Google reprappro conducting plastic experiments. Go to the github link they provide and examine the pdf file where they describe the methods and uniformity they achieve. It may save you some time.

An example of a process to mix plastic with a filler is described in the paper, "A Simple, Low-Cost Conductive Composite Material for 3D Printing of Electronic Sensors."

This is extremely helpful.  Many thanks.  Last night we discussed whether we should use a chemical method or just heat the plastic, mix the glass in, and then chop the plastic back down to pellets.  This should provide some needed insight.

I think we've decided on testing three concentrations, each pre-mixed in a different color of PLA.  We're hoping the different colors will make it easier to tell where the filaments need to be cut and which parts we should be testing.  We'll package up about 20-30 meters worth of each type, though we really only need about 3-4 meters of whichever ones extrude decently.  That should be enough to do whatever tests are needed and print a ton of scaffolds.

It's going to work better to mix the PLA and the glass chemically, and then to run the resulting plastic through the filastruder. This will create a more uniform mixture. An example of a process to mix plastic with a filler is described in the paper, "A Simple, Low-Cost Conductive Composite Material for 3D Printing of Electronic Sensors."  Google for it, it's available online. This silly forum won't let me link it.

In that paper they dissolve PCL (polycaprolactone) in DCM (dichloromethane.) PLA is also soluble in DCM, although using DCM on something that is supposed to dissolve inside of someone's body would probably cause the FDA some consternation. But we're talking proof of concept here, right?

One thing to consider if you're going to stick with a thermal based mixing approach would be to use ecorene, which is a powdered form of PLA. Using that as an input should help get a more even distribution of filler/plastic.

dubbsd wrote:

wouldn't the glass act like sandpaper causing metal contamination?

I don't think 20 micron glass is going to break off pieces of metal 10x bigger than itself.

Plus, I'm working something that would solve that problem altogether.

WHAT ARE YOU WORKING ON?!?!

Quit teasing us!

I doubt very much that they're going to put filament made on a $200 extruder, regardless of the barrel material, into the human body. The FDA will laugh that one out of the building!

This is meant for one-off prototypes and lab research, before paying a company megabucks for medical-grade stuff. For one off-prototypes, it'd work fine.

That's correct for this stage.  In fact, the glass we're intending to use for this test batch of filament is just your standard, run of the mill stuff you'd find in any bottle.  This is mostly to show that 3D printing is capable of making a scaffold at a tiny fraction of the current cost and piquing a bit of curiosity in it.  The lab spends a ton making and researching these things, and we're thinking the people who write the checks might be as fascinated as we are.  At that point, getting the right machines built becomes a little easier, as then this wouldn't just be out-of-pocket, on-the-side inquiry anymore.

This is meant for one-off prototypes and lab research, before paying a company megabucks for medical-grade stuff. For one off-prototypes, it'd work fine.

dubbsd wrote:

wouldn't the glass act like sandpaper causing metal contamination?

I don't think 20 micron glass is going to break off pieces of metal 10x bigger than itself.

Plus, I'm working something that would solve that problem altogether.

It will quite happily grind off pieces of comparable size though, and that could add up to a sizable amount of contaminant. Silicate powder affixed to a surface (here a flowing polymer) is basically 400 grit sandpaper.

Could you extrude through a glass nozzle though? You can drill through glass, though it's not easy, and that could eliminate the trickiest part where contamination could occur. Glass could handle extrusion temperatures as long as the ramp up and down is slow enough.

I don't think 20 micron glass is going to break off pieces of metal 10x bigger than itself.

Plus, I'm working something that would solve that problem altogether.

Worst case scenario some of the parts could be made in ceramic.  I'm reading now about ceramic augers, so apparently there is some precedent to work from, and I could probably get access to a shop that works with it.

Sounds like an interesting project you've got going.  Can't wait to see what you come up with!

wouldn't the glass act like sandpaper causing metal contamination?

I don't think 20 micron glass is going to break off pieces of metal 10x bigger than itself.

Plus, I'm working something that would solve that problem altogether.

wouldn't the glass act like sandpaper causing metal contamination?

That's a good point.  The materials used in the final extruder, if filament can be made, may need to be pretty unusual or at least coated.  Same would go for the printer parts in this case.  This really is turning into quite the adventure.

Thanks for bringing that up.  I'll look into it.

What does a scaffold look like? I am picturing a pretty small structure what precision would be required?

They're tiny grids with lines about 200 microns wide at most.  My account is too new to post a picture, I think, but here's a shortened, mangled url to a picture of a few:  h ttp://goo.gl/m2YYE

As you can see, they're just begging to be 3D printed.