Thank you for creating content that is innovative instead of just copying what others have done. You are contributing to the knowledge of the printing community. I just began supporting you on Patreon. Thanks!
Aside from the actual content itself, which is very thorough and informative, I wanted to commend you for the way you presented it. Your video editing skills have greatly improved. I enjoy your use of time lapse, speeding up, and splicing shots together.
Excellent! Creeping is too often overseen in 3d-print material tests but has so much impact when dealing with plastics. Thank you! Btw: in my experience the gap between PETG and ABS is truly noticable in real life parts, therefore I moved to ABS. But might depend on the brand.
@@alejandroperez5368 And? ABS is still better at creep resistance than PETG. Prusa uses PETG it's because they had the freedom to design a printer with parts that only required certain specific properties, and had the freedom to design both the models and the polymer blend to meet those needs. Besides, they're mostly there for aligning parts, almost none of them are load-bearing. The Voron team doesn't sell printers, they can't hire factories to produce specific metal parts for their printers or fabricate their own filament blends, they're just a grouplet of hobbyists designing printers from parts that are available off-the-shelf, so they have far more limitations. The Voron team had to design their printers around the minimum properties of common off-the-shelf filament blends and chose ABS, despite the eternal nagging by new people about the possibility of using polyamide or PETG - their printers simply won't perform well with those.
Fantastic video, I really appreciate the scientific methods for testing of the material properties. A DTI gauge jig for all printed peices would be a massive time saver for future videos. Keep doing what your doing,you've planned out and documented your finding like a pro. You deserve more viewers to the channel,keep it scientific and the nerds will keep on subscribing. The torque test was very useful,and the reletively low deformation temperature for PLA was eye opening. Very much looking forward to the ABS results in your upcoming video. Thanks for taking the time to experiment and document your findings for us.
The compression creep tests are fascinating. I had a functional print with a live hinge compressed around a pipe, and over time the PLA it was printed in migrated away from the bolt until it became impossible to tighten it anymore. I was hoping to find PETG or ASA had less creep under compression, but evidently not. I guess the best that can be done is to print the compressed area at 100% infill and use the largest washer I can fit in that area. Or carve it out of wood... In any case, your data and collection methods are really handy, thanks for the data!
I’m new to 3d printing and trying to decide what my go to filaments are going to be for differentiate cases. I’ve watched a couple of your videos in addition to those from other creators now and yours are far the best thought out, scientific and controlled. Subscribed and looking forward to new videos and watching your testing methodology refine over time.
One theory about the immense creep of the Nylon samples: as we all know, it takes up a lot of moisture (up to 10% in weight), which it does after printing as well. I recently stumbled upon a statement of a material science student, who investigated the mechanical properties after leaving the parts out in the open. He stated, that the elastic modulus was almost half compared to freshly printed ones. So the creep of Nylon could actually also be down to moisturizing, maybe it would be interesting to see how „old“ parts would perform the same test
Great video. Please continue to do these creep tests with other materials! There are so many interesting filaments out there and hardly any information on creep.
Thank you very much, very helpful and careful tests. I am printing a spring for a battery holder. The PLA spring deformed under the small tension of holding the battery. Based on your test I'll try PETG.
Thanks for the info! The screw test is extremely relevant since it's very similar to the kinds of loads some printed parts face when they're actually in use. I'm curious about whether annealing would help the nylon (or PLA) too.
Outstanding work! Clever, thorough tests. There are many channels with discussions about the physical properties of different filament types, but few conduct meaningful tests.
Thank you you Igor for an excellent and super helpful test. I have also found ASA to be most useful for outdoor parts that see heat but always wanted to see actual tests like yours.
I really like these additions to your tests. I'm all for tests that are more relatable to real world scenarios rather than ISO lab tests. For the torque test maybe you can include the total amount (added up) of degrees turned after 6 days. This would relate to "tightened additional 2 full turns".
I'm missing PACF and PCCF filaments. Those are commonly used in printers parts, and after some people complaining about PACF creeping, my choice could be PCCF.
I got some filaments from eSun for testing. ABS+, ePA and ePC (not CF), there will be a comparison video too. But carbon fiber tests will be included in near future. Only waiting for some hardened nozzles to arrive.
It would be similar as in enclosure. And maybe you will find this one interesting too: not outside, but in car: ua-cam.com/video/QdAKd_YbsjI/v-deo.html
Wow, that's really professional! Your results correlate with the general opinion the engineers have at my work. PA (or Nylon) has wildly different creep depending on ingredients, if everything fails, add glass fiber :) Thanks for going through all that work, analyzing and documenting! This should have more views.
This is awesome. Might I recommend a slightly different "screw" test? Instead of testing how much you can rotate the screws at a given torque, a more in-machine relevant test might be to test the torque required to break the nuts free. Procedure: Tighten nuts to a specified torque (using your high-repeatability click wrench), then check the maximum torque required to "break free" using your digital one. If the material softens, there won't be as much pressure against the threads, and it should have a lower "break free" torque.
Hm, great point, if I understand correctly. Instead additional tightening the screw, more important is the torque for loose the screw. That also a realistic case too. Perfect would be if I could measure the pressure between 3D printed object and the nut or washer for example. Imagine you use screw to join two parts, you want that pressure between two parts to be constant and strong. Now, that's hard to measure. What I can easily measure, is additional twisting the screw (which is in linear corelation with that pressure), as you can see in video. Maybe at the end of the test, I could measure the unscrewing torque too.
This would be very difficult to get repeatable results since a great deal will depend on layer adhesion. You are describing a shearing test in essence.
@@MyTechFun the most relevant statistic for screws in plastic is how the clamping load decreases over time, this is usually what's speced out in injection molding plastics tech sheets. However this is sometimes hard to measure, requiring a load cell/strain gauge
Great test. Just a question about the nylon. It looks like it was not dried before use, looks like you can see that it has bin bubbling when extrudes. You also did not mention if the nylon have had time to equalize with the humidity. Freshly printed nylon get softer over time cause it absorbs moisture.
Nylon was dried, but probably not enough. 4h at 70C (it is recommended 12h at 70C, but I thought that it stay dry, because it was dried before packing in vacuum bag)
@@MyTechFun i doubt it will affect the result much since it is tested over a period of time. It will absorb moisture and get soft. Directly after it was printed the piece that was printed with wet nylon would prolly lose.
