Ideally, you’d want a spiral filling mode here (a kind of vase mode). But this doesn’t actually solve the isotropy issue since even with this setting, the printer will just stop filling the cylinder for each layer (to continue printing everything else on that layer). However, the idea in this article is, IMHO, stupid. There’s no additional strength to be gained from more than 2 layers (maybe 3). Worse, because it’s injecting plastic in a very lightweight container, it makes the whole part weaker: the infill wall aren’t made to support injection pressure and the wall layer will simply split in that case.

When the printer goes over an existing layer of plastic, there’s a remelting action that gives strength to the bounding with the previous layer and the new layer. Here, since the injected plastic (which is hot, but clearly cools quickly), will contact a very cold layer of plastic (the infill bottom), no such remelting will happen, you’re just layering 2 different material here. You’d get the same results, IMHO, to simply embedding plastic nails in your part, like his tests on impact resistance have shown.

Moreover, the hot mess of the injected plastic here has a strong thermal gradient, so its exterior skin cools even faster when it contacts the infill layer while the interior content stays molten. So it won’t really fuse with the infill’s container, unless pressure is increased which, in turns breaks infill’s own interlayer adherence.

It would be better, IMHO, to simply fill only one or 2 layers of plastic instead, so that the plastic that’s injected is hot enough to actually remelt with the cold “previous” layer plastic. Typically, you’d want to inject the equivalent to 2 (or 3) lines into a 2 layer deep hole while still following the infill’s wall so that the plastic hot section is always in contact to some plastic instead of hoping for the best here. You’ll fill with a spiral going inward (from the infill’s wall container to the center of the cylinder) so that the exterior skin is always in contact to a hot or cold region.

I don’t have an impact resistance test bench so I can’t prove this, but that seems intuitive to me.

I would imagine the print time would be quicker his way but yeh who knew solid plastic is stronger than hollow plastic

Also I think we need to look at non-plastic and cheap alternative fillers. Something as simple as sawdust and wood glue or paper pulp and a glue or even coffee grounds (I remember some people trying them as filament filler).

FTFY.

The Hackaday summary doesn’t really delve into it, but these aren’t about increasing strength as such, they’re about making parts more isotropic. FDM prints typically suffer a lot from having a weak axis due to being composed of a series of generally pretty strong layers that are relatively weakly bound together – all of these strategies are aimed at automatically creating bridges between those layers (z pinning in particular is literally about pinning the layers together) to mitigate that. There are ways to make parts much stronger than this of course but these strategies could be incorporated into slicers since they just use the extruder that’s already laying down the part, albeit potentially with a modified hot end in this case.

I was scrolling looking for this. All the strength of uv resin and not nearly as much heat.

So, this could be part of the “make it strong” profile on a slicer”, or you could make it stronger by….

Having a stock of pins/dowels in the right size, remembering to use them, being capable of using the slicer or other CAD to put in a properly dimensioned hole, and installing it post print without cracking the part.

I wonder which one has better scaling a user friendliness?