single head colored printing research

Hello
I bought a 3D printer (not RepRap but i can modify it still).
I have this idea:

A filament has portions of different color on it, joined together.
Some input sensor tells the machine what color is about to enter the extrouder right now.

An algorithm makes sure that the color that we want to see on the outside of our part is visible on the outside of the part.
Unwanted color portions of the filament are used to print inside the part where nobody will see them.

Maybe some means of mixing colors of a filament may be used inside an extrouder. At this point I have a very crude prtotype of the extrouder head and I contacted some manufacturers of the filament and asked them if the filament I am describing possible to manufacture. They said that it is but it will require some modefications to their equipement.

I turned my attention to those machines that can make filament at home from colored plastic pelets. Unfortunately there are no commercially available ones.
This may be the future of 3D printing.

I am a very poor student that had to stop going to college because I cannot afford to pay tuition.
There isn't much I can do.

If RepRap communit is interested in this reserach, please help.

Hey,
I don't seem to see a question but I'll participate in your discussion.
First of all check out :
[richrap.com]

Then what are you trying to do? you want to fuse say white-blue-red filament together? With a said given length? This is very shape dependent. I can't seem to see a good use for it as so many issues will rise.

There are commercially available DIY / open-source filament making machines, [www.filabot.com] for example.

Concerning the "algorithm" this can be a 3 year thesis project for a phD. I've assisted to a phD presentation this year close to that topic.

No. I am trying to make a printer that uses a single filament of varying color and prints the right color at the right time to make a surface of the plastic part a desired color.

There is a manufacturer that will make such filament if they see a reason to. At this point everything is experimental.

I have no way of doing everything myself.

tleneel Wrote:
-------------------------------------------------------
> Hey,
> I don't seem to see a question but I'll
> participate in your discussion.
> First of all check out :
> [richrap.com]
>
> Then what are you trying to do? you want to fuse
> say white-blue-red filament together? With a said
> given length? This is very shape dependent. I
> can't seem to see a good use for it as so many
> issues will rise.
>
> There are commercially available DIY / open-source
> filament making machines, [www.filabot.com]
> for example.
>
> Concerning the "algorithm" this can be a 3 year
> thesis project for a phD. I've assisted to a phD
> presentation this year close to that topic.

This is not so hard to do considering that it has to only be done once and than all printers will be able to do it. This is a part I can work on, actually, but I would not start unless I see a lot of public interest. I don't really need it myself. It is not the 1st priority in my survival as an individual.

Your idea is interesting but I think it's more of a problem than a solution. How to manufacture an evenly 'striped' filament? Because of the way filament is extruded, it would have to be dyed perhaps?

Much better to either improve upon dual extrusion methods and software, or concentrate on a color-mixing nozzle instead.

grumpenstein Wrote:
-------------------------------------------------------
> Your idea is interesting but I think it's more of
> a problem than a solution. How to manufacture an
> evenly 'striped' filament? Because of the way
> filament is extruded, it would have to be dyed
> perhaps?
>
> Much better to either improve upon dual extrusion
> methods and software, or concentrate on a
> color-mixing nozzle instead.

There is no need to extrude an evenly spaced color filament. There will have to be some kind of low-resolution camera or a light color detector to tell the machine where one color ends and another color begins.

I wrote an email to one of the Mexican filament manufacturers and found out that there may be a way to do this.
The machine that extrudes filament can be fed different batches of different color ABS or PLA pellets and extrude a filament that has portions of different color. Ideally, those color areas should vary every 40mm with a blurry transition zone of about 4mm.

Edited 1 time(s). Last edit at 09/18/2013 07:59PM by Tarakan.

When the plastic heats to the point of melting, won't the colors bleed and make it impossible to get accurate colors where you want them?

They will. But when a part is printed, there is inner volume of the part. So this mixed color plastic can be used on the inside of the part, while pure color can be used on the outside. Everything is simple and printers wont have to be modified. Only firmware and a new filament.

tleneel Wrote:
-------------------------------------------------------

>
> There are commercially available DIY / open-source
> filament making machines, [www.filabot.com]
> for example.

Don't forget the filastruder which has been shipping for a while and is about a 3rd of the price.

As for the rest of the topic. The major issue that I see is that this idea relies on the preconception on more plastic is used inside the part than on the perimeter. Say for example you have a 3 colour filament. Due to the transition zones, this means you are unlikely to have more than 25% of each solid colour available (and even this is doubtful given how long it seems to take to clear an old colour out of the print head when I change colours). So we are assuming that less than 25% of the plastic volume will be on the surface, but for many optimised parts! this may not be the case. Bearing in mind that if your surface is a light colour, then you will need to use this colour for multiple perimeters else's the dark colours inside will shine through.

C

It is not perfect by far but the idea is that the machine may be used almost without any hardware modifications. There will have to be some color change detector but thats it. No need to add any special extruders.

Look at dyeing filament, works good for PLA.
I'll cite him again, but look up richrap. He has done work at dyeing PLA filament which sounds like a better option to what you are trying to do.