This is something i needed! I would like to add untreated PLA vs annealed PLA comparison to the test, i wonder if annealed PLA will perform better. Also a material that often gets neglected in tests is HIPS. It is very inexpensive! It's supposed to shrink less than ABS in printing, and similar to ASA, and be more rigid and less smelly than ABS, though smell potential exists, and high shrinkage potential exists too. I think i would like to see more of it. I think there can be a massive difference between brands, as opposed to ASA or PETG or most non-silk PLAs, which is unfortunate. I do like the C-shape test, in the absence of other tests, so if i had to choose just one test out of 3, i would make a decision based on it alone, and it's fine to me as-it-is; but i think screw test is interesting too, it looks like it will provide interesting and distinctive results eventually. I also actually liked all the tests though. Maybe if you use a piece of double sided tape to secure the nut, nylon can survive the 2-point test too, well enough to make a measurement, with enough room for it to bend down. But it also looks like the most fragile of the tests, with potentially needing a different thickness of the sample depending on modulus to be comparable. 45°C is an excellent choice for a hot box, as it's more or less representative for example also of use inside a car or inside a computer. Normally say you can expect the air in the computer chassis to become 40°C under load and SSD or hard disk can also reach 45°C. Currently i have a 3D printed SSD retainer in my PC and will probably have a 3D printed CPU cooler mount in my next PC. Ideally i won't have a 3D printed backplate but we'll see. I recall Stefan (CNC Kitchen) did have a 3D printed CPU cooler retainer and it failed on him, and it was i think CF Nylon? In hindsight, not surprising. He also experienced his Voron fail in exactly the same manner, with material creep under mild heat! I actually don't terribly mind PLA being used for part cooling fan assemblies and such, because they aren't really load bearing, and also can survive some permanent deformation. But i do think PLA is less ideal.
Thanks for your detailed comment. Yes, any kind of nylon (CF nylon) have this creep problem, it is good in impact tests for example. 2 point bending test I will replace with 3 point, only I have to solve the problem of accurate reading (with dial indicator, more accurate). And on my list is testing of annealed PLA vs PLA (and vs Nonoilen filament, because it is PLA type but heat resistant up to 110°C without annealing, tested earlier). For ABS testing for voron parts probably I have to go higher, up to 50C (accorting to response to my question in Voron group) and for longer time, not only 2 hrs. Mayabe 4-6h.
@@MyTechFun Thanks for the great video! I'd also be very interested in results for annealed PLA. Nonoilen seems to be similar to extrudr GreenTEC Pro, maybe that would also be a good candidate for testing...
Nice Video. I would suggest to normalize the results to get a % value (eg. Platic defimation/lenth). I have done similar measurements with another geometry. In my measurement PLA and PETG performes quite similar (20°C). But annihling the parts used to have an extrem impact on the creeping. It more than halves the creeping (also on PETG). The effect of the annihling was much much bigger than the difference between the materials. With Extrudr NX2 I also have found a PLA, wich dos not deform so much during the annihling.
Fantastic methodology. I really respect your attention to detail. I'd only suggest that your graphs be modified to show just the deformation for readability, ie. for test one subtract 12mm. Very informative.
Great work & valuable results! Thank you! For C-ring test, maybe you could modify sample and/or fixture so top ear mounts to fixed bar. Then measure lower ear to bar (similar to cantilever flex test between sample & fixed table) Cheers!
Very interesting, Egor.. Thank you. I don't have a 3D printer but I DO have a curious mind. That's all that was really needed to enjoy your vid. GR8T post editing by the way. That was a chor in itself. Cheers from So.Ca.USA, 3rd House On the Right.
I thought the screw test was useful and would happy to see if you can develop it a little more to get more meaningful numbers from it. Maybe have the screws hold 2 pieces together with a pressure sensor in between (Arduino/Raspberry Pi pressure pads are pretty cheap). That way you can test if the clamping force is actually weakening over time, if the additional twisting to the same torque is actually increasing pressure or not, etc. Something else I'd like really like to see is some testing in the cold. I like the higher temperature tests you do. No one seems to test for the cold, even though many people print things that are meant for outdoors or to be kept in a car. How do prints with different filaments change in strength, flexibility, and brittleness at or below 0°C?
I got your point, and you are right. I have some pressure sensors, BUT it is not so easy to use them, I need two equal or one with hole. Place it simmetric etc. That's why I measure this torque, because it is in linear correlation with that force between two parts. About cold test, great idea, definitely I will test it, winter is comming, I hope I will have -10°C outside :-) And I am curious not only for creep test, but also tensile, layer adhesion etc.
Thanks for conducting the tests. I have a question regarding the torque test. Your data shows that ASA on day 2 achieve an a average angle of 66 and then on day 4 it’s 30 degrees and when heated it’s back up to 65. Actually it looks like your day 4 averages for all materials appears to be better. Any ideas on what the cause might be?
I think that's normal. If there wouldn't be heating, that rotation angle would be smaller and smaller after every test, since the material is more compressed (this was printed with infill, as I mentioned). With heating, this angle jumps, but then it would be again smaller and smaller after every tightening.
I modded my ender 3 to dual belted Z. Bottom belt holders were first from ASA and cracked a bit due to weak design (except the crack they were working fine). I reprinted buffied ones with PETG as I ran out of the ASA. I noticed that initially I had to spend more time leveling the bed than with ASA ones. Also now after few months I had to re-level the bed a bit. So at the next printer service I will replace the buffied PETG parts with buffied ABS ones.
At room temperature, I have experienced significant creep with PLA during a year. It permanently deformed until it broke, well under its tensile strength.
Would like to see each sample bent the same distance of deformation for a week, then see which one returns more closely to it's original shape. Repeat it with heat.
Great test, interesting results I love your screew test, very visualy apealing I think the Nylon (PA12?) gives a strange results, I dont belive it is only Nylon in that filament! I use mostly PETg in my prints At work it is PETg and some ASA (For outdoor use) Thanks for sharing your expirences with all of us :-)
Super video! Small improvement: when testing the retightening. Don't clamp both nuts at once. I think since this does not only lock the nut from rotating but also in vertical direction against the other nut.
It is important to lock only nuts, to prevent them from rotating. So all rotation is made is visible (and measurable) from the other side. If I would hold only one nut, 3d printed object may rotate too during testing.
