Any graphs for layer adhesion? What is the difference in stress and/or elongation when you try to break it in either direction what is the impact resistance. Yeah chopped fiber parts are stiffer. But that is not what takes the cake all the time. What we need is a print off. Right now we are getting ready for an off season competition but maybe after that lets find some real world tests publish an STL and print it on the markforged and the S5 and I will do it on my Chiron and then have someone test them. If I am nasty I am suggesting the part I am printing right now cause I know it wont fit any of the other printers so lets find one that fits the smallest of the ones available. It definitely would be interesting. maybe something that has to be accurate to like some meshing gears as part of it etc. And then I still ask the question if you have a finite amount of money what are you better off with 1 5000 printer or 10 500 printers?
I think if you were to test with gears, using continuous fibers would make a huge difference in testing, and as far as sheer strength in the not layer direction though, onyx would take the cake. However, I think it’s interesting that you point out layer adhesion, which alone would be a more interesting test.
to the point of money, I’d agree you probably are better off with three prusas and a prusa mini than having only a markforged, but if you have the funds for 7 prusas or a markforged, 3 prusas and a prusa mini, I’d go for the latter option. Imo unless you are using 3d printing as your main method of fabrication, there will be a point in which you will end up having too much printers. I think for most teams, having 3 ish printers is around the optimum amount, and past that, you get less value out of each printer.
Or 31 $160 dollar printers
In our case 3DP is currently the main production method. Here is the 415mm tall item I just printed
Its HIPS and will hopefully lift our robot. its 415 mm tall 350 g of HIPS bought at 10/kg so 3.50 of plastic printed on our $400 Chiron. And it has been running turning out parts for close to 14kg of HIPS 2kg of PETG and about 2kg of PLA. Printed with a .8 nozzle sliced with prusa slicer .32 layer height .88 outside perimeter width up to 1.5mm inside perimeter width for extra layer adhesion and strength. So finding a balance between strength and accuracy. We reprinted this because the last one was a little too short… but here is a test with that one
The general wisdom from MF users is to have a 2nd printer that prints everything out of PLA at ‘draft’ settings in order to do rapid prototyping of fit/function. Then the final parts can be made out of Onyx. This speeds up iteration on parts without greatly increasing the cost of iteration.
I bought a pile of MPMDs (blame Ryan Dogneaux!) to do just that, and to print small stuff cheap.
Totally Agree. If it was up to me I’d choose ten $500 printers and the occasional 3DHUBS order to a Markforged printer when it is absolutely required.
I blame @AllenGregoryIV he started it
I believe that you know a ton about 3d printing and what you are doing looks like amazing work, however you continue to bring off topic data into this thread. I am not looking at how efficiently I can print or how many printers I can run at one time because a majority of out work is machined. This thread was created to discuss the properties and quality of different printers when working with carbon infused plastic because this is what we are looking at. We have tested some prints on our UM3 and the prints are decent but I am looking at the differences between what we can do currently and what something like and Onyx 1 can do. Thank you for all your input on 3d printing but this thread is not the place for it.
I love this idea and if we continue to work with our NylonX this will most likely be how I do it.
Then I wish you good luck with the Onyx 1
I run a HS engineering program and we have Raise3d N2 printers as our workhorse printers for student designs. Similar to you, I started down the path of printing MaterHackers NylonX on the Raise3d printers.
I was fortunate to receive a grant for a MarkForged Pro, which offers continuous strand fiberglass in the Onyx parts.
Are you interested in setting up an experiment comparing the MarkForged with NylonX? I am game for having a few students work on a project like this.
Continuous strand and Nylonx is not the needed comparison. Onyx vs. many of the other nylon carbon fiber filaments would be nice. There are already many lab tests published for the material strength properties. What the test do not tell is printability. The question that needs exploring is can a hobby sub 1000$ printer using the many alternative carbon nylon filaments match the printing performance of Onyx and the entree level Markforge printer. I’m not talking about small pulleys , gears and other small prints but large 3d plates that will see significant stress and abuse. Can a hobby printer match the performance on large parts. Parts that would fill the Markforge bed. The issue is warpage and layer adhesion on large parts. My experience is that yes you can. However, an open non-modified hobby printer can not. The hobby printer would have to have some type of enclosure to keep the whole print at an elevated temp. It has to print hot above 260C. This means all metal hot end. Thermistors are terrible above 250C. Should have a RTD. Bed adhesion is another issue. What bed material is best. Is there a Superior bed glue? As more and more teams get in to 3d printing structural parts for their robot, this becomes important because the buy in cost is substantial. The knowledge for printer mods is substantial also.
My take from last season and our large 3d plate printing experience is that “yes you can”. If you heavily modify a hobby printer. Even then you have get everything perfectly dialed in and keep it at that level consistently. This was not easy, but we did achieve it.
That brings me to this year’s off season project. We are redesigning our swerve module. Very similar to the killer bees effort. It is mostly printed. While the killer bees have almost all printed. Ours because we wanted large 4" wheels and it will be CVT is larger than the killer bees module. The students wanted to print the 2 main plates with carbon nylon. I more or less vetoed this because the time, effort and risk to go for it during build was to much. Instead we limited the design to 2.5D plates that will be routed. If we had a expensive Markforge printer would that have changed our decision? Could a Markforge print do these large plates? These are the questions to answer. A good test would be to see if a modified hobby printer could match the killer bees prints. That would be a good test project.
