[newaccount74]

In my experience printing overhangs is rarely an issue.

The big issue that I have is that the orientation of the layers determines the strength of the part. Any vertical tension pulls the layers apart, so sometimes I wish I could somehow print a part in two orientations to make it stronger...

[gerdesj]

You can get some surprisingly strong prints. I have a plastic storage unit in my garden (don't judge me).

The doors are about 4'h x 5'w. The hinges broke on the left door a few years ago and I printed some more in PLA from a Thingieverse design (1). They have put up with quite a lot of weather (UK, South West). One did fail but not by layer separation - the hinges sheared diagonally across the layers, ie the "threads" snapped instead of separating. I printed another and a huge lump of plastic carries on being useful instead of land fill.

I generally use a gyroid infill but that is as far as I go off piste from the defaults. Prusa MK3S - self assembled, stock everything. Filament "Sunlu" ie Amazon random and it seems to work well.

If layer printing is an issue for you then you might want a different process or look into inter layer adhesion. I'm no expert on that but I did study Civ Eng at college so I have an idea about shear and moments n that. I generally print small parts where strength isn't an issue. If you need a reasonable challenge to test your gear, try printing a small tank and turret (2) pick one with a long barrel and print it with the barrel vertical.

(1) https://www.thingiverse.com/thing:2187167 (2) https://www.thingiverse.com/thing:3553160

[dekhn]

For outside stuff I recommend PETG over PLA. PETG is a bit stronger, doesn't melt in ambient temps (California, not UK), and takes much longer to degrade outside.

[mft_]

With respect, I think you're missing the point. No-one doubts that one can print objects which are strong, but it depends on being able to orient the layer lines such that the directions of relative strength and weakness are appropriate for the use case. This is a well-recognised weakness (no pun intended) of 3D printing, even with well-engineered and -set-up printers.

In the case of the hinges you linked to, if you were to set yourself the challenge of printing them standing upright on one thin edge, with the hinge barrel running horizontally, the layers would run parallel to the barrel and ultimately run vertically when attached to a standard door. This would make it likely that the hinges would break much more easily once the weight of a door was applied. Happily, this orientation would be crazy, and the obvious orientation (with the large flat areas flat on the print bed) is a strong one.

[naasking]

> but it depends on being able to orient the layer lines such that the directions of relative strength and weakness are appropriate for the use case.

I think this is sometimes a bit overblown though. CNC kitchen also demonstrated that you can achieve nearly 80% of the horizontal strength in the vertical direction, you just have to print HOT and SLOW. So "strength" profiles just have to be different than "fast printing" profiles.

[mft_]

Thanks for sharing the concept and the link - interesting read.

However, since the overall message (as regards strength) seems to be something like 'if you fine-tune your print settings to be unrealistically slow, you can mostly but not completely overcome the issue of lower strength due to layer adhesion' - I'm not sure it changes the argument, or makes much difference for most people, in most situations :)

[naasking]

I see it differently: when strength along all axes matters, you no longer need to take more time to carefully design and orient your print to overcome the strength limitations of FDM printing, or overbuild the part on the Z axis and waste filament. My design time is more important than printing time.

Most people are not printing non-stop or printing a functional part on a deadline, so I think it absolutely changes the argument when they need the extra strength.

[mft_]

Fair point :)

[newaccount74]

I tried to figure out where that figure was from, and spent about an hour or two watching CNC Kitchen videos on Youtube. What a ridiculously inefficient way of sharing test results!

I guess Youtube makes it easy to monetize your work, but as someone interested in results, this is rather frustrating.

[naasking]

Pretty sure it was this one, describing the properties of the "transparent glass" printing method:

https://www.cnckitchen.com/blog/transparent-fdm-3d-prints-ar...

So higher than normal flow, high temp, slow feed rate, and very little cooling showed remarkable strength along layer lines, up to 93% of horizontally printed specimens.

[newaccount74]

That blog post is even more frustrating than the video. No structure, you need to read a giant wall of text to figure out the result, at that point it's probably faster to just watch the video.

I hate that so much info about 3D printing requires watching hours of videos (even if the videos are just 15 minutes long, you'll end up watching a bunch of them because you don't know which ones contain the info you are looking for)

[zbrozek]

There are better and worse plastics for this, though you're right that it is always a consideration. TPU and PET self-adhere extremely well if you need strength in that axis.

[engineer_22]

Cnckitchen (original post) has done a lot of research into anisotropies of various plastics and settings, with surprising results. Check out his YouTube for more information

[progman32]

Or even alternate to create a weave.

[WillAdams]

Or just do 100% infill and bake in salt and re-melt to get a solid part?

https://hackaday.com/2020/09/23/reforming-3d-prints-with-sal...