Great test! All 3 the methods show similar results (though not always the same magnitude but that is to be expected). What isn't recorded in the state after taking the stress away. At the end of the video it is clearly visible how much elasticity is lost by PLA after just a short heat cycle. But I am surprised that Nylon is so poor and doesn't return to its original state. Polycarbonate is known to be very creep resistant but is brittle and sensitive to oils so would be a poor choice for your Voron build. PC-ABS might be the ideal material upping some properties of ABS (ASA) but not being so brittle as PC itself. That obviously leaves CF-PETG and CF-ABS or CF-ASA as options. The surely are stiffer and very creep resistant . Whether they don't become too brittle is a question.
Thank you sir, very informative video. I believe all the impact tests in the world are kinda useless compared to this because most functional printed parts are just holding something, silently deforming over time. For even more comprehensive creep test I would like to see how part thickness influences creep for a given material. (So, for example, maybe PETG creeps more than ASA in a 2mm thick part, but less in a 6mm part) Another test would be to measure twisting creep.
Twisting creep - great idea, thank you for the tip. Thickness effect to creep.. not sure if I will test that, more likely I will test is the correlation between thicknes and stress linear as with metals.
For the compression test, maybe the results would be more useable if aggregate result was calculated? That would show how much in total the bolts had to be tightened.
Yes, I agree, I noticed that later, now they are different type from first two. But, at least, excel table is downloadable, so it can be done by user too (*I will add this graph to downloadable excel)
In my V0, chamber temps are around 60C, I'd recommend setting that as an temp goal for testing, and I'd test over a 24hr period, as 2hrs at 45C doesn't compare with most of my prints. Good methodology though!
He's explained that 45°c is plenty to print ASA in an enclosure. So that's his goal. Sounds like you have a different setup. Maybe you should do the test yourself. Edit: Consider also if you have fume extraction your chamber temps aren't going to stay at 60c unless you have a heating element.
I appreciate this test a lot as creep is an oft overlooked issue when designing plastic parts. However, and feel free to discuss, I have the impression this test is incorrectly setup because it does not account for the amount of stress relative to the yield stress of the material that the samples experience. For instance PLA is really known to creep a lot but because it is also strong and stiff, it does not deform greatly in these tests, since you are loading it with a different percentage of its maximum allowable stress than the other materials, which could be loaded closer to their yield points leading to increases in creep rate. The total amount of creep deformation at a fixed percentage of the material yield point would be more indicative, as well as the total deformation after creep recovery.
Great test you done there. Very thorough . I thought why is ABS not there. Then later you said why. Well done. Did you think these results would have been any different?
Try to anneal PLA - grind table salt in coffee grinder, pack it around the PLA part. Put in oven at around 100 Celsius for 20 minutes. I tried it and PLA parts do become much stronger, though I didn't measure exactly by how much.
I'll try annealed PLA too (comparison with regular PLA and Nonoilen filament). But not sure if I will anneal it in salt, I know it will deform but I will see the dimensions.. still new to annealing..
@@MyTechFun In my experience part doesn't change dimensions if packed in fine ground salt. You don't need to remelt the plastic, 100 degrees Celsius (212 F) is enough.
Great video and glad to know you are building a Voron. I am building myself a 2.4 300mm spec also. For the 4th ABS in your tests I can highly recommend "3DJake niceABS" or "Formfutura TitanX". I have been using 3DJake niceABS for my Voron parts, it is really good! There is also the option for "Amazon Basics ABS" which is cheap and I think more difficult to print, as it is not modified (no ABS+, ABS plus, ABS Pro, ABS no warp etc) it is just regular ABS, but it might be surprisingly good in properties because it is not modified. Finally, I would be interested to see NonOilen compared to PLA, especially because of the results in this video for PLA after heating and your previous video where you boiled NonOilen. I bought NonOilen from 3DJake because of your video by the way.
Hm, that would be useful info, yes nonoilen resists high temperatures. So far, i was only thinking about nonoilen for food safe applications on my workplace (food engineering department). But from the other side, it is very expensive material, so I would prefer ASA instead (if no need for food safety). Oh, so many tests to do...waiting in a row.. pity that creep test is so slow process :-)
Thanks for the helpful information. I wonder how PLA would perform after heattreating. It probably wouldn´t spike that bad from heating but i wonder if the general creepresistance would go up or down
Usually PLA is annealed, never heard about nylon annealing. The problem I have with annealing, is that objects deform a lot, not good for mechanical parts
Nylon is semicrystalline and changes properties so much when annealed that manufacturers often times publish both datasheets, un-annealed and annealed. The reiforced versions (glass filled, carbon filled) can be annealed even at Tc without suffering dimensional inestability. For the unfilled versions, because the HDT is so high compared to Tg, the deformation annealing at Tg or a little bit above Tg is not too noticeable
Oh.. and when annealing at Tc, the time period is much shorter. You can place a natural color sample of the same polymer with a similar wall thickness next to the part you're annealing and see when the indicator sample gets opaque, that let's you know the if crystallinity has increased. You can get different degrees of crystallinity, but I haven't tested that too much.. just got to see different levels of opaqueness and subjectively percieved different properties, but yeah.. I can't assert anything in that regard
MyTechFun: did you post the follow-up video about different ABS's you're going to use for your voron ? i'm very interested in what you chose and tested there !
Not in one video, but I tested eSun ABS+, ASA, Polymaker in separate videos, but results are comparable with each other. www.mytechfun.com/videos/3dprint
Can you make a video about 3d printed hose fittings / tap fittings like double nipple or connector, and compare how strong the material againts water pressure. PLA, PETG, ASA, Nylon etc. 100% infill or 100% perimeter. Thank you.
I already printed some fittings for my fathers pesticide sprayer (1 bar) from PETG and it works now more than 2 years. But yes, this would be very interesting test.. great idea, thx.
I think for the screw test, the better visualization would be to show the running total angle, so day 1: 30°, day 2: you get another 50°, so plot 80° and so on
That was my first plan, but pointers are too big, they will hit the other bolt. But I need two screws to lock the position of the nut on the other side. I can redesign the graph, so it will show summary result. I'll upload this graph to my website, link in description.
screw clamping test shows the progressive creep, and how *at room temp* all samples had a high initial creep that then slowed down, but none were able to retain a clamping tension over time
Printed a filament guide in PLA+, but it warped to the point where it wasn't really doing its job - probably because I print in an enclosure with elevated temperatures. So I've reprinted it in PETG, hope that lasts a bit longer. If not, I might have to add a rib to the design to stiffen it. I don't really have the capability to print some of the other materials at the moment, mostly need to figure out ventilation.