Everyone is correct in assuming that acquiring expensive printers will not be a magic bullet to success. 3D printers are just like CNC mills in that regard - they require some expertise in design to use.
Keep in mind that designing for 3D printing is a bit more involved than “just 3D print the plates”. For example, standoffs or other 3D shell structures could be integrated into the same face that holds bearings. This makes the overall assembly stronger and more rigid regardless of what material / printer is used.
A team I know is experimenting with some practice bot 2-NEO drivetrain gearboxes printed out of PLA on a Prusa. The idea seems preposterous at first, and yet their design flipped traditional gearbox design on its head and is nothing like the gearboxes we see in FRC today. I’m quite excited to see the results of the driving.
Any pics? Did they print just the plates or the gears too?
That sounds great!
I wouldn’t presume to pre-release their design for them, so the community will have to wait on it. They only printed the gearbox itself; everything else is COTS.
From my experience printing some parts out of NylonX on Prusa MK3 and MakerGear M2, I’ll say that one of the parts we made wasn’t turning out as strong as I would have hoped. I strongly suspect that a MarkForged part would have come out stronger. I was running it on a 0.6 mm nozzle in an attempt to get better layer adhesion, and while that seemed to work really well on the MakerGear, which seems to be able to crank out filament faster with the larger nozzle, I think I needed to run that hotter on the Prusa to get the desired result. In any case, this type of experimentation is exactly the point that MarkForged users make about being able to buy the thing and immediately print great parts. Some of our NylonX parts turned out great, but I have some homework to do before I feel super confident with that filament.
One of our mentors has a Pulse XE, and he likes the NylonX parts he makes. He makes some bigger stuff that he uses for car stuff that he’s very happy with. He has also had to do some experimenting to get it figured out.
I think larger nozzle sizes and layer widths and layer heights are a way to make NylonX parts as strong as the stuff that comes out of a MarkForged One. I don’t know if I’ll get to it this year, but I’d love to experiment with a 0.8 mm nozzle on a Volcano hot end to see how strong a part can get with respect to the same sized part on a 0.4 mm setup.
We ran PLA gearboxes on our drive and arm all year. They weren’t exactly light, but PLA is a surprisingly rigid material, and the gearboxes worked reliably. I don’t have a cleaned up model of the versions we ran, but the preseason versions are here:
IDK what the dimensions of the markforged are and I might be off topic but usually printing Plain Nylon is tougher to print than carbon filled as it tends to warp more. I have not printed carbon filled for the simple reason I have not seen the need. So here are some prints Printed with a .8 nozzle
The white is Bridge Nylon about 120mm in diameter
about 2kg of a mix of Bridge and Hobby king mystery (fishing line $20/kg) Nylon Biggest Nylon part is 186 millimeter diameter ring gear. Printed with a .8 nozzle (yellow is Hips 219mm max diameter
All the white(ish) is some kind of Nylon (bridge, 910, CX12) the rest is other materials (mostly PETG the light blue ABS) . This was a test platform The axle holders are about 245x215mm the longest of the straight pieces is 267x54mm The compound planetary is about 6 in in diameter. The compound planetary on this one is the only one that was printed with a .4 Nozzle
All the Nylon was printed on a prusa type printer with a garolite build plate and an E3D V6 hotend and brass nozzles except the .4 was a copper nozzle. Now How big is big? Now i know some of them were gears some small like the planets in the last planetary some huge like the rings in the big planetary and some 6 in tall like the compound planetaries in the big planetary. And some rectangular. Printing Nylon can be tricky but it can be done. Even on a hobbyist (cheap?) printer
With larger Nozzles you get to establish the max volumetric E to not print too fast depending on what material you print. And keep the layer height between 25% and 50% of the nozzle diameter and the layer width at about 150-200% nozzle width except the outside layer where you want it 110% nozzle diameter for best accuracy. So for a .8 that would be layer height between 0.2 and 0.4 outside perimeters at .88 and the remaining ones between 1.2 and 1.6mm use 0.2 layer height for accurate things like gears the rest as close as possible to .4 accuracy allowing. Here is a turorial on how to dial in the Max Vol. E.
or if you prefer I did a video a while back
Works on any filament some rough numbers in mm3/sec on a V6
Bridge Nylon 9-10
Other Nylones <=8 (the more flex the lower the speed)
On a Volcano about 2x above speed.
Speed depends on temp and brand. Higher temp will give better layer adhesion lower temp less stringing and cleanup. If you have the choice print at the higher layer height and slower speed as slower speed also improves adhesion.
Example 13mm3/sec PETG at .4 layer height an 1.5 layer width = (1x.4x1.5) .6mm3/mm extruded filament so 13/.6 = about 22mm/sec print speed. Now you can print it in the same time with .2 layer height and 44mm/sec print speed. The first will give you better layer adhesion. the 2nd will do better at overhangs and surfaces quality due to the lower layer height. With that on a prusa type printer you should be dialed in in about an hour - figuring on some short test prints. The toughest thing for me in switching to a .8 was the actually switching of the nozzle