Is there any existing technology that takes a variety of filaments of different colors (say a CMYK blend), and then blends them in different ratios to achieve whatever color you want?

I had this thought the either day, and I figure it's either stupid, already exists, or genius.

all 23 comments

sorted by: hot top controversial new old
[–] 9 points 2 months ago (13 children)

It's already a thing in orca and prusaslicer at least

  • source
  • hideshow 13 child comments
  • [–] [S] 4 points 2 months ago (12 children)

    Are you referring to Full-Spectrum? Because that's not really what I'm talking about.

  • source
  • parent
  • hideshow 12 child comments
  • [–] 6 points 2 months ago (11 children)
  • [–] [S] 6 points 2 months ago (10 children)

    That was an awesome read, but it does seem to be an evolution of the dithering in FullSpectrum.

    What I'm talking about is a print-head that would combine multiple melted filaments into blending hotend, so the extruded plastic would be a blend

  • source
  • parent
  • hideshow 10 child comments
  • [–] 8 points 2 months ago* (last edited 2 months ago) (4 children)

    That would be impossible to control. paper CMYK printers don't even do it like that. You'd need some sort of blending reservoir with active mixing to get a uniform color with that approach, and you can't make it continuous (you'd lose control of color) so you're basically making small individual batches of filament and flushing the mixing chamber between colors. you've now come full circle back to multicolor printing with one extruder like Bambu et.al. does now, but with more steps.

  • source
  • parent
  • hideshow 4 child comments
  • [–] [S] 1 point 2 months ago (3 children)

    You may be right, but I may attempt of to build some POCs to explore the concept, because I think the challenges are manageable

  • source
  • parent
  • hideshow 3 child comments
  • [–] 4 points 2 months ago*

    I'd mostly worry about controlling color transitions, because you can't really control the actual color properly when changing the mix for a different color. I think you'd end up needing some sort of purge between colors to avoid weird unwanted color transitions. I don't think it would provide lower material use or a speed benefit to the existing MMU multicolor solutions because of this.

  • source
  • parent
  • [–] 4 points 2 months ago (3 children)

    Filament won't mix in the nozzle, it is laminar flow. That's why the multi-color filaments still print multi-color instead of becoming a mix.

    Also, as someone else pointed out, color printers do not mix ink, they put tiny dots on the page - like pixels on a monitor. In fact, basically everything you look at today with color is using tiny dots of color and dithering to make it look like a different color. So, it sounds more like your problem here is that FDM printing does not have the resolution needed to make the dithering effect indistinguishable at the distance you would like. Which may be fair, but it is more likely that someone invent a way to significantly increase the resolution of FDM prints than it is that they would find a way to mix any color you want from CMYK filament "on the fly".

  • source
  • parent
  • hideshow 3 child comments
  • [–] [S] 1 point 2 months ago* (last edited 2 months ago) (2 children)

    Im aware that ink printers don't mix ink, but it's the resolution issue with FDM that makes me think filament blending is the way to go

    I've been contemplating how to blend the liquid (or near liquid) filament without introducing chaotic flow, which fortunately I have some experience with from previous projects, and in effect I believe that the filament stream could be twisted, it wouldn't bleed strictly laminar but can be done in a way that can avoid turbidity. It wouldn't be perfectly "blended", but with some uneducated guesses I estimate it could get the color resolution down to about 0.08mm.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 2 points 2 months ago

    I think the process is prohibitively complicated for filament FDM printers. From what I understand, mixing fluids that are in laminar flow would require a large working area which you don't have in a 3d printer. On the fly, from the printer, is SO MUCH more than what a 3d printer is already doing. It would basically be a color mixing chamber, at melting temperature, and then extruding back into a new filament at the correct diameter to run through the rest of the printer - may as well do that separately from the print process.