I couldn't find any information about Vostok. Another Russian development, like Voron? (means East) *update, I found it, it will be Idex printer, but far for finished product v0.4
@My Tech Fun There are some youtube videos with 870 pla test,... it's look like "better" than abs, and obviously is easier to print. So, if we have a durable material and not hard to print, the non profesional printers (as me) have a good material to avoid the problems with other filaments. Děkuji :)
@@shvirik Thx. Sounds like the Nonoilen, which I already tested (only not the strength, but the temperature, it can be even boiled or heat up to 110°C)
@My Tech Fun Yeah ... it looks similar, but I won't use the water to make a tea :) The 870 also performs better with the heat: ua-cam.com/video/uMIfU4obfo8/v-deo.html (an example) Taking into mind that an 870 PLA could be bought for around € 22 / kg ... I think it is reasonable to use it for parts that should be durable without making your life difficult. (for printers noob as me).
So wait, everyone keeps telling me how bad PLA is at creep yet from skimming this video it seems better at creep than PETG which is what people keep telling me to switch to. Am I misunderstanding the results?
Wish some carbon fibre petg was in this, though I imagine petg will say something about this Also Polycarbonate as an alternative to ABS, due to the no fume printing.
There will be PC, PA, Carbon fibre etc materials. As I mentioned, this is just my first test to see the Creep specification for these basic materials. And also I am still searching for optimal creep test method..
Happy to hear there will be PC+ in future tests. I find it to be a very nice material to print and the temp resistance and strength is very good, but not sure about creep (until you inform us that is 👍).
At half way through, I might see one issue. I don't know the climate there, but I noticed the test pieces were sitting on the table in direct sunlight.
The screw test is extremely valuable, but there are 2 major issues with you current set up. First, the torque is way to high. This is the sort of torque you would apply to something like metal, not plastic. Second, the second torque is clearly being affected by the first. The screw holes need to have a division between them to absorb deformation - like the grooves put in concrete sidewalks. There are very good affordable dial indicators out there. In the US, Shars carries really nice ones that take a fair bit of abuse, and are accurate and repeatable. There should be an outfit that sells quality Chinesium stuff somewhere in your buying zone. Look for the places that hobby machinists get tools, not the places that Makers do (they are clueless about this sort of stuff). I would not buy digital unless you had some really compelling reason.
You are right about screw test, I'll insert a groove between. What should be the torque for M5 in plastic? I'll buy dial indicators and rebuild the test setup a little bit
in my opinion, there is no point in doing such long tests, test more: ABS vs ABS + vs ASA different PLA company from PLA + / HT-PLA ... and take the average for a given type of filament because there will definitely be differences and resins of various companies, ordinary vs transparent vs ABS-Like vs PLA (average) nice measurement after raising the temperature = ideal for use in, for example, a car
Nylon absorbs a lot of water from the air and becomes _lot_ more flexible. So basically your test with the Nylon was invalid, because you were measuring the change in elasticity and not creep.
A much neglected but important topic. Your methodology is sound, intuitive, logical and illustrative. Thank you.
What do you think about Delrin plastic about holding after weeks under load?
Thank you for creating content that is innovative instead of just copying what others have done. You are contributing to the knowledge of the printing community. I just began supporting you on Patreon. Thanks!
Thank you Gene, I appreciate that! More similar test are comming.. with different materials (PC, PA, ABS+ etc)
Aside from the actual content itself, which is very thorough and informative, I wanted to commend you for the way you presented it. Your video editing skills have greatly improved. I enjoy your use of time lapse, speeding up, and splicing shots together.
Excellent! Creeping is too often overseen in 3d-print material tests but has so much impact when dealing with plastics. Thank you!
Btw: in my experience the gap between PETG and ABS is truly noticable in real life parts, therefore I moved to ABS. But might depend on the brand.
The Voron community only uses ABS because of it's creep resistance
@@WhenDoesTheVideoActuallyStart Prusa printers use PETG. So, you tell me.
@@alejandroperez5368 And? ABS is still better at creep resistance than PETG.
Prusa uses PETG it's because they had the freedom to design a printer with parts that only required certain specific properties, and had the freedom to design both the models and the polymer blend to meet those needs. Besides, they're mostly there for aligning parts, almost none of them are load-bearing.
The Voron team doesn't sell printers, they can't hire factories to produce specific metal parts for their printers or fabricate their own filament blends, they're just a grouplet of hobbyists designing printers from parts that are available off-the-shelf, so they have far more limitations.
The Voron team had to design their printers around the minimum properties of common off-the-shelf filament blends and chose ABS, despite the eternal nagging by new people about the possibility of using polyamide or PETG - their printers simply won't perform well with those.
Fantastic video, I really appreciate the scientific methods for testing of the material properties.
A DTI gauge jig for all printed peices would be a massive time saver for future videos.
Keep doing what your doing,you've planned out and documented your finding like a pro.
You deserve more viewers to the channel,keep it scientific and the nerds will keep on subscribing.
The torque test was very useful,and the reletively low deformation temperature for PLA was eye opening.
Very much looking forward to the ABS results in your upcoming video.
Thanks for taking the time to experiment and document your findings for us.
The compression creep tests are fascinating. I had a functional print with a live hinge compressed around a pipe, and over time the PLA it was printed in migrated away from the bolt until it became impossible to tighten it anymore. I was hoping to find PETG or ASA had less creep under compression, but evidently not. I guess the best that can be done is to print the compressed area at 100% infill and use the largest washer I can fit in that area.
Or carve it out of wood...
In any case, your data and collection methods are really handy, thanks for the data!
I’m new to 3d printing and trying to decide what my go to filaments are going to be for differentiate cases. I’ve watched a couple of your videos in addition to those from other creators now and yours are far the best thought out, scientific and controlled. Subscribed and looking forward to new videos and watching your testing methodology refine over time.
