People have been experimenting with 3D printed molds for fiberglass and carbon fiber for a while now, but these molds really aren’t much different from what you could produce with a normal CNC mill. 3D printing opens up a few more options for what you can build including parts that could never be made on any type of mill. The guys at E3D are experimenting with their new dissolvable filament to create incredible parts in carbon fiber.
For the last year, E3D has been playing around with their new soluble filament, Scaffold. This is the water-soluble support material we’ve all been waiting for: just throw it in a bucket of warm water and it disappears. The normal use case for this filament is as a support material, but for these experiments in composites, E3D are just printing whole objects, covering them in carbon fiber prepreg, vacuum bagging them, and allowing them to cure. Once the carbon fiber isn’t floppy and gooey, the support material is dissolved in water, leaving a perfect composite part.
E3D aren’t that experienced with composites, so they handed a bit of filament off to So3D for some additional experimentation. The most impressive part (in the title pic for this post) is a hollow twisted vase object. This would have required a six-part machined mold and would have cost thousands of dollars to fabricate. Additional experiments of embedding ABS parts inside the Scaffold mold were extremely successful.
As you would expect, there are limitations to this process. Since E3D are using a dissolvable mold, this is a one-time deal; you’re not going to be pulling multiple composite parts off a 3D printed mold like you would with a machined mold. Curing the parts in a very hot oven doesn’t work — Scaffold filament starts to sag around 60°C. Using prepreg is recommended over dry fabric and resin, but that seems to be due more to the skill of the person doing the layup rather than an issue with materials.
16 thoughts on “Soluble Molds For Composite Parts”
Epoxy curing will generate heat due to the chemical reaction. It doesn’t take a whole lot of epoxy mix to go beyond 60C (140F). http://epoxyworks.com/index.php/controlling-exotherm/
You also want to bake at a higher temperature than the operation temperature of your molded part.
I don’t think they reached the ‘operating’ part of the equation yet.
And carbon fiber is not usually a material you put in high heat use is it?
Plenty of motorbikes have carbon fibre mufflers, granted it’s probably only the outer layer but I still wouldn’t want to touch one for fear of being burned…
Exotherm depends strongly on the volume vs surface area of the epoxy as well as on the cure rate. A half-cup of fast epoxy in a cup will exotherm pretty quickly. I’ve had it happen in five minutes, melting the cup. The same material spread over a square foot will barely get warm. And a slow-curing epoxy doesn’t exotherm as much or as quickly.
Reminds me of the old children’s arts-and-crafts of putting paper dipped in wallpaper glue on a balloon and then when it’s dry you don’t need the balloon anymore.
That’s also a fairly common way of making hollow CF parts. A bladder is covered in a plastic sheet, then prepreg, then put into a mold and inflated.
I’ve also heard of forming styrofoam inner molds, then dissolving the mold out with gasoline/toluene/xylene – not always environmentally correct but effective.
Acetone is safer if I recall correctly.
Looks like a perfect tool for Formula 1 teams to build even more complex front wing structures.
They cute their components at 120c and at 60psi positive pressure a bit beyond 60 c!
60 c softening point is still a bit low fot this composite fabricator
Not all resins require temps that high to cure. You also have the option of post curing to a higher temp after separating the part from the mold.
I don’t have time to read through the whole thing but why is this stuff any better than already-available PVA?It desolves in water as well and works great.
This material apparently has less crosslinking when left in a hot nozzle so it doesn’t clog. It also claims to have improved adhesion and flow properties over standard PVA.
Original link broke, here’s a working link. The captions are broken too. Seems like e3d migrated their blog platform.
Thanks, updated the post.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)