Z-pinning to increase isotropy of part strength

[INDIVISUM]

The common problem 3d printed parts are that their tensile strength is anisotropic. i.e. parts are weaker in the Z direction than in the X/Y directions. Z-pinning increases the strength and toughness if parts in the z direction compared to not using Z-pinning. to quote the research paper (linked below): This initial investigation of z-pin printed structures with unreinforced PLA showed 20% increase in strength and 100% increase in toughness in the z-direction. The fracture surfaces of z-pinned samples were very rough, indicating that the pins successfully re-routed the crack front as it propagated between layers

As described in the research paper:
“The z-pinning process begins by printing a part in a conventional fashion that contains multiple voids in each layer that are intentionally aligned in the z-axis (Fig 1a). In the example shown, the voids are staggered in depth – one penetrating 6 layers deep, the other only 3 layers. Prior to deposition of the next layer, the print head is positioned over the deeper of the two holes and material is extruded until the void is filled (Fig 1b). The shallower of the two holes remains unfilled. Then the next 3 layers are conventionally printed, maintaining the aligned holes across layers (Fig 1c). The print head is then positioned over the deeper of the holes (different x-y location) and material is extruded in the z-direction to fill the void (Fig 1d). The process is repeated for the following 3 layers, allowing for deposition of a second pin into the same hole (x-y location) as the original pin (Fig 1e). Although this approach allows for deposition of continuous material across multiple layers, the z-pins have a finite length (6 layers in this example) which results in a “seam” at the interface between successive pins at a given location. The seam position can be varied among neighboring pins by staggering either the pin length or the starting layer of the pin.”

Please refer to this research paper:
https://www.osti.gov/biblio/1808415

[shakezilla19]

Brilliant idea!

Wish I knew how to code because I'd love to make this happen. Maybe final implementation could be an infill setting something like "Z-Pin Rectilinear".

The big problem appears to be how to tell the slicer where to add the pins or to make it automatic. I imagine the research group just sliced their model then hacked the gcode to have an action similar to an initial purge line which is pretty easy but very time consuming. A semi automatic approach seems to be the next step. Something like a "paint Z pin" feature in the slicer. The paint locations would then add a "purge line action" to the areas after every 5 layers, or so, and there is your Z-pin.

A basic Gcode to me would look like:
Prior Gcode....
G1 X, Y ; Move to Pin1
G1 E1 ;Extruder creates Z-pin,
G1 X, Y ; Move to Pin2
G1 E1 ;Extruder creates Z-pin
ect....

[foresterr]

Wow, awesome. That would be great for functional parts. You probably wouldn't want to have entire model using this (it would be practically 100% infill), but an additional "shell" of this structure right under the walls would strenghten the model without increasing use of the filament dramatically.

In layman terms, it's leaving small vertical holes in the model, and then pumping molten filament straight into them, right? I guess getting the amount and pin length right for all cases might be tricky.

[vgdh]

Hmm. But where the internal air would go when the upper of the bore sealed with the filament during z-pinning process.
It's need some experimental testing to check if that possible.

[TMRolle]

Maybe you could use U-shaped holes? Or just a small horizontal vent? May be a non issue too, most prints are not fully airtight, plus air is compressible, plus nozzles can build up a fairly large amount of pressure, on the order of a few hundred PSI according to this paper: https://pubs.aip.org/sor/jor/article/63/1/141/848523/In-line-rheological-monitoring-of-fused-deposition#_i27

[Anthony-Bec]

+1

[gundy360]

It would be interesting if you could have the z pinning right next to the shell and tie it into the rest of the part in an interlocking way. To more effectively tie into the rest of the part for the layers leading up to the pinning, you could do the regular infill with a small additional circle around the pinning site being printed amongst the infill, but not breaking it. That way when the pinning actually happens, the plastic interlocks around the sparse filaments in the pinning region further tying the layers together. The circular paths amongst the previous layers would just act to contain the z pinning plastic. If venting could be achieved by leaving a small space on the first layer of the feature to allow excess to escape into the infill region on the bottom.

[KestutisTauckela]

Thank you for sharing your idea, especially with images and such a detailed explanation. Impressive!
I had very similar idea with only small "enhancement": shape is a vertical dumbbell instead of a line.

Let's say for standard 0.4mm nozzle we leave 1.2mm disc of emptiness in first layer, 0.4mm disc in second layer, 0.4mm disk in third layer and again 1.2mm disc of emptiness in fourth layer. Then move nozzle directly above the empty dumbbell shape and fill it all with plastic in one go. In my theory, that should create a sort of rivet style solid structure, binding two middle layers firmly and somewhat reinforcing first and fourth layers. Whole structure does not need to be exactly 4 layers. It could be more (perhaps configurable via parameters). Not sure about 3 layer structures... Needs to be tested.

I really hope someone will find time to implement this. I'd like this idea to be free and benefit every Orca Slicer user.