One theory about the immense creep of the Nylon samples: as we all know, it takes up a lot of moisture (up to 10% in weight), which it does after printing as well. I recently stumbled upon a statement of a material science student, who investigated the mechanical properties after leaving the parts out in the open. He stated, that the elastic modulus was almost half compared to freshly printed ones.
So the creep of Nylon could actually also be down to moisturizing, maybe it would be interesting to see how „old“ parts would perform the same test
Would like to see: Nonoilen, Carbon fiber blends, "low odor" ABS (eg Novamaker), PC blends printable below 250 C (eg PC-PETG blend like zhoupu brand)
Amazing, thank you for testing this. Creep is something I've noticed for tight tolerance parts loosening over time.
Great video. Please continue to do these creep tests with other materials! There are so many interesting filaments out there and hardly any information on creep.
Awesome testing! I'd be interested to see how polycarbonate would do.
I just got some ePC, ePA, ABS+ from eSun, so there will be a polycarbonate test in very near future (not only creep test)
These type of tests is always good, especially if you like me like to print functional parts and build things like rc cars etc, so I like this
Thank you very much, very helpful and careful tests. I am printing a spring for a battery holder. The PLA spring deformed under the small tension of holding the battery. Based on your test I'll try PETG.
I was thinking about this when designing a part today. Keen to see the results
Thanks for the info! The screw test is extremely relevant since it's very similar to the kinds of loads some printed parts face when they're actually in use. I'm curious about whether annealing would help the nylon (or PLA) too.
Outstanding work! Clever, thorough tests. There are many channels with discussions about the physical properties of different filament types, but few conduct meaningful tests.
Thank you you Igor for an excellent and super helpful test. I have also found ASA to be most useful for outdoor parts that see heat but always wanted to see actual tests like yours.
I really like these additions to your tests. I'm all for tests that are more relatable to real world scenarios rather than ISO lab tests. For the torque test maybe you can include the total amount (added up) of degrees turned after 6 days. This would relate to "tightened additional 2 full turns".
Wow! This video is a treasure of information! Much respect and love.
Subscribed!
a valuable resource for the 3d printing community for years to come
Good information, I've only printed with PLA & PLA+ we use them for 3D printing guns and parts
Would be interesting to know the stress in the material in the tests. Is there a critical stress below which there is negligible creep?
I'm missing PACF and PCCF filaments. Those are commonly used in printers parts, and after some people complaining about PACF creeping, my choice could be PCCF.
I got some filaments from eSun for testing. ABS+, ePA and ePC (not CF), there will be a comparison video too. But carbon fiber tests will be included in near future. Only waiting for some hardened nozzles to arrive.
I also am very interested in the carbon fiber blend filaments. I look forward to the results.
Great job, one of the best videos I've seen so far. I would love to see the weight test outdoors on a hot summer day
It would be similar as in enclosure. And maybe you will find this one interesting too: not outside, but in car: ua-cam.com/video/QdAKd_YbsjI/v-deo.html
Perfect topic for some of my current projects ... thx :)
You not only have an excellent channel with excellent videos, but you also have a new subscriber.
I really like your experiment designs.
Wow, that's really professional! Your results correlate with the general opinion the engineers have at my work.
PA (or Nylon) has wildly different creep depending on ingredients, if everything fails, add glass fiber :)
Thanks for going through all that work, analyzing and documenting!
This should have more views.
Very much anticipating the ABS test.
This is awesome. Might I recommend a slightly different "screw" test?
Instead of testing how much you can rotate the screws at a given torque, a more in-machine relevant test might be to test the torque required to break the nuts free.
Procedure: Tighten nuts to a specified torque (using your high-repeatability click wrench), then check the maximum torque required to "break free" using your digital one.
If the material softens, there won't be as much pressure against the threads, and it should have a lower "break free" torque.
Hm, great point, if I understand correctly. Instead additional tightening the screw, more important is the torque for loose the screw. That also a realistic case too.
Perfect would be if I could measure the pressure between 3D printed object and the nut or washer for example. Imagine you use screw to join two parts, you want that pressure between two parts to be constant and strong. Now, that's hard to measure. What I can easily measure, is additional twisting the screw (which is in linear corelation with that pressure), as you can see in video.
Maybe at the end of the test, I could measure the unscrewing torque too.
This would be very difficult to get repeatable results since a great deal will depend on layer adhesion. You are describing a shearing test in essence.
@@MyTechFun the most relevant statistic for screws in plastic is how the clamping load decreases over time, this is usually what's speced out in injection molding plastics tech sheets. However this is sometimes hard to measure, requiring a load cell/strain gauge
Excellent job, very informative and entertaining to see, thank you and will be waiting for the ABS and other materials test. subscribed!
Great test. Just a question about the nylon. It looks like it was not dried before use, looks like you can see that it has bin bubbling when extrudes. You also did not mention if the nylon have had time to equalize with the humidity. Freshly printed nylon get softer over time cause it absorbs moisture.
Nylon was dried, but probably not enough. 4h at 70C (it is recommended 12h at 70C, but I thought that it stay dry, because it was dried before packing in vacuum bag)
@@MyTechFun i doubt it will affect the result much since it is tested over a period of time. It will absorb moisture and get soft. Directly after it was printed the piece that was printed with wet nylon would prolly lose.
Great test! I would like to see annealed PLA as well, together with Colorfabb XT or HT and Taulman Alloy 910.
This is something i needed!
I would like to add untreated PLA vs annealed PLA comparison to the test, i wonder if annealed PLA will perform better.
Also a material that often gets neglected in tests is HIPS. It is very inexpensive! It's supposed to shrink less than ABS in printing, and similar to ASA, and be more rigid and less smelly than ABS, though smell potential exists, and high shrinkage potential exists too. I think i would like to see more of it. I think there can be a massive difference between brands, as opposed to ASA or PETG or most non-silk PLAs, which is unfortunate.
I do like the C-shape test, in the absence of other tests, so if i had to choose just one test out of 3, i would make a decision based on it alone, and it's fine to me as-it-is; but i think screw test is interesting too, it looks like it will provide interesting and distinctive results eventually. I also actually liked all the tests though. Maybe if you use a piece of double sided tape to secure the nut, nylon can survive the 2-point test too, well enough to make a measurement, with enough room for it to bend down. But it also looks like the most fragile of the tests, with potentially needing a different thickness of the sample depending on modulus to be comparable.
