New infill patterns in experimental Cura builds

[burtoogle 508]

I was asked to implement the Schwarz P and Schwarz D TPMS infill patterns by a user who wishes to compare their properties with the already implemented Gyroid pattern. This I have just done and so I'm just letting people know in case anyone wants to give them a go.

 

This is Schwarz P ("Primitive") and I don't think it is so good for general infill as it requires a lot of travels because each layer consists of a bunch of closed loops.

 

 

This is Schwarz D ("Diamond") and, like Gyroid, each layer consists of wavy lines so it requires fewer travels.

 

 

Both of these patterns have shallow slopes and so to get the best visual quality they should be printed using thin layers and fat lines (i.e. 0.1mm and 200% infill flow).

 

A known issue is that the Connect Infill Lines setting doesn't work very consistently with these patterns. More work required, there.

 

As ever, my Cura builds can be found at https://www.dropbox.com/sh/s43vqzmi4d2bqe2/AAADdYdSu9iwcKa0Knqgurm4a?dl=0 Please read the README.md file there.

 

For more info on TPMSs, a good page is http://facstaff.susqu.edu/brakke/evolver/examples/periodic/periodic.html

 

All feedback is welcome!

 

[Smithy 1,135]

As always, faster than the police allow, thank you Mark!

[Cuq 162]

Thanks a lot for your work @burtoogle, 

 

I  like your new Schwarz D  pattern.   The Gyroïde pattern is already nice but the Schwarz D pattern are faster and it generate less wobbling movements.  On the other hand I am not convinced by the Schwarz P(rimimitif) pattern.

 

 

Edited January 15, 2020 by Cuq

[burtoogle 508]

Hi @Cuq. Yes, the Schwarz P isn't really much good as a general infill but apparently it has potential applications in medicine and other niche areas. Personally, I will keep using gyroid as my day-to-day infill.

[Noisettetbou 1]

Hello,

I'm testing your last Cura master with Schwarz D infill and it's the best I've got tryied. Gyroid was good but this one is better for mechanical and really faster. And your others options are very great and convenient. I hope Ultimaker uses your work.

Thanks a lot !

[burtoogle 508]

Hi @Noisettetbou, thanks for the feedback. What kind of objects are you printing using the Schwarz D infill?

[Noisettetbou 1]

Hi,

I print mechanicals objects and I'm working on a project of blocks of construction for houses who needs this kind of infill. Anyway even if I print a character or a little piece, I would prefer use this kind of infill for his speed ans his strong.

[bagel-orb 61]

I've been studying triply periodic minimal surfaces (TPMS) quite a bit.
There are numerous other surfaces. This site has a very extensive list:

http://facstaff.susqu.edu/brakke/evolver/examples/periodic/periodic.html#gyroid

 

One important criterion is that they have no overhang, but even gyroid surface violates the overhang angle.

We might want to relax the criterion of being minimal, and alter the gyroid slightly in order to satisfy overhang constraints.

However, that would require some mathematical trickery which I haven't delved into yet.

Another idea is to rotate the patterns if there is one orientation in which there is considerably less overhang.

 

@Noisettetbou Please share any results you found. I am very interested.

 

It seems to me the Schwarz D is very anisotropic; does it have those straight diagonal lines only in one direction?

Edited February 13, 2020 by bagel-orb

[Noisettetbou 1]

@bagel-orb Hi. My results are that this infill is very fast and strong. The straight lines change of direction (rotation of 90°). I don't doing some pressure tests but I think is rigid like gyroid.
Note that with the Cura master version you can set the infill vertical scaling.

 

[bagel-orb 61]

According to my analysis the Schwartz D is exactly as printable as the gyroid, with normals concentrated around 45* overhang. The Schwartz D surface is definitely more stiff, though, seeing as it has straight lines in it.

 

Can't find an easy stl for the Schwartz P surface for my automated analysis. Perhaps I will have to generate it myself in MATLAB just like I did for the gyroid.

 

 

[bagel-orb 61]

I verified that the Schwartz P also has the same shape of the overhang angle occurrence plot.

 

Are there more TPMS which are defined by an implicit function?

[n2d7 0]

Hi, Sorry for being naive, but where can I find the code for Schwarz P infill implementation? I was going through the github files and couldn't find it. 
I would like to try implementing few custom infills, and was hoping to get direction from the your implementation. 

Many thanks!

[tinkergnome 924]

4 hours ago, n2d7 said:

where can I find the code for Schwarz P infill implementation?

 

Take a look at the "mb-testing" branch:

 

https://github.com/smartavionics/CuraEngine/tree/mb-testing/src/infill