Hey Guys

My name is Jan and for the last few months I was working on a new open-source 3D printing slicer/modeler called Chisel, geared towards high-speed printing of lightweight efficient 3D structures.
It works quite differently from traditional slicers which get STL triangulated mesh as input. Unlike those, this slicer works directly with parametric CAD primitives (bezier/nurbs curves) and requires bit more information from user to generate gcode instruction.

To generate a gcode file, you need to supply 2 front & back surface patches + infill-options and the software will automatically generate a strength optimised lightweight panel.

Why did I use such more complicated approach instead of just using any other slicer with adjusted infill type & density to suit your specific needs ?

One reason is that due to common slicers being so generic, they lack crucial information about model in order to generate truly optimised infill geometry - for example when considering object with thin long beam cross-section, they don't know which sides of the perimeter are top/bottom in order to generate triangulated truss structure of infill.

Another advantage of being more narrow focused and having more information about object being sliced is that it enables fine tuning of things like wall-line width, infill-line-width (separately for any type of infill) and minimising travel moves resulting in very fast prints -> it's always just a single perimeter followed by infill pass/passes.

For example lightweight core panel which will serve as composite core later will have very thin perimeter, thinnest possible to still print without holes/defects as it will be later laminated with composite (glass/kevlar/carbon) skins which are much, much stiffer & stronger then any plastic.

On the other hand a standalone panel which won't be reinforced afterwards will get much thicker perimeter line as it will be the primary load bearing component of the object.
All infill types can be freely combined, confined to just specified intervals within the panel, etc.

Last but not least, unlike most other slicers, Chisel is very fast, enabling slicing very tall and detailed objects with ease - a 1000mm high object sliced with 0.2mm layer height and high-resolution smooth curved skin won't crash your slicer or take 20 minutes to finish slicing.

Roadmap of soon to be implemented additional features

* Batch printing of multiple panels at the same time
* Automatic generation of skirt/brim
* Automatic generation of support structures to tie together tall objects/close c-shaped profiles
* Non-planar printing
* Import of bezier/nurbs curves/patches from other programs, eq automatically extracting them from .step CAD files, etc.
* Automatic generation of resin infusion channels on core skin/skins, requires really good tuned printer for core to stay watertight, but my preliminary testing showed it's indeed possible, in a worst case with simple post processing step (sealing of the composite core).

You can even combine best of both worlds - Chisel supports cutting patches in any direction, primary to support printing of very high objects in multiple parts, but you can even export specified section of your object and further process it in your favourite CAD tool (adding holes, etc.), export to STL and then slice it with a standard 3D slicer, while most of the object will be sliced with Chisel.

Some examples of panel infills possible with Chisel:

Corrugated infill panels, showcases Warren truss like structure


Cross-corrugated panel, good for panels with single outer skin


Sine corrugated infill structure, enables engineering more compressive compliance into panel


Sine modulation of corrugations in Z direction, for increasing compressive stiffness


Tube with corrugations oriented optimally to resist crushing forces, enables printing very light and still stiff tubular cores (that's how I made SUP paddle shaft core)


Pictures of some finished composite product with 3D printed cores:

Small canoe for children, 2.2M length, 1800g complete weight and very stiff, fiberglass skins + balsa wood trim


SUP paddle, 440g weight and comparable shaft-paddle stiffness to paddle of top brands made with foam cores


Snowsled for children, 0.8m long and printed in single piece, fiberglass skins


This one is not yet finished, but my biggest project so far, 5.5m long prototype of very fast, fin stabilised paddle craft:


I hope some of you will find this useful as well, the project is hosted on github and the only way to use ti currently is directly via programming language API as shown in the readme, but I'm working on more user friendly interface/support of importing parametric models from CAD files (STEP, IGS...)

Welcome to the forum and good luck with the 3D printing software. Its a growing market and I can imagine 3D printed cores could be very useful in some complex shaped objects.  Be interesting to see how the 3D printed cores compare mechanically to a "typical" foam core material or similar you might see being commonly used in composites. 

Warren Penalver
Easy Composites / Carbon Mods - Technical Support Assistant

Thanks Warren !
I did some preliminary testing (shear & compressive strength) on core samples printed with PLA, which is pretty stiff plastic (~3.6GPA flexural modulus), but nothing extraordinary (it also has quite poor temperature resistance & resistance to impacts):
3D printe core samples

For core with density of around 110kg/m3, compressive & shear properties were comparable to AIREX T10 foam, I have not yet tested samples of much higher performing materials/filaments (glass filled polypropylene and glass/carbon filled polyamide), but I'm very curious about how much better they will perform.

There are quite a few different filament materials you could experiment to get the performance you want.  

I think the fact you can do complex shaped cores relatively easy once drawn up in CAD is an advantage in its own right compared to traditional flat core materials. Sure its not necessarily going to be world changing for composites but as 3D printing advances there will certainly be a place for such core materials in specific composites processes where other standard cores just aren't suitable.

Warren Penalver
Easy Composites / Carbon Mods - Technical Support Assistant