45°C is an excellent choice for a hot box, as it's more or less representative for example also of use inside a car or inside a computer. Normally say you can expect the air in the computer chassis to become 40°C under load and SSD or hard disk can also reach 45°C. Currently i have a 3D printed SSD retainer in my PC and will probably have a 3D printed CPU cooler mount in my next PC. Ideally i won't have a 3D printed backplate but we'll see. I recall Stefan (CNC Kitchen) did have a 3D printed CPU cooler retainer and it failed on him, and it was i think CF Nylon? In hindsight, not surprising. He also experienced his Voron fail in exactly the same manner, with material creep under mild heat!
I actually don't terribly mind PLA being used for part cooling fan assemblies and such, because they aren't really load bearing, and also can survive some permanent deformation. But i do think PLA is less ideal.
Thanks for your detailed comment. Yes, any kind of nylon (CF nylon) have this creep problem, it is good in impact tests for example. 2 point bending test I will replace with 3 point, only I have to solve the problem of accurate reading (with dial indicator, more accurate).
And on my list is testing of annealed PLA vs PLA (and vs Nonoilen filament, because it is PLA type but heat resistant up to 110°C without annealing, tested earlier). For ABS testing for voron parts probably I have to go higher, up to 50C (accorting to response to my question in Voron group) and for longer time, not only 2 hrs. Mayabe 4-6h.
@@MyTechFun Thanks for the great video! I'd also be very interested in results for annealed PLA.
Nonoilen seems to be similar to extrudr GreenTEC Pro, maybe that would also be a good candidate for testing...
Nice Video. I would suggest to normalize the results to get a % value (eg. Platic defimation/lenth).
I have done similar measurements with another geometry. In my measurement PLA and PETG performes quite similar (20°C). But annihling the parts used to have an extrem impact on the creeping. It more than halves the creeping (also on PETG). The effect of the annihling was much much bigger than the difference between the materials. With Extrudr NX2 I also have found a PLA, wich dos not deform so much during the annihling.
Fantastic methodology. I really respect your attention to detail. I'd only suggest that your graphs be modified to show just the deformation for readability, ie. for test one subtract 12mm. Very informative.
Yes, I am using that method now in newest videos (deformation, difference between two days)
Great work & valuable results! Thank you! For C-ring test, maybe you could modify sample and/or fixture so top ear mounts to fixed bar. Then measure lower ear to bar (similar to cantilever flex test between sample & fixed table) Cheers!
I was waiting for a video just like this! Thanks
Very interesting, Egor.. Thank you. I don't have a 3D printer but I DO have a curious mind. That's all that was really needed to enjoy your vid. GR8T post editing by the way. That was a chor in itself. Cheers from So.Ca.USA, 3rd House On the Right.
I thought the screw test was useful and would happy to see if you can develop it a little more to get more meaningful numbers from it. Maybe have the screws hold 2 pieces together with a pressure sensor in between (Arduino/Raspberry Pi pressure pads are pretty cheap). That way you can test if the clamping force is actually weakening over time, if the additional twisting to the same torque is actually increasing pressure or not, etc.
Something else I'd like really like to see is some testing in the cold. I like the higher temperature tests you do. No one seems to test for the cold, even though many people print things that are meant for outdoors or to be kept in a car. How do prints with different filaments change in strength, flexibility, and brittleness at or below 0°C?
I got your point, and you are right. I have some pressure sensors, BUT it is not so easy to use them, I need two equal or one with hole. Place it simmetric etc. That's why I measure this torque, because it is in linear correlation with that force between two parts.
About cold test, great idea, definitely I will test it, winter is comming, I hope I will have -10°C outside :-) And I am curious not only for creep test, but also tensile, layer adhesion etc.
Thanks for conducting the tests. I have a question regarding the torque test. Your data shows that ASA on day 2 achieve an a average angle of 66 and then on day 4 it’s 30 degrees and when heated it’s back up to 65.
Actually it looks like your day 4 averages for all materials appears to be better. Any ideas on what the cause might be?
I think that's normal. If there wouldn't be heating, that rotation angle would be smaller and smaller after every test, since the material is more compressed (this was printed with infill, as I mentioned). With heating, this angle jumps, but then it would be again smaller and smaller after every tightening.
I really enjoy your investigations! Nicely done.
I modded my ender 3 to dual belted Z. Bottom belt holders were first from ASA and cracked a bit due to weak design (except the crack they were working fine). I reprinted buffied ones with PETG as I ran out of the ASA. I noticed that initially I had to spend more time leveling the bed than with ASA ones. Also now after few months I had to re-level the bed a bit. So at the next printer service I will replace the buffied PETG parts with buffied ABS ones.
Thanks for sharing your experience, very useful info.
At room temperature, I have experienced significant creep with PLA during a year. It permanently deformed until it broke, well under its tensile strength.
Very interesting. Maybe test ASA with 3 ABS materials to have comparison. I am looking forward for ABS as it is the only material I print.
Excellent videos as always Igor!
Would like to see each sample bent the same distance of deformation for a week, then see which one returns more closely to it's original shape. Repeat it with heat.
Something similar you can see in this video too, just before results summary.
Great test, interesting results
I love your screew test, very visualy apealing
I think the Nylon (PA12?) gives a strange results, I dont belive it is only Nylon in that filament!
I use mostly PETg in my prints
At work it is PETg and some ASA (For outdoor use)
Thanks for sharing your expirences with all of us :-)
There will be more testings in this topics. Few SLA comming soon, then eSun PC, PA, ABS+ then 4 ABS brands etc..
Super video! Small improvement: when testing the retightening. Don't clamp both nuts at once. I think since this does not only lock the nut from rotating but also in vertical direction against the other nut.
It is important to lock only nuts, to prevent them from rotating. So all rotation is made is visible (and measurable) from the other side. If I would hold only one nut, 3d printed object may rotate too during testing.
@@MyTechFun yes true! But I mean only one nut? Unsure
@@MyTechFun nevermind, with only one nut clamped, you could not messure the rotation since the part itself maybe turns as well
Great test!
All 3 the methods show similar results (though not always the same magnitude but that is to be expected). What isn't recorded in the state after taking the stress away. At the end of the video it is clearly visible how much elasticity is lost by PLA after just a short heat cycle. But I am surprised that Nylon is so poor and doesn't return to its original state.
