Don't forget to share this video on Facebook, Reddit, Twitter and other social media!

@THEOGGUNSHOW

3 жыл бұрын

I would love to see you redo this experiment with the following parameters: After printing has completed, place each unit into a container and fill with plaster of Paris (just as you would do in mold building). Allow to cure and then place into the heating chamber and slowly ramp temperature up, until the appropriate annealing temperature is reached internally. Then, gradually lower the temperature, until room temperature, remove the plaster and then perform stress tests as before. Thanks

@seville2k

3 жыл бұрын

What happens if you vapor smooth this filament?

@user-mc6rf7do7z

3 жыл бұрын

try these threads in solvent vapors

@elmariachi5133

3 жыл бұрын

Great tests as usual! Why so negative about the results? Besides not being improved by annealing, the hybrid material gives great results as it is. ABS without warping, nearly the layer adhesion of pure PC and the ability for vapor smoothing? That's all I ever wanted..

@anotherriddle

3 жыл бұрын

Impressive dedication to rigorous testing! I'm so happy I found your channel :) I have an idea that might be worth trying: You can embed a 3D print in a fine powder, compact it and then put it in an oven at a temperature past the glas transition temperature. The idea is that the surrounding powder takes the shape of the 3D print and holds the shape when the polymer can't hold it on it's own anymore. Think lost form casting, but with keeping the original plastic part instead of replacing the volume with metal. I tried this a while ago but unfortunately didn't have access to a 3D printer for long enough to really test this. However, first results (after a couple of failures) were promising. Here is what I tried, that allmost worked: I used an ABS print with essentially 100% infill, put it in a ceramic flower pot and embedded it dry cement powder. You have to vibrate it to fill all the gaps and compact the powder (just like with lost form casting). Then put the whole thing in an oven just like you did here but you can use higher temperatures. Afterwards you can easily brush off the powder from the part. The only reason I used cement was, that I knew it can hold the temperature, and it is a very fine powder, so it preserves detail. I'm sure other powder could work too, maybe better, but it was the best idea I could come up with at the time Mistakes I made: In the beginning I used too little infill and the print collapsed inside the powder form because, obviously the plastic doesn't stick to the powder well. Then I used too complex of a geometry and didn't surround the print well enough. Ideas that might work, but I didn't try yet: Use fine table salt as the embedding medium and fuse it together with water mist (fine spray water bottle or ultrasonic). The water dissolves a tiny bit of salt and crystalizes when the water evaporates and those tiny crystals make the remaining salt grains stick together. I found that ABS sticks to this salt surface (kind of). My hope was that the almost molten plastic sticks to the salt wall well enough so I could use lower infill percentages. (Also you can easily dissolve salt with water.) Maybe you can make use of some of these ideas. I'm sure if someone can it's you. (probably someone else allready came up with something like this, but at the time I couldn't find anything similar) I hope to continue my tests in a month or so, when I have my first 3D printer. I wish you all the best with your 3D printing projects!

I swear you do more scientific 3d print testing than the manufacturers do. Awesome work!

@overloader7900

Жыл бұрын

Proper science takes human effort and money, you know, manufacturers don't like using humans

Nothing quite like that "Wait, what?" moment you get from hearing someone is 3d printing their filament for 3d printing.

@mophie6941

3 жыл бұрын

exactly me 😂😂😂

@OrangeC7

3 жыл бұрын

That's exactly what I wanted to say lol It's just awesome the kinds of things that 3d printing allows us to do

As a materials scientist studying additive manufacturing, this is really impressive. I'm lucky to have a lot of expensive tools at work but what you're able to accomplish with your DIY testing tools is inspirational!

@rmp5s

2 жыл бұрын

So...when are you going to make some videos with those expensive tools, man?! lol

I think that you missed the opportunity for annother interesting test. As far as I know the properties of plastics get worse when you heat them repeatetly to their melting point. Because of this phenomenon I would expect your printed and then reprinted Material to have worse properties, than the Material that is only printed once. Maybe you could try this out by printing abs filament, from your abs filament, (maybe even repeat this process n times) and then test the material properties of the testsamples printed from the reprinted abs. I would be interesend in your thougths on this. Grüße aus dem Norden😜

@imhavoc

3 жыл бұрын

This! Science. Baseline. Yes!