    However, I think this would work so much better without starting at filament and using pellets instead. There are pellet hopper FDM 3d printers out there, though not nearly as common as filament printers. But you could have a system of 4 hoppers of pellets that let out measured ratios into a primary hopper that agitates and mixes the pellets (while still solid) and then feeds them through print head like normal. Although, the normal size of pellets might be a bit big to get a good mix that isn't just a splotchy mess, once the hoppers and mixing is figured out you should be able to find a path to smaller pellets.

    That said, I do also like the idea someone else had of using a neutral color filament and then having a process of adding pigment somewhere along the way to the nozzle.

  • source
  • parent
  • [–] 4 points 2 months ago

    No idea if if ever took off, but e3d had a hotend that could mix 2 filaments. I've never seen one in person, but messed around with the chimera a bit when it was first launched.

    https://www.filastruder.com/products/chimera-cyclops?variant=1035850192

  • source
  • [–] 2 points 2 months ago

    It's not here yet that I know of. But I feel like it's going to get figured out in a couple of years. I daydream I had was not to blend but to have a filament that would change color with different temperatures.

  • source
  • [–] 2 points 2 months ago

    Binder jet printing is an older 3D printing technology in which adhesive is sprayed on layers of powder to build an object. A mix of ink can be included in the spray to produce detailed color combinations.

    https://en.wikipedia.org/wiki/Powder_bed_and_inkjet_head_3D_printing

  • source
  • [–] 2 points 2 months ago

    I feel like I definitellyyy saw a CMYK printer in a VOD of some printing convention a couple years ago? It was presented by some company and looked huge and expensive, wish I could remember who they were.

  • source
  • [–] 2 points 2 months ago (1 child)

    What I'd love to see even more than that is some kind of pigment cartridge system. You'd use natural, unpigmented filament and the printer would inject pigment in different mixtures as it prints. Just seems... I dunno. More efficient?

    I remember seeing a YouTube video (that I might search for and find again if I remember) where some company had developed (and probably patented) an approach that involved inkjet-coloring as it printed. I remember the colors looked pretty pale and washed-out looking. Not vibrant the way I would have hoped. But the approach might be one that could be improved on. 🤷

  • source
  • hideshow 1 child comment
  • [–] 2 points 2 months ago

    I remember seeing a YouTube video (that I might search for and find again if I remember) where some company had developed (and probably patented) an approach that involved inkjet-coloring as it printed.

    This is HP and their Multi Jet Fusion or MJF line.

  • source
  • parent
  • [–] 1 point 1 month ago

    It kinda depends on what your goal is. Is it making a custom filament? or are you thinking of blending colors on the go? And are you looking for commercially available solutions?

    If you're looking for commercial solutions, then your guess is as good as mine. I know you can get filament recycling systems, where you crush failed prints and supports before then extruding usable filament. But I think that's not what you're looking for.

    Going the tinker route instead, I think it was CNC kitchen that experimented with extruding their own filament by printing spiral of thick filament and then running it through a 1.75mm nozzle. They even got their logo embedded, and it was detectable in the finished prints layers.

    But if you're looking to blend on the go then maybe some manual g-code manipulation and custom hotends could work. But given that I asked about experiences with dual input hotend like the one below, and all the answers were along the lines of "don't even bother", I guess that you having to do a triple or quad input hotend won't make it more likely to succeed.

  • source
  • [–] 1 point 1 month ago (1 child)

    You would need some mixing chamber. That holds liquid plastic and churns it to achieve the results.

    Not to mention having all the filament the same plastic composition.

    Injection molds sometimes get stuck plastic in the hearing unit, but it doesn’t mix well and is streaky. Without some staging in a chamber it likely wouldn’t work.

    Even with clear filament and dyes, it would taking mixing to incorporate the dyes.

  • source
  • hideshow 1 child comment
  • [–] [S] 1 point 1 month ago

    I've done some experiments, and I've concluded that completely mixing the filament is unnecessary, and that with the rudimentary experiments I've been able to indicate I can get a color resolution less than 1mm within a reasonable nozzle size, no chamber necessary.

  • source
  • parent