Polycarbonate is known to be very creep resistant but is brittle and sensitive to oils so would be a poor choice for your Voron build. PC-ABS might be the ideal material upping some properties of ABS (ASA) but not being so brittle as PC itself.
That obviously leaves CF-PETG and CF-ABS or CF-ASA as options. The surely are stiffer and very creep resistant . Whether they don't become too brittle is a question.
I just got from esun ePA, ePC and ABS+ so this will be one of my next testings. Carbon fiber...yes but a little bit later..
Thank you sir, very informative video. I believe all the impact tests in the world are kinda useless compared to this because most functional printed parts are just holding something, silently deforming over time.
For even more comprehensive creep test I would like to see how part thickness influences creep for a given material. (So, for example, maybe PETG creeps more than ASA in a 2mm thick part, but less in a 6mm part)
Another test would be to measure twisting creep.
Twisting creep - great idea, thank you for the tip. Thickness effect to creep.. not sure if I will test that, more likely I will test is the correlation between thicknes and stress linear as with metals.
Amazing tests again, great job!
For the compression test, maybe the results would be more useable if aggregate result was calculated? That would show how much in total the bolts had to be tightened.
Yes, I agree, I noticed that later, now they are different type from first two. But, at least, excel table is downloadable, so it can be done by user too (*I will add this graph to downloadable excel)
In my V0, chamber temps are around 60C, I'd recommend setting that as an temp goal for testing, and I'd test over a 24hr period, as 2hrs at 45C doesn't compare with most of my prints. Good methodology though!
He's explained that 45°c is plenty to print ASA in an enclosure. So that's his goal. Sounds like you have a different setup. Maybe you should do the test yourself.
Edit: Consider also if you have fume extraction your chamber temps aren't going to stay at 60c unless you have a heating element.
I appreciate this test a lot as creep is an oft overlooked issue when designing plastic parts. However, and feel free to discuss, I have the impression this test is incorrectly setup because it does not account for the amount of stress relative to the yield stress of the material that the samples experience. For instance PLA is really known to creep a lot but because it is also strong and stiff, it does not deform greatly in these tests, since you are loading it with a different percentage of its maximum allowable stress than the other materials, which could be loaded closer to their yield points leading to increases in creep rate. The total amount of creep deformation at a fixed percentage of the material yield point would be more indicative, as well as the total deformation after creep recovery.
Great test you done there. Very thorough . I thought why is ABS not there. Then later you said why. Well done. Did you think these results would have been any different?
ABS is similar to ASA, but I am curious is there any difference between 4 most popular brands, that will be tested
Try to anneal PLA - grind table salt in coffee grinder, pack it around the PLA part. Put in oven at around 100 Celsius for 20 minutes. I tried it and PLA parts do become much stronger, though I didn't measure exactly by how much.
I'll try annealed PLA too (comparison with regular PLA and Nonoilen filament). But not sure if I will anneal it in salt, I know it will deform but I will see the dimensions.. still new to annealing..
@@MyTechFun In my experience part doesn't change dimensions if packed in fine ground salt. You don't need to remelt the plastic, 100 degrees Celsius (212 F) is enough.
Great video and glad to know you are building a Voron. I am building myself a 2.4 300mm spec also. For the 4th ABS in your tests I can highly recommend "3DJake niceABS" or "Formfutura TitanX". I have been using 3DJake niceABS for my Voron parts, it is really good! There is also the option for "Amazon Basics ABS" which is cheap and I think more difficult to print, as it is not modified (no ABS+, ABS plus, ABS Pro, ABS no warp etc) it is just regular ABS, but it might be surprisingly good in properties because it is not modified. Finally, I would be interested to see NonOilen compared to PLA, especially because of the results in this video for PLA after heating and your previous video where you boiled NonOilen. I bought NonOilen from 3DJake because of your video by the way.
Hm, that would be useful info, yes nonoilen resists high temperatures. So far, i was only thinking about nonoilen for food safe applications on my workplace (food engineering department). But from the other side, it is very expensive material, so I would prefer ASA instead (if no need for food safety). Oh, so many tests to do...waiting in a row.. pity that creep test is so slow process :-)
How much can we avoid creep by designing to lower stress? Maybe using more of a cheaper material is equal to less of a more expensive material.
I don't know yet, this is my first creep test. But a good subject for future testing.
@@MyTechFun Yeah, material is one thing and geometry another.
Thanks for the helpful information. I wonder how PLA would perform after heattreating. It probably wouldn´t spike that bad from heating but i wonder if the general creepresistance would go up or down
Really useful information!
Was the Nylon annealed? I've always wondered how much of a difference in creep of annealed vs un-annealed PA would there be
Usually PLA is annealed, never heard about nylon annealing. The problem I have with annealing, is that objects deform a lot, not good for mechanical parts
Nylon is semicrystalline and changes properties so much when annealed that manufacturers often times publish both datasheets, un-annealed and annealed.
The reiforced versions (glass filled, carbon filled) can be annealed even at Tc without suffering dimensional inestability. For the unfilled versions, because the HDT is so high compared to Tg, the deformation annealing at Tg or a little bit above Tg is not too noticeable
Oh.. and when annealing at Tc, the time period is much shorter. You can place a natural color sample of the same polymer with a similar wall thickness next to the part you're annealing and see when the indicator sample gets opaque, that let's you know the if crystallinity has increased. You can get different degrees of crystallinity, but I haven't tested that too much.. just got to see different levels of opaqueness and subjectively percieved different properties, but yeah.. I can't assert anything in that regard
MyTechFun: did you post the follow-up video about different ABS's you're going to use for your voron ? i'm very interested in what you chose and tested there !
Not in one video, but I tested eSun ABS+, ASA, Polymaker in separate videos, but results are comparable with each other. www.mytechfun.com/videos/3dprint
Could you do the same for a wide range of resins? It seems to be that all tough resins in the market creep like crazy
Hm. Good idea. I am curious now
Great video. What about CF filled materials, especially PETG+CF, Nylon+CF or Glass Filled ASA?
CF will be tested too
It would be nice to see results of PP, too
Awesome, thank you so much.
Can you make a video about 3d printed hose fittings / tap fittings like double nipple or connector, and compare how strong the material againts water pressure. PLA, PETG, ASA, Nylon etc. 100% infill or 100% perimeter. Thank you.