@jayphone1

3 жыл бұрын

I asked this myself as well: Does the material degrade in properties when you leave it at high temperatures in an oven for a few days.

@JuliaC-sp5qk

3 жыл бұрын

@@jayphone1 I don't think the annealing process would degrade the material, as it isn't heating it above the melting point. Could be wrong tho

@nf794

2 жыл бұрын

As long as you dont go over the temperatures in which the molecular chains are damaged No. In fdm we use thermoplastic plastics. These can be melted and reshaped endless Times. In theory that is.

While you didn't get a dramatic strength improvement, you did get a slightly stronger end result. More importantly, the combined material was much easier to print than PC alone with much of it's strength. Thank you for doing this huge project!

@eelcohoogendoorn8044

3 жыл бұрын

Yeah I see printability as a huge potential application of this concept as well. An inner core optimized for strength and an outer core optimized for layer adhesion. Putting some nice long carbon fibers with horrible flow properties and just pushing those out of your brass nozzle would be pretty cool, for instance. Not sure if there are many promising material combinations though.

@dr.doppeldecker3832

3 жыл бұрын

@@eelcohoogendoorn8044 i love the embedded fiber idea! Probably needs a more heavy duty nozzle and pump to smoothly print fibers. If the print is one continous motion from start to end, wouldn't it be possible to have just one long fiber throughout the whole filament? I think one continous fiber filament printed in some kind of criss cross pattern could be very strong!

@rileyneufeld7001

2 жыл бұрын

I think if Stefan used a smaller nozzle diameter like a .2 or even a .1mm nozzle he could've experimented with higher PC to ABS ratios and really get the PC to show it's strength with the bed adhesion/warp resistance of the abs making large prints possible. A 75-25% PC-ABS mixture would really help the strength values.

The finish is amazing! Specially from a filament that has layer lines on it and two different materials on it.

"Four and a half Days" lmao literally goes ahead and binds the timer in place with a magnet

@etch3130

3 жыл бұрын

the timer has an infinity time setting too.

Dude that filament is worth it just for the cosmetics. It looks super good for such a technical filament

I have a vac chamber that I use for resin degassing and infusion, and I've done quite a lot of work on trying to make 3DP parts stronger and/or impermeable using various coatings and resins. FFF parts come off the printer very leaky. My best results for strengthening FFF parts have been by resin filling: - printing them with sparse porous infill - coating the outside with plasticote or similar to seal it - drilling two or more holes in the top (entry and vent(s) - putting a silicone funnel if the entry and a small tube in each vent and sealing them in - filling the funnel with thoroughly degassed slow epoxy or PU resin - putting the whole thing under vacuum for most of the working time of the resin. - returning it to atmospheric pressure a few min before the resin gells

I would say given how the dimensional accuracy held it would be interesting to see how much temperature it can take before it loses that

@osimmac

3 жыл бұрын

yeah 135c might be enough to improve the strength significantly

@Mrcaffinebean

3 жыл бұрын

I was about to say the same. The heat treat temperature is the variable that likely has the best ability to improve the results. None the less the idea has legs and I’m sure the kind of filament will be out soon and most testing will be possible

@osimmac

3 жыл бұрын

OH! perhaps he should anneal it after printing it the first time as well? The i was thinking about how the people in the research paper printed a large core and then heated it up and drew filament from it, perhaps a heat treatment step before printing with it again help fuze it together?

@ARVash

3 жыл бұрын

@@osimmac not a bad idea, plenty of room for experiments IMHO

@Bahnamoon

3 жыл бұрын

Maybe place it in a vaccum chamber while annealing it? May help layer adhesion

I love how you test new concepts of improving 3d printing performance. I believe I will become a patreon. I would do the test again with higher temperatures. Because I believe that Tg is not enough to fuse layers together. You probably need to reduce the material viscosity quite a bit. As a first step I would try to get the temperature when the multimaterial parts deform too much. And then heat up to just below this temp. And maybe there is a way to increase the PC content, because we need ABS only at the interface. Is there a way to paint PC filament with ABS slury? Or you could print 2.85 mm filament.

