It would be great to test not only "just a wall" but wall + infill + wall. For example how 2 mm "wall" with 2 walls and infill stack against 2mm just walls
I’d seen a test for this I believe from cnc kitchen, the infil mostly just holds up top layers and didn’t contribute as much to over all strength as you’d expect were the conclusions I remember from the video
Glad that someone finally addressed this along with infill in another video. I end up remixing and even redesigning many prints because even though the are exactly what I am looking for they tend to have very thick parts that are complete overkill for their purpose. making them up to 1/4 as thin of what the were originally still leaves them plenty strong.
If they use a plate to distribute the force the vertical walls will be under compression and the piece may sustain more force. I think pressing in the middle like a beam supported on both ends stresses more the part. I don't think the setup is so important as far as it is consistent for all tests.
Yes, failure was equivalent to taking a wall and bending it until it snapped. Rather different. And I'm here because I designed a load bearing part and looking for some data on what it could expect to support.
I've been printing panels of 2mm thickness to be used as lids for aquariums. It's good to know that I can shave it thinner and still expect a decent amount of strength.
Beyond this we need to look at infill orientation and line direction. I've found in some cases if I print with grid at 0 & 90 degrees I get very different strength than if I print the exact same model with the grid at 45 & 135 degrees. I have found that by changing the infill line orientation I can often reduce the infill %, use fewer walls and still get the strength I need.
@@Dongaz When the item is under compression grid infill with the infill lines vertical (ie like tubes standing on end) and running 45 and 135 degrees worked best. I used this to print off a set of bed post lift to raise a bed 8 inches higher (for an elderly friend). 30% infill and 5 wall and top/bottom layers and they've been in use for 6 months with no cracks or bending.
As I have already commented on the video on Infill patterns, I would still like to see these tests in compression along the z axis as viewed from the printbed, if your test equipment can handle it. Different materials, wall counts, infill patters, and infill percentages would also affect those tests, but we should expect higher strenght compared to the force being aplied in x/y.
I would love to see a wall thickness vs water tightness. Design a part that has several features. Like elbows corners etc and put water pressure inside until it fails.
This needed a second test set that used even pressure across the entire cube. especially on the last samples that failed due to the top wall failure. They likely could even double the results if pressure was applied across the entire top.
I like that you added the wood bottom to distribute the pressure better by the wood being able to deform. However, you still have lateral movement in your press. This looks like one from Harbor Freight. It's a good start, but needs some tweaks to make it a reliable and repeatable tool. Both moving cross beams, the one with the bottom of the ram and the one items get placed on, need to be better constrained. You can add roller bearing blocks and adjust those, or make them spring loaded. But, that is time consuming for a cheap tool. A quick fix is to add a sheet of UHMW to the inside of both uprights. Just push the beams to one side and measure the gap. Divide that in half and round it down to the nearest thickness you can get. Attach it with screws above the travel of the top beam, and a piece of double sided mounting tape at the bottom. That should give you some great results. Also, using the last 2 sets of holes at the bottom of the uprights, some 3/4 inch black iron T's, black iron pipe, and some all thread (from the electrical department at big orange box store), you can make front and rear cross bracing to pretension and stabilize the uprights so that the lateral force of the press trying to drift under load is not substantial.
This is an amazing test. For a lot of my functional parts, i will often do a 2mm wall thickness even if the part will not be under any intense load. Example a camera shroud printed with PETG and a 2mm wall thickness is virtually indestructible. Combine that with gyroid infill of about 20% and you have a winner
"yo momma so fat that she can break a square with 5mm walls" i'm so glad you guys are able to do much more scientific tests now that you've got the equipment for it!
Just a suggestion for future experiments - try comparing experimental results like in this video to theoretical strengths from FEA. You might want to switch to a stiffness measurement instead of failure load to enable a comparison like that. That would be interesting to compare FDM parts to theoretical isotropic designs.
I'm interested in the compression of the cube rotated so you test it how the layers were added. The print orientation, however that is described. The strongest possible orientation.
I would absolutely love to see a comparison of ASA creep performance. I know folks in the 3dprinting subreddit and other 3d printing communities would love to see this. Polycarbonate and PC with infill is another interesting one. I have a tie holder, for example, and after some time, the holder is beginning to sag with PLA. I would expect ASA to perform better, but I don't know. Creep is not a well documented by most manufacturers. Can you guys consider doing comparisons of various ASA colors and brands? I think there is a strong community interest in eSun, Polymaker, and Hatchbox brands. I would guess black, white, and grey would all be great candidates. Green, blue, red, orange, if you wanna get fancy.
Good video but I would be interested to see where the failure occurred on each test. Was it always in the same place or changed as the thickness went up.
Heh. Saw this pop up in my feed about an hour ago and was sighing because of the effect/affect confusion in the title - this is one of my pet peeves and it drives me crazy. Especially because in *my* accent they sound completely different. However, clicking into the video just now, I see it's been fixed. Thumbs up for whichever editor caught the mistake!
Great video, and good data. Suggestions: Please show the failed parts. (at least one representative sample) A slo-mo of part failure would provide insights into a how part fails. This could be useful as a "how to design for" type discussion. Did part have: clean break, sheer area, distortion, layer separation, etc. While the thin walls failed before the bridge gave way, at some point a thick-walled variation was stronger than the bridge. Unknown is how much the distortion of thin walls contributed to earlier failures. Think a thicker bridge, even on thin-walled variations would have provided more consistent wall focused test data. (ie: this was both a compression test and deformation test) If it was the bridge area that was being tested, these tests would be mostly test failures. ;) Testing like in video is ok for simple reference, and basic education, but less usable if building a reference data set.