I already printed some fittings for my fathers pesticide sprayer (1 bar) from PETG and it works now more than 2 years. But yes, this would be very interesting test.. great idea, thx.
I think for the screw test, the better visualization would be to show the running total angle, so day 1: 30°, day 2: you get another 50°, so plot 80° and so on
That was my first plan, but pointers are too big, they will hit the other bolt. But I need two screws to lock the position of the nut on the other side. I can redesign the graph, so it will show summary result. I'll upload this graph to my website, link in description.
@@MyTechFun ah, I think you can just add it on the graph, you don't have to do it with the pointers
@@LightSpeedIII added, new excel table uploaded
screw clamping test shows the progressive creep, and how *at room temp* all samples had a high initial creep that then slowed down, but none were able to retain a clamping tension over time
Thanks!
Thank you! I am glad, you find this video useful!
Printed a filament guide in PLA+, but it warped to the point where it wasn't really doing its job - probably because I print in an enclosure with elevated temperatures. So I've reprinted it in PETG, hope that lasts a bit longer. If not, I might have to add a rib to the design to stiffen it. I don't really have the capability to print some of the other materials at the moment, mostly need to figure out ventilation.
Cool video. But please do not hurry with Voron, wait for Vostok, or at least look at RatRig V-Core 3 :)
I couldn't find any information about Vostok. Another Russian development, like Voron? (means East) *update, I found it, it will be Idex printer, but far for finished product v0.4
Great job!
Please, check for the next test a 870 PLA filament.
Hm, not the first comment who also suggested this. What is so special about this PLA?
@My Tech Fun There are some youtube videos with 870 pla test,... it's look like "better" than abs, and obviously is easier to print. So, if we have a durable material and not hard to print, the non profesional printers (as me) have a good material to avoid the problems with other filaments.
Děkuji :)
@@shvirik Thx. Sounds like the Nonoilen, which I already tested (only not the strength, but the temperature, it can be even boiled or heat up to 110°C)
@My Tech Fun Yeah ... it looks similar, but I won't use the water to make a tea :)
The 870 also performs better with the heat:
ua-cam.com/video/uMIfU4obfo8/v-deo.html (an example)
Taking into mind that an 870 PLA could be bought for around € 22 / kg ... I think it is reasonable to use it for parts that should be durable without making your life difficult. (for printers noob as me).
Good learning, Thanks
great information thank you
Excellent!
What about polycarbonate? PC in my experience is very elastic, but also rigid.
PC is tested in this video: ua-cam.com/video/caPEp_3uHoE/v-deo.html
Have you considered carbon parts or glass infused filaments?
Yes, sure, step-by-step. Carbon fiber filaments will be tested too.
I thought your aluminum clamp to capture the dimension was a great idea
Did you test this with Nylon 6 CF and Nylon 12 CF? I would love that test
Maybe here: ua-cam.com/video/cQb-hbr1KYY/v-deo.html&ab_channel=MyTechFun
So wait, everyone keeps telling me how bad PLA is at creep yet from skimming this video it seems better at creep than PETG which is what people keep telling me to switch to. Am I misunderstanding the results?
Wish some carbon fibre petg was in this, though I imagine petg will say something about this
Also Polycarbonate as an alternative to ABS, due to the no fume printing.
There will be PC, PA, Carbon fibre etc materials. As I mentioned, this is just my first test to see the Creep specification for these basic materials. And also I am still searching for optimal creep test method..
Happy to hear there will be PC+ in future tests. I find it to be a very nice material to print and the temp resistance and strength is very good, but not sure about creep (until you inform us that is 👍).
Incredible!
greate work! thanks!
It seems your Nylon was quite Wet in these tests, at least it appeared quite fuzzy to me
Yes, it was dried before placing in vacuum bag, so I dried it only few hrs before print, but looks like, was not enough
@@MyTechFun it probably would have been enough but nylon will "expire" in just a few hours in open air. somtimes even in the middle of a print
At half way through, I might see one issue. I don't know the climate there, but I noticed the test pieces were sitting on the table in direct sunlight.
Climate: inside 22-24°C. out approx 15-20°C day temperature. Effect of the sunlight is not too strong any more.
The screw test is extremely valuable, but there are 2 major issues with you current set up.
First, the torque is way to high. This is the sort of torque you would apply to something like metal, not plastic.
Second, the second torque is clearly being affected by the first. The screw holes need to have a division between them to absorb deformation - like the grooves put in concrete sidewalks.
There are very good affordable dial indicators out there. In the US, Shars carries really nice ones that take a fair bit of abuse, and are accurate and repeatable. There should be an outfit that sells quality Chinesium stuff somewhere in your buying zone. Look for the places that hobby machinists get tools, not the places that Makers do (they are clueless about this sort of stuff). I would not buy digital unless you had some really compelling reason.
You are right about screw test, I'll insert a groove between. What should be the torque for M5 in plastic? I'll buy dial indicators and rebuild the test setup a little bit
Appears ASA and PETG fared the best.
28:05 yellow =PLA orange = PETG blue = nylon white = ASA green = ???
Green is PLA but not creep tested. I was experimenting with dimensions on this filament as explained earlier.
That wasn't creepy at all...
in my opinion, there is no point in doing such long tests,
test more:
ABS vs ABS + vs ASA
different PLA company from PLA + / HT-PLA ... and take the average for a given type of filament because there will definitely be differences
and resins of various companies, ordinary vs transparent vs ABS-Like vs PLA (average)
nice measurement after raising the temperature = ideal for use in, for example, a car
Nylon absorbs a lot of water from the air and becomes _lot_ more flexible. So basically your test with the Nylon was invalid, because you were measuring the change in elasticity and not creep.
Igor, Prusament Carbon Fiber PC! That’s what I want to know!
CFPC from Prusament is easy to print and is really strong.
Hello Igor
Why Nylon for these tests...
Nylon is almost a semiflex material...
🤦
I included it as one of 4 basic 3D printing materials, for this first creep test.
oooof voron. I bought ratrig.
What's the problem with ratrig? That's good printer too
@@MyTechFun No problem. In fact I see benefits for ratrig that's why I got it.
Can someone tldr the results to me
Anneal the different materials and redo the tests.
I will, but I don't like annealing, because the object deforms a lot during this process. Not good for mechanical parts.