@AlexanderTasch

3 жыл бұрын

Isnt this what he actually did? In the video arount 8:30. If its to much the parts will deform. Depending on the application this is a even bigger problem.. :(

@jayphone1

3 жыл бұрын

@@AlexanderTasch Yes you are probably right. I thought that up to 165 degrees where the PC failed would be still quite a margin to raise the temperature further. Might be no the case for the material mixture. The best way to check would be to heat up the multimaterial benchy till it looses it's shape. But maybe it's not too far from 130.

@psionicxxx

3 жыл бұрын

@@jayphone1 I agree, the same thing I was going to write. Try slightly higher temp, those few degrees inside the ABS transition phase may be crucial for the proper layer fusion

@Balorng

3 жыл бұрын

Point is, PC already have excellent interlayer adhesion, better than *tensile strength* of ABS in fact. Why not simply print with pure PC than? If you really want good interlayer adhesion, warp-free annealing and care about strength, a combination of PC outer layer and gf/cf filled PC core will work best. Same for Nylon I presume. A russian startup as already making coaxial filaments, I am already testing a sample of pc/abs (50/50 ratio), it does look promising.

@AlexanderTasch

3 жыл бұрын

@@Balorngthere is more than one point to consider. There is no "point is" in 3d printing. One idea is that even very thin layers could in theory benefit from stability while annealing. Laminar flow provided that's a benefit. Also your approach has to consider the interface when "just" doing a multi material print as a new point of failure. This approach could be available for most printers without or minimal mods.

As always, in awe of your methods. Thanks so much for sharing!

Your dedication is most admirable, and your data marks the progress in manufacturing in this new field, pointing the way to advancement. Gut gemacht!

Whaou I am blown away by the quality of the video and your research! Really one of the best channel about mechanical engineering in fdm 3d printing, congratulations 😊

WOW, I really appreciate how much real work effort you put into your videos. Quality work!

For a more complex internal PC geometry perhaps you could print the 'filament' at 200-300%, then draw it down to 1.75mm through progressively smaller dies? I've seen it done with wire for jewellery - might be a way with plastic eg if heated slightly.

Amazing test, thank you for all the work that went into the making of this video.

This is not my idea, I saw it on another video by Teaching Tech, but it would be really cool if the big 3D printing KZreadrs (you, Makers' Muse, 3D Printing Nerd, Teaching Tech, etc.) did a 3D printer build off, kind of like the Scrapyard Wars of Linus Tech Tips. I don't know how you would organize it, I I just think that would be a very entertaining series of videos.

@ameliabuns4058

3 жыл бұрын

I'd love to participate in something like that even tho I'm not a big KZreadr or anything

@CNC Kitchen, you need to test annealing the pc-abs mix at gradually higher temps until it deforms, then pull back some 5-10Cº, instead of the pre-test on a different oven. Maybe even with shorter annealing times, if that's not detrimental in your experience. What do you think? :)

Love your scientific approach, Stefan. Other channels often just print a single benchy and extrapolate wildly

Really amazing content as usual. This must have been quite a job both time and equipment wise to pull off, thanks for the informative content.

I love the scientific method you use for all of your testing.

Always love your testing of new concepts. Great video!

i do love your videos man. Such a fantastic resource for people who use 3D printers for engineering and not just pretty model making. Thank you ! Let's hope this PC-ABS core blend end's up as a purchasable filament.

Really appreciate your efforts Stefan.Yes, ofcoure I will share the video.

The most informative 3d printing channel. Thank you Stefan!

Stefan, you're insane but in the best possible way.😎 Thanks for the giant amount of time you have invested and the shared results.... Keep up the good work 🖒🖒🖒

Thanks for putting in all that effort!

Great work you have done, as always 👍 Thanks for sharing 👍😀

This channel just keeps becoming more advanced. Wow!

As always, a very thorough video, even if you don't reach the results you hoped for, I find your videos both entertaining and educational, keep up the good work!