Awesome video. I’d like to see an updated video on waterproof 3d printing. There is different info out there on material choices, over extrusion, wall thickness & more on best practice. However, I haven’t seen any comparison. I imagine this would also be a near zero cost in the test equipment required. Anyway thanks for all the info & hard work you put into these videos
perhaps same test but using large flat steel plate in between to isolate strenght of material from point load? and bonus round -> print with settings you think is the strongest per unit of weight
Newton is the correct unit. To get kg on earth Newton/9.8. pounds is a measurement of force, change the gravity and the scale changes. This is why a 360lbs man weights 60lbs on the moon. But will always have a mass of 163kg.
Could we see this test repeated, but on longer timescales, and under varying temperatures? I'm really interested to see how each wall thickness holds up under load over time, and when exposed to high and low temperatures, as well as thermal shock.
This is very useful information. I want to make replacement parts for laptops where spares aren't readily available. How do 3D printed parts handle high temperatures (e.g. 100°C)? Some laptop casings get brittle and fall apart near the heatsink over time due to the heat.
Excellent videos about wall thickness and infill. Could you tell us about strategies to increase stiffness? I tried to design 3D-printed lightweight drone frames and found that there is probably a limit in size (5 inch drone) caused by low stiffness. Low stiffnes leads to vibrations which makes the drone unusable for filming or even causes heavy oscillations confusing the flight controller.
I always wanted to know which filament is the stiffest, or which one has the highest hardness on a scale, and not just PLA, but specific types of PLA and so on.
When you print a 1mm wall what nozzle have you used, 0.5? I usually design wall thickness as multiples of nozzle withs, so 0.8 or 1.2 would be the the closest to 1 mm. Is it me overthinking the problem or do you print with different nozzle size?
Nice video, I would have one request for future videos. Could you also add the weight on Kg. It is hard to think in mm together with Lb for us that do not live in the USA.
would be great to see the differences with different print orientation, and also test other forces: tension, torsion, shear (in x and y), fatigue. look up orthogonal arrays to reduce number of tests you do but still get the same results. there's a good explanation of this by NightHawkInLight (Multivariate Experimental Design)
Considering straight blowback weapons exist that use printed bodies and operate reliably for hundreds or even thousands of rounds it's no surprise 3D printing can be quite structurally sound if designed for such.
Just a quibble about the "piano" comment - upright pianos usually weigh about 450 lbs, and grands up to around 7 feet weigh up to 750+ pounds. My family owns a piano store 🙂
would like to see this test at 10C, 20C, 30C, 40C, 50C, so it can be applied in real world applications. 50C is an engine room of a small power station etc. 40C is top of the environmental range 30C is about the range most items would be in. 20C is the standard dimensional calibration laboratory temperature, which means this is the temperature to verify the dimensions of the test objects. 10C because no one tests the strength of plastics going down in temperature, only up. 0C and below would be kewl too lol
Most of the world uses the metric measurement system. Please could you give stats in metric? I have no comprehension of "pounds" other than British currency.
Useful data but I'm looking forward to the data from the crowd sourced tension compression tester that won't have the artifacts from hand pumping a hydraulic bottle jack. Fund the Kickstarter, everybody!
I like you vids but your strength testing approach is driving my OCD nuts. Your piston is slipping sideways making shear the cause of the failure. Make a bracket/sleeve to maintain the direction of force exerted by the piston.
I am guessing that the infill is zero percent and then the walls dense. With 0.4mm nozzle then 3mm all would be almost 10 lines wide ? That is the setup?
Unfortunately, most of my 3D printed parts that have failed, don't fail with compression, rather the layer lines fail. I would like to see the opposite test, where you pull 3D printed parts apart, and what would be the best way to strengthen layer line separation and shearing. Anyhow, thanks for the video, great information from tests like these.
Not sure what your name is so I'm going to call you Bob. Can I call you Bob? Ok, Bob, question: Why didn't you put any effort into showing a relationship between the results? Does the strength grow linearly, or logarithmicly, or does it diminish after a certain thickness? WHAT is happening!?!? If you would have just tried, you could have seen how beautifully the results line up linearly. Almost like if they were made up. They have an R^2 = 0.9975 To know what the expected strength (in this case) of the cube would be is Strength = 1388.9*(wall thickness) -1079.3 Where "wall thickness" = 1, 2, 3, 4, 5. So at 1.5mm = 1004 N, 2.5mm = 2393 N, and so on... Come on, you spent all this effort in getting the data, but no effort analyzing it.
Congratulations for the channel, we learn a lot from your videos! I haven't seen in your videos the testing equipment that you have. Do you have a video showing it? Why do the graphs have those saw tooth form? Is it coming from your pressing device? Are you using a manual press? I think is not good applying the force in this "apply and release" way. Pressing with a constant speed would be better because it may allow you to see how the load changes before breaking meaning the structure is failing and changing rigidity. Something similar to what you see in this video: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-eewlYa6IQPg.htmlsi=tnBUrnQuI5eCG9dX&t=377 If pressing with a constant speed is not possible, at least you have to maintain the force applied. Maybe using a hydraulic manual press with a good check valve?
@@slant3d Confusion is likely related to using N neutron's which is force, while referencing objects of known mass. Generally to most people, lb (pound) is a unit of mass, not force. Problem is "lb" is often used interchangeably to imply weight and mass; at times incorrectly. lbf is pounds of force, while lbm is pounds of mass. It's an imperial pun, to confuse a mass of "slugs" with their weight in "stones". 😉 In the video using objects of known "mass" was a great general reference.
