Lab LogoFletch's Fab Academy 2014 Blog

11. Composites

This weeks assignment is to design and make a 3D mold (around 1ft square) and produce a fiber composite part in it.

So I decided to make some 'Gruffalo' monster feet type stilts for my 5 year old son.  We initially did some process tests as a group which I've described towards the end of this page.

Oli and the feet

11.01 Making the Monster Feet

After our process tests I decided that the best method was to keep a foam core inside the composite as this provided a very rigid final part that was most likely to survive 'abuse' from my kids.
As part of this assignment it seemed like a good time to start to learn a more 'artistic' 3D modelling package as most of my 3D work up to now has been done using Kokopelli, so I started to learn Blender.  I modeled a very poor monster foot in Blender by first creating a sequence of spheres that roughly matched the shape I wanted, squashing them, combining them into one single mesh and then using the 'sculpt' mode to create a more foot like object.
Foot in blender
I then exported the STL file form Blender and used PartWorks 3D to scale it to the maximum size that would fit on the foam offcuts we had available.  I ignored the fact that the object had a lower surface that the shopbot wouldn't be able to mill and just set the machine up to cut the top surface and then cut out around the object leaving tabs so that I had a thick foam foot with a flat underside.
STL file here (sorry it's nearly 800Kb!).

Foot model setup in partworksMilling foam foot

I machined two foam feet.  I used a spindle speed of 12000rpm, 9 inch/sec feed rate for the roughing pass, 4.5 inch/sec feed rate for the finishing pass.

I then proceeded to cover them with two layers of hessian each.  I decided to use the Smooth-Cast 310 as my resin because we had pretty much run out of West System Epoxy in the lab.

Before covering each foot I pushed a piece of 0.5 inch PVC tubing through the foam at about the mid point of the foot to leave a channel for the rope.  I cut some teeth into the end of the tube with a craft knife so that I could rotate it like a drill and get it through the foam easily.

The first foot I tried to cover with with single sheets with darts cut so that they folded to fit.  This was a difficult process as the Smooth-Cast material has a very low viscosity and doesn't provide much natural adhesion between the hessian and the foam.  It was difficult to keep all of the layers in place and we ended up using staples to hold layers whilst I tried to assemble it all.
Foam footFoot in vacbag

After a couple of hours at room temperature the part was cured enough to remove from the mold.  It was very rigid but required quite a bit of cleaning up of edges with a Dremmel and sanding to remove excess fabric that had 'escaped' in my messy assembly process.
Foot excess material

As the first foot proved so difficult to assemble I decided to try a slightly different method for the second one.  I cut the fabric into 1 to 2 inch wide strips and soaked them in a bath of Smooth-Cast before wrapping them around the foam mold.

Testing fabric wrappingSoaking fabric

Wrapping soaked stripsVac Bag

This second attempt was much easier to assemble without resorting to extra pairs of hands or staples.  Although the final result had about as much excess material that required removal.

After a bit of tidying up here are the final finished feet.  My 5 year old has promised me that he will paint them during his Easter holidays.

Final feet

11.02 Rubiks Snake Segments

One of my possible final project ideas is to make a self folding Rubiks Snake.  So i did a quick test making a few Rubiks Snake segments.  I used an open one part mold and vacuum bagging during the curing process.

Rubiks SnakeMy snake

I created a very simple test mold in Kokopelli with space for three snake segments.  After cutting the mold I tried to see if I could seal the small surface bubbles with spray paint, but this wasn't very successful.

Kokopelli moldSeal attempt

I measured and cut fabric to fit so that I had two layers all over the mold.  I then covered the surface of the mold in Vaseline to help release the parts before vacuum bagging all three segments at once.

Snake fabricSnake segments demold

Even with the Vaseline it required a bit of force to remove the parts from the mold and I suspect that I wouldn't be able to use the mold more than a few times.  After a quick clean up and sanding I drilled some holes in the center of the square faces and assembled my very short snake.


11.03 Process Tests / Experimentation

Composites don't seem to come up that often in our lab so we decided to experiment as a group with the processes involved before deciding exactly what to do for our individual work.
In the lab we had the following materials:

11.03.1 Test 1 - Open Mold

We found a large wooden object in the lab and decided to use it as an open mold, just layering sheets of hesian and using the West System 105 Epoxy.
Wooden moldLasercut fabric

As the object was almost a cylinder (it was actually a truncated cone) we lasercut the fabric to fit.  We decided to do 2 layers with a different cut pattern for each layer to produce overlaps.  We covered the wooden mold in cling film to stop the epoxy sticking to it.
We then worked the Epoxy into the fabric with an old paintbrush and layered it onto the mold, left over epoxy was poured ontop and worked into the fabric.
Working epoxy into fabricOpen mold

Final object de-molded.

When we initially de-molded the object it was still slightly flexible.  Over an additional 24 hours it became much more rigid.  Once it had fully hardened we all had a go at standing on it to test it's strength.  Not a particularly scientific test, but fun.
Strength test

What we learnt:

11.03.2 Test 2 - Vacuum Bag

For our second test we decided to mold around a foam core using the smooth-cast 310 and a vacuum bag.  We cut a simple block shape in the foam with a craft knife, layered a hand cut square of fabric with darts cut into it, poured on the mixed smooth-cast and used a wooden stirrer to spread it around.  The whole thing was rather messy, so we assembled it all on a piece of cardboard and inserted it 'cardboard and all' into the vacuum bag.
For the vacuum bags we used 'clothes storage' bags and a standard vacuum cleaner to generate our vacuum.  The vacuum had to be 'topped up' every 30 minutes or so as the bags were slightly leaky.
VacBag testVac Bag object trimmed
What we learnt:

11.03.3 Test 3 - Strength

We covered a cardboard tube from the center of a roll of paper towel to see if we could make a structural component.  We used two strips of hessian winding in opposite directions and Smooth-Cast 310.  The final object wasn't very pretty but it did take my weight.

Structural tubeTube testing

After a while it finally failed!  When we came back to this part the following day it felt much more rigid.  With more care and longer curing time I think we could make genuine structural parts out of these materials.

Failed tube

11.04 Conclusions

It's messy.
Care is required to make the final parts 'tidy' without the need for lots of post process cleaning of edges etc.


Oliver has painted his monster feet!
Painting the feet