@vidznstuff1

3 жыл бұрын

He didn't "hope" for results - the results were claimed in a scientific paper. Did you miss that part?

Great study. Don't get frustrated by the results. That's why they call it "re"-search! I'm sure this is very temperature dependant. Being above Tg does not guarantee diffusion. I bet you just want to get as high as you can without softening the PC. Keep up the good work.

And they said i was being crazy printing my own filament! But excellent work! Do you happen to have the elongation relative to the stress available to you? would be interesting to see how that changed.

@uiopuiop3472

3 жыл бұрын

:buttofking: :pogv2:

Wow. You are really adding value to the table :-) Keep up the good work

Wow, this is amazing, printing your own filament!

Love you cnc kitchen stay inspiring!!

Can't imagine the work this overall effort took. Even the results were disappointing, I think the effort was well worth it.

That's a really neat use for the tool changer!

Fascinating work. Thanks!

Even null result is still valuable data! Thank you for putting the huge effort into the tests and saving us mortals from having to do them ourselves.

A huge amount of work indeed!

3D printing composite filament (wow!) plus outstanding experimental work. 100 thumbs up.

@xl000

3 жыл бұрын

they will cancel to 0 thumb up..

The amount of work and time that went into this... Ich ziehe meinen Hut...

Stefan is not only making KZread content but actual R&D

@CNCKitchen

3 жыл бұрын

Thanks!

an amazing amount of experimentation and work. hats off. shame the results show no material improvement. thanks for the video

a true science as usual, please keep it up Stefan

Nicely done. A clear engineering mindset. I really enjoy your work. Ausgezeichnet.

This is really briliant 👍

Sehr cool. Du machst das richtig gut. Gucke dich immer wieder gerne.

Your tests are really rigorous. I appreciate your methodology.

Stefan, your bed heating demonstration was both enlightening and supremely German. I loved it.

Amazing work

I would like to point out that, for the tensile strength test, ABS/PC composite had the best failure mode of the three samples. Not the best performance, but the best failure mode. After quickly reaching what appears to be the yield strength, the neck turns white and keeps stretching a LOT. The ABS sample turned white on a very small section then immediately fractured. PC necked some but had no color change. The ABS/PC composite should, irl, make a loud snap or sudden jolt when overloaded, then the stressed section will turn white and stretch significantly before falling apart. This will give the user a lot of time and warning during and after failure.

This would have been an interesting application for the GOM Correlate unit you had... scan each Benchy before and after annealing to show the distortion. Might be tough to get it to register the post-anneal ABS though. An amazing amount of effort, as usual. You are a great example to others. Take care, and thank you.

fantastic work

Amazing analysis, and really well done as usual!. I think this material might retain its strength more over time, and may lead to longer lasting parts rather than stronger parts. A similar approach is done to pvc id cards, straight pvc only lasts about 2 years and becomes brittle and cracks, the better cards we use are a pvc/pe 60/40 sandwich, they last twice as long.

Dude you are a wizard, nice job

Fantastic Work! Thank you!

Thanks for the video

very nice study, thank you

Very impressive. It's very refreshing to see new concepts being applied to improve 3D printer/filament technology, alas, the results were not as promising. However, please continue with the full Engineering prowess you exhibit on each of your videos. Thank you.

amazing work, thank you. Would be curious to see your review on the E3D toolchanger, an interesting machine for experimental printing.

As always, an amazing video and incredible work behind it. I have a suggestion for future ideas: what materials / combination of materials do survive common disinfection processes, such as alcohol, bleach, or the most aggressive of all, autoclave! I've been designing substitutive components for medical use this quarantine and I've yet to find a material that survives autoclaving. I've tried ABS and PP, without success, and I think PC will not make it. Great work and keep doing these amazing videos! Congratulations from Granada, Spain!

Absolutely fantastic devotion to testing this. Well done. I would love to see your findings with different core shapes, specifically when the pc is also part of the surface area. I also wonder if 130 degrees was the correct temp as it was a guess. Thanks again, I loved the video

Since PLA is kind of the gold standard for annealing, it would be promising to redo this experiment with PLA and an ABS Core. If the endresult would have 80% of the properties of annealed PLA while retaining dimensional properties, that would be really nice.

