagating throughout the whole part. Just an idea though 😜
Average overall temp of the print area could have been more uniform. Also further cooling time between layers of each part potentially allowing for a better bond to the prior layer.
Once you know the answer to why the Multi was stronger, can you please make a follow-up video to explain these results?
Very possible answer: It looks like it's more melting during traveling (retraction, stops the extrusion, more time to melt the filament = better adhesion)
@@MyTechFun Thank you for the reply. Once you said the answer, it makes a lot of sense. But before that I was scratching my head wondering why.
Hey! about Strength Difference ( 5:58 ) ( Explanation )
it may be the Temperature and / or Humidity Difference, in the room when you print ( maybe the Sun Look in to the room )
+ thanks for the Videos : )
No, it's the more melted filament during retraction+extruding paused during traveling
I've recently started printing a fully playable violin out of PLA, and I don't like having to restart my printer every time a part finishes. So, I put as much of the pieces to print at the same time as possible and it will end up using about 120 meters of filament and take around 50 hours. Is it ok to print it anyway or should I separate the prints into about 13 hour prints? Oh, and yes, it does have quite a few supports.
When I print multiple objects I set it for one at a time. It will print the first object completely then go on to the next object. But you need space between each object for the printhead to move around.
Hi Igor, maybe sequential multi printing could be a nice compromise in between the 2 methods. Its Obviously needs a bigger (or at least medium) size bed. And i am not sure if its possible to do with different objects, or they have to be identical. (i am a beginner)
By the way... I have no faint idea how this result is possible. Crystallisation by reheating each upper layer? Some kind of annealing-hardening. I think there is chance you just explored a phenomena in 3D printing! I think it worth to do some further experimenting.
The difference is that in the single the layers are printed on hot layers, in the multi the layers are printed on cold layers.
Multi printing gives the layers time to cool down. Allowing it to become solid layer by layer. Material use can also play a huge factor.
Unexpected results. Several interesting hypotheses in the comments. Fascinating. If it is plasticity then a full plate should see consistently higher layer adhesion, which is useful for me. If it is longer melting time due to travel time then a full plate with short travels and larger parts will not see a strength improvement. I guess.
Yes, after vacation, I will repeat this test with Bambu printers.
i expected the multi-object to print with better quality and possibly better layer adhesion, but my best guess would be either the fan is higher on the smaller layers, or sometimes when the cross-section is small, the printing won't turn out as well and it will end up with a slightly lower cross-sectional area although I'm not sure this would make a difference of 10%.
it would be interesting to see if you modified the gcode so that it printed the object and then retracted and waited for about 20 seconds before printing the next layer if you would get results more on par with the multi-part print.
Yeah, it looks like more melted filament after retraction and extruding pause during traveling is the main reason
Testing idea: 2 pipes in vase mode, 1 at printed at normal speed, 2nd pipe printed at very slow speed.
Have you tried this test with 100% infill? I'd like to see if there's a difference. I did some flashlight and Leatherman holsters at 100% infill out of petg carbon filament and the singles seem to be way stronger them the ones printed in multiple.
Not as multiple print, but something like this yes: ua-cam.com/video/6S1qt2wiuIw/v-deo.html
Did the temperature change while testing? Where both sets of samples tested at the same temperature?
My biggest concern is with the scale. Balances and scales warm up over time. Scales can give different readings depending on how long it has been on. I'd suggest weighing a known amount between each test. I'd also suggest alternating between testing the single printed items and multi prints. This way any variable which changes with time should effect both groups the same.
Thank you for sharing these tests with us.
All testing was done in less than 30 minutes, afternoon approx 15-16h in shadow. No big changes in temperature or humidity..
@@MyTechFun My chemistry scales need to be on for about 15 minutes before they provide consistent readings. I'd be surprised if your scale needs to warm up but it is possible.
I'm surprised by your results (as you were) so I'm just trying to think of things which would effect the data.
Hey. Try this. Set your infill feed to an odd number like 89%. Now edit the code. Insert a G4 P250 before each G1 F1234 or what ever it is. This creates a spot weld dwell point. Much stronger layers. I would like to see the results. Works for me.
Very interesting results! It really shouldn't matter, and I doubt it's the reason for the unexpected results, but perhaps for testing something like this you should print the "single" parts at the same 5-point locations on the print bed instead of printing them all in the center.
It looks like it's more melting during traveling (retraction, stops the extrusion, more melted filament = better adhesion, but 10%, that has to be tested somehow)
For this type of printer, in the y direction the extruder is in the same position for different objects at different y. I think the significant difference is the difference in temperature when objects cool down when moving between them. Position on the bed in should not make much difference.
