One reason I hate to use 3D printed parts in my 3D printer

I checked on Son of MegaMax at the makerspace today after someone reported odd behavior. It took about 5 seconds to spot the problem:



Yeah, I know, the problem isn't the parts, it's my poor design...

That part will be replaced with an aluminum one within a week. Never again!

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Have you considered printing with stronger material? colorfabb XT or something similar. I know you have the 'all metal' mantra, but 3D printing technical parts can and should have its place

This was MG94 ABS.

The only remaining 3D printed part (other than some LED mounting brackets and and things of that nature) is the X axis belt tensioner. I'm going to step up the effort to replace that, now, before it, too, fails.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

What about pet or petg just started using it and the parts are very strong blows abs out of the water. Surface finish isn't as good and its quite stringy but mechanically it's very tough.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Fixed! A couple hours int he metal shop at the makerspace and SoM is back up and running fine.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I might not be so quick to call your original 3D printed bracket a "poor design". This doesn't look like a failure from normal wear and tear - fairly large chunks of plastic broke away, indicating that maybe it hit something pretty hard. Better to have an easily replaceable 3D printed part break than ending up with mangled metal.

Edited 1 time(s). Last edit at 07/02/2016 11:28AM by LoboCNC.

There's nothing for it to hit at the back of the machine (left side of the picture), except the motor mount (behind 1" foam insulation board), but if it had gone back that far, the bearing block would probably have come off the guide and spilled balls all over the place. Maybe not. There is a dent in the foam... Hmmm. Now I have to figure out why it went back that far.

There is a hard stop at the front, just beyond Y=0, but if it had hit there it would broken the coupler in the opposite direction and the damage would have looked quite different.

There used to be a hard stop at the back of the machine, but apparently I neglected to reinstall it after doing some work on the Y axis many months ago. That's going back in.

I'll think about a redesign of the printed part- busted plastic is probably better than twisted metal.

Edited 1 time(s). Last edit at 07/02/2016 12:27PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Longer screws would probably help with the original design - less chance of the part separating if the screws have more plastic to spread the load through. It does make you wonder if it's worth having endstops at all travel limits and always active.

That's a good point James. Having min and max endstops on all axes is easy enough but mainly they're disabled after homing for ABL purposes. Maybe a secondary set of "safety switches" which just activate the kill switch if hit, they can be just beyond the travel of the regular homing switches/sensors.

Or perhaps as was being said in another thread about head crashes, since most don't cause that much damage maybe its a low frequency low severity risk. Maybe more risky if the machine is loaned out/resides in a makerspace.

Edited 2 time(s). Last edit at 07/02/2016 02:46PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I'd be a lot more comfortable with more edge distance on those fasteners regardless of the materials used. Even better would be enough more edge distance to accommodate nuts and, ideally, washers. Tapped holes are nice for part count but they don't make a particularly strong joint.

Something definitely caused the carriage to move too far back- there's a perfect impression of the bearing block grease nipple in the foam board at the back of the machine.

I was thinking about the secondary kill switch thing. It seems like a good idea, but definitely a low frequency type problem, so worth the trouble? It took a couple hours to fabricate a new coupler after a few false starts, so a set of kill switches, at least on the Y axis with the beefy motor (425 oz-in hold torque) and drive screw, are probably in order.

I agree, the screws could/should (?) have been longer, and nuts and lockwashers could/should (?) have been added, or maybe threaded inserts... Plastic may be best for this part to prevent major damage in the event of failure. I'll have another go at the design and print a couple copies. If I add washers and nuts with the plastic, it may get strong enough to result in major damage in event of a failure. Maybe I got lucky that the design was just poor enough to fail in an unspectacular way.

Edited 1 time(s). Last edit at 07/02/2016 11:47PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

When E3d were testing their volcano nozzles, they found that parts printed with wider and thicker layers were very much stronger than "normal" layer heights. [e3d-online.com]. I haven't tried it myself but found it interesting and thought provoking. Don't know what layer height you used on the failed part but as it is a functional part and doesn't need to look pretty, then maybe thicker/taller layers might make it stronger? Just trying to be helpful but I expect I'll get roasted as usual for daring to make a suggestion.

Well sounds good to me, I try to print functional parts with 0.5 nozzles at 0.4 layer height. They are usually pretty strong, and print much faster. Been thinking about drilling out a nozzle to 1mm or similar and seeing what kind of parts I can produce.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
deckingman
When E3d were testing their volcano nozzles, they found that parts printed with wider and thicker layers were very much stronger than "normal" layer heights. [e3d-online.com]. I haven't tried it myself....

I did try it after buying a volcano and using the 0.8mm nozzle and they (E3D) were right. The printed parts were stronger. Specially the 100% infill parts are very strong.

Quote
DjDemonD
Well sounds good to me, I try to print functional parts with 0.5 nozzles at 0.4 layer height. They are usually pretty strong, and print much faster. Been thinking about drilling out a nozzle to 1mm or similar and seeing what kind of parts I can produce.

I've been doing a lot of research into this lately. It seems one big limiting factor is how fast the hot end can melt the filament and that is largely dictated by the time the filament spends in the hot end. A longer melt zone helps. One needs to determine the melt rate for the hot end in mm^3/sec then divide that by the nozzle area to determine what speed you can use. The bigger the nozzle, the slower speed you have to use so the time to print an object may not be any quicker than using a smaller nozzle and faster speed. However, there may be strength advantages due to better layer adhesion. I'm also told that T Glass can look better when printed at bigger, fatter layers but I have no personal experience of this.

Being into machining before additive manufacturing, I was a little wary of plastic parts myself, but having a simple cheap part designed to fail first is a great thing to save harder to manufacture and replace parts.

But if you want that plastic brace strong run the screws through to washers and nuts and use the compression of the plastic as its strength, rather than tensioning the bond between the printed layers which is creating as many 'fixtures' (attachment points that can fail) in the object as there are layers. Make the plastic part not the tension strength but instead the geometric guide for tensioning rods, m3 screws with washers and nuts. Sort of like earthquake retrofit in old buildings, reinforce through compression. It will then be able to break other things before it does, if that is your desire.