@Gengh13

3 жыл бұрын

Shouldn't it be ABS core with PLA exterior?

@6AxisSage

3 жыл бұрын

Would those materials play nice together?

@extremewirehead

3 жыл бұрын

Like he noted at the start, maybe the reason it isn't done is because these 2 materials don't fuse well? As I recall (for other manufacturing uses) they don't really fuse.

@fabianbinder3681

3 жыл бұрын

@@Gengh13 that’s what I wrote?

@fabianbinder3681

3 жыл бұрын

@@6AxisSage I always figured yes, because one time I got this weird Chinese filament which the vendor said was ABS, but didn’t dissolve in Acetone at all. Since it was high temperature resistant and nothing like PETG (e.g. splattering when you brake it) I always assumed the Chinese factory made a mistake and put a considerable amount of PLA into the ABS pellets therefore ending up with a kind of „hybrid“ between those two (printing temperatures had to be in the middle of ABS and PLA to get usable prints, too). However, this assumption might be totally wrong.

Who would downvote someone putting this much work into just trying to figure out if you can make a stronger material at home!?

I'VE BEEN WAITING FOR THIS EVERSINCE I SAW YOUR TWEET!! :3

you know hes pro when he prints his own filament

GENIUS!

always wondered about heat treating in a deep column of water, sand etc, just about any pressurized environment. might help with warping

Good work!

Awesome! Thank you for testing this out =D maybe looking at carbon reinforced filaments distortion during annealing could be interesting... hmmm...

I'd love to see more experiments done with this composite filament technique. It's clear that the oven step is going to have to be hotter in order to truly get rid of the layer lines.

I have that thermocouple thermometer. For the cost (cheap) it's a really good instrument. Of the many cheap ones I have it seems very accurate and has extra features others dont

Good informative video!! Multi materials are Interesting technique in 3D printing field!!

I finally see why the channel is called CNC KITCHEN

Very nice experiments! It put me to think!

wow I want some that is so cool

Thanks for the great video, I'd pretty much stopped watching most 3d Printing videos, there seemed to only be rehashes for already covered ideas. One test I'd like to see, is a mix of PLA/ABS, or even with PETG, something the average 3d printer can do - I REALLY didn't think you'd still get separation of materials after extrusion, I thought it would just mix in the nozzle, and was very happy you I was wrong.

PC-ABS is actually what the stock Joycon shells and the backplate of the switch are made out of

@mr2good4name

3 жыл бұрын

Its a pretty common plastic blend for injection molding. at work we produce a lot of parts for car interiors with Bayblend and similar material.

@vidznstuff1

3 жыл бұрын

This is not a blend, or did you miss that? this is an attempt at making a composite. The 3D printing community should run with the methods Stefan came up with here and maybe with that hive mind we can come up with a real composite that provides a unique hybrid of properties.

@mr2good4name

3 жыл бұрын

@@vidznstuff1 i know but what B10K PlaysGames said about the joycon shells is a blend ;)

This is really cool, and props to you for managing to recreate the research so well with no special parts Edit: oh, it seems I was a bit too optimistic :( How about trying better cross sections next? a square has less interaction that a star or the shapes used in the paper.

50secs ago? noice! youtube, u did well this time ;)

PLA and TPU would be very cool

That e3d printer is awesome

Nice work.

Nice video! Appreciate the time taken to test this! I'm curious if annealing the parts in a vacuum would help to pull the air out from inside. Or... in a much more difficult option, print the part in a vacuum to begin with.

I'd be curious to know what the PC-core ABS does in terms of adhesion strength when exposed to acetone for a time longer than usually used for smoothing. The PC could allow it to keep shape while the ABS could melt and fuse. Another interesting question: When welding plastics, like damaged PP canoes, argon atmosphere is used to prevent oxidation of the plastic that prevents fusing. Could it be that a similar effect - oxidation by surrounding air forming a non-fusable layer on filament - is happening during FDM printing? Would prints done in argon fare better on layer adhesion?