Do you think that they would be stronger if using a different style of infill settings
Or maybe which infills have stronger properties
are you using any part cooling? With a single item, your fan blows constantly on the part, rapidly cooling the part and slightly decreasing (theoretically) the layer adhesion. With multiple parts, your part cooling fan spends more time away from the part allowing the layers to cool more slowly and granting (again theoretically) better layer adhesion. A larger sample size and playing with the printing speeds/cooling settings could confirm this.
The mistery is solved in the meantime. Added to description of video
I have idea, MAYBE , while travelling to next part -> the retracted material in the nozzle gets more melted / more hotter , and when extruded on the part, it will stick to layer better ?
Something similar i had in my mind as well. A bit similar to the annealing-hardening process, but here its happening on each layer, one by one.
That's exactly what I got in my mind during the last sleeping ;-) Now I am thinking how can I test this theory.. not only the retraction, but during travel (after retraction) extrudion stop, more time to melt the filament
@@MyTechFun maybe print few parts without any Retraction settings , and then try few of them with retraction with e.g. different retraction speeds, so when you will select VERY slow retraction speed, it will take longer time to retract and again print . THere will be need to enable retraction at each new layer i guess (not sure if prusa slicer have this ) but Cura have checkbox for it.
Is the plastic a 'glass' or is it crystalising when cooling down. Perhaps there is an unknown effect occurring printing multiple parts? Is there better bonding at slightly lower temperatures? Is there some oxidation occurring? Could you put the printer in a sealed bag with nitrogen gas flush to eliminate oxygen?
It looks like the reason is more melted filament during retraction and pause in extrudion during traveling. So when it starts the layer, more melted filament sticks better to existing layer.
Was ur weather differnt during printing my next best guess is cooling actualy helps
Mistery is solved in the meantime. The reason multiprinting was stronger: During traveling, the retraction and extruding pause will melt the filament more.
The multi part print has better layer adhesion because the layers have time to cool.
You have to think about that 3rd layer below your two working layers. On a single small part print it's still soft when the new layer goes down so it's getting moved and heated a little and compromised. It's like wiggling two objects you are trying to glue together.
Perhaps the air temperature or humidity were different enough between the prints to affect them.
3D printing was done one after each. When last single was finished, I started with multiple one. No significant change in temp or humidity. Also testing was finished in less than 30 minutes
I bet it had more to to with the surface of cooled layers being not as smooth as freshly printed. So the next layer has more surface area to stick on
Or retraction and stop extruding during traveling melted better the filament
@@MyTechFun then should lower extraction speed have a similar result in stronger bondage. But didn't found a source to prove this and I'm just 1 month into printing
My thought was a result from the item sitting longer on a heated bed; more fusion?(lol)
When printing with hot material into a not cooled down layer, the result is less precise. The structure is not as regular for single prints and the layers do not overlap optimally. My question would be, did the you see any differences in print quality between the S and M objects? Did the S objects look as smooth as the M objects where they broke?
Same visual quality but this depend of the printer too. The layer adhesion mistory is solved in the meantime. More melted plastic due the retraction and pause in extruding..
fascinating the multi ones are stronger also is the strength more if you print it on its side ?
also could the reason for the multi comes out better is because there is more time between each layer as the printer travels between the four ? does it make sense the material being colder at the time of added layers increases its strength ? maybe can look at that.
Yes, that's exactly the reason (Suggestion I got in one of the comments earlier and I already tested it). Retraction + travel = more melted filament inside the nozzle.
I hate it when my testing doesn't give me the expected results
Nothing to hate! Just an unexpected learning journey
Is it possible the S failed at a lower value because they were printed first (so the filament was more on the outside and not as dry)?
It looks like it's more melting during traveling (retraction, stops the extrusion, more melted filament = better adhesion, but 10%, that has to be tested somehow)
@@MyTechFun You can test it inserting a z hop and pause between every layer
That good engineer quip - nice. Sadly, I can't help you, as I am not. I'm curious to see if anyone has a good theory as well. Crazy stuff.
Mistory is solved. More melted filament when retract and pause extruding during traveling.
Okay, that makes sense. So does it follow that you could potentially achieve similar results with the single parts by increasing your nozzle temp? Super intriguing, really drilling down into the details of print settings and impacts on strength. Thanks for all you do!
Hello Igor. Thanks for share this info for us. What do you thing what happening if U will playing with the minimal layer time? ( ex: setup the slicer software : layer time under 15s --> the printing speed will automaticly decrease till not reach the 15s ) It can help for the small cross sections or not?
Yes, that is one of the possible reasons, just don't know how slowing down make weaker layer adhesion, it should be stronger.
And probably this one will be the main reason: It looks like it's more melting during traveling (retraction, stops the extrusion, more time to melt the filament = better adhesion)
Very unexpected result
Also not what I would have guessed... Funny right?
Yes, but thanks to very smart comments, mystery is solved (retraction and pause during travel to other object, more melted filament = better layer adhesion)