@CNCKitchen

3 жыл бұрын

It might, but the layers that should fuse together as already bonded air-tight, so I think the impact of polymer oxidation shouldn't be too high.

Love your videos Stefan! Super interesting video, shame the mechanical yield and fracture performance was not greatly enhanced. Perhaps an interesting investigation may involve a PLA and PC (or other core) mix? I imagine that crystalline domain size will grow larger at temperatures further from the glass transition, as the available energy for amorphous to crystalline activation increases? Would love to see the comparison! Super nice video man

Cool experiment.

That's a lot of work and a nice oven. My only suggestion would be to create "giant" filament and extrude it through a 3mm nozzle to remove any voids and inconsistency and let you create more complex structures. I wouldn't expect the results to be different...but you did ask.

Very cool video. I'm sure it really was a TON of work, too. Thanks for the effort. I'd love to see what the overall strongest filament+processing procedure there is looks like. Like, a "how close can we get to milled aluminum" video. That'd be really interesting as, there are always trade-offs of course, but it wouldn't surprise me if we could get pretty close these days!

if you splash acetone on polycarbonate it breaks like sugar. but as soon as the acetone evaporates it is strong again. so if you ever need to break out of a polycarbonate box, make sure you have acetone and a hammer handy.

I chuckled when you threw some ham and cheese on the print bed. CNC kitchen indeed.

That's a really cool idea! Some of my immediate thoughts have already been mentioned, but here: Unless the diffusion is supposed to work on the PC as well (if the heat treatment temperature was that far below the PC's Tg I don't think it would have a major effect), wouldn't the expected layer strength improvement be limited to the strength of the ABS? It does show an improvement over normal ABS in both treated and non-treated tests (it is a composite when printed after all), but if the Heat Treatment is still below the Tg of the PC (a compromise, as mentioned), and the polymers are still phase separated, then it seems like the ABS is the main component that's supposed to fuse the layers. My point is, if PC-Core filament is already stronger in Z than annealed/heat-treated ABS (as shown in the tensile test results), and the diagrams from the paper seemed to show only the ABS component flowing/fusing, I don't think I'd expect to see a huge improvement, particularly with a high cross-sectional area of PC. The test results might be different with a lower percentage of the cross section being PC or a higher grade of ABS, since the PC would still retain the shape, but the ABS would be a larger contributer to the strength of the printed part. This is probably part of the reason for a star pattern in the original paper instead of a core, so that the improvement in strength could be more easily demonstrated. EDIT: I think this might also play a part in why the impact strength saw a more noticeable improvement, since the PC parts showed a lower impact strength (but with lower scattering), likely due to the decrease in ductility, but the PC-Core parts actually showed an improement and lower scattering, likely the best indicator of layer fusion in the results.

@YourAverageJoe1000

3 жыл бұрын

Another note (to avoid an extra-long comment): Do you think a pressure treatment regime could bring out some of those desired properties as well? For example, heating the parts to the treatment temperature for 3-7 days, but slowly decrease the pressure (to draw any gases from the print, whether it's typical polymer volitiles like Bis-Phenol A in the ABS, or porosity between the layers), hold a partial vacuum for a few hours, and then slowly increase the pressure to be greater than atmospheric (to compress the parts and encourage layer fusion)? This would also mean varying the temperature during the test (to stay below the PC's Tg at all pressures), but it might be interesting to see.

Very interesting video!!! Thank you 🙏🏻. Maybe temperature + vacuum will helps? I mean: vacuum the oven, heat, delete vacuum staying heated for some time - done. ?

Thanks for doing all of that work to give us all that info. I would love to see more on low cost ways to do annealing. I'm on a very fixed income, so any tips would be appreciated.

Wonderful video, thank you. What about acetone smoothing the pc core abs?

I wonder whether having the outermost layer of something even lower Tg than ABS would allow a more pronounced melt fusion of the layers. Also, a material with fast melt flow. I have a hunch that PLA on the outside could be really good - it qoes fully liquid a sufficient temperature

This really makes me curious as to whether you could make 'Safety Plastic' by combining PETG and Ninjaflex. They print nicely and bond well at a similar temp range.