Skip to content

Wildcard Week (aka resins, epoxies and composites)


We are going to use sheets of foam (PIR), cut them on the shopbot and then layer them with fibres that we glue together epoxy. There are some limits to the millingbit and size, but the milling goes very fast. so that is great.

To speed things up: make the rough cut somewhat more precise then you normally would. And don't do a smooth cut: you can sand the roughness away by hand.


Design and produce something with a digital fabrication process (incorporating computer-aided design and manufacturing) not covered in another assignment, documenting the requirements that your assignment meets, and including everything necessary to reproduce it.

Assignment 1 - Group assignment:

  • read the safety sheet.
  • put on a lab coat
  • put on a mask
  • put on some glasses
  • were gloves

Together we create a whole slew of wooden planks of approx. 10x15cm. Each one we coat with a different substance. This way we want to test the release possibilities.

Sadly I had to leave early, so I couldn't follow the test all the way. The next morning though I saw that a whole lot of more things had been made: like a frisbee for example.

There were great difference between the release agents: some didn't do anything and the epoxy had bonded with the wood really well. Others were much better. Notably vaseline and cling foil work well. Although the latter one has the issue that it causes wrinkles...

Assignment 2 - make something myself.

As my daughters b-day is coming up, I wanted to make her a unicorn for on her bike. I searched online for nice 3D models, and found a few that looked nice. But after some back and forth discussion with Michelle, I discovered that there are certain features in these designs that are not possible to make with subtractive machines.

I do want to venture back into this road and find smartness around the issues that I have found. So some interesting things to note are:

  • You can make an body from any STL file in Fusion360. But you do have to import it as mesh and then convert it. There are many tutorials online how to do this.
  • Before you can make it into a body you do have to reduce the amount of points in your mesh
  • You can cut the mesh pretty well if you only need part of the design.

There are many discussions to be had on how you make your final model. In the Waag it turned out that some people advocated for making a positive head in PIR and then put fabric/fibers on it that are drenched in epoxy. This would resemble what you have done as kid with paper mache, but now with chemicals.

The other faction was more in favour of creating a negative of your body in PIR and fill that up with fabric and epoxy. The only problem there becomes that some fabrics do not stretch well; a piece of linen for example. And then folds may occur.

The last part that remains to be discussed is wether you put a positive mold on top of the negative mold so you can put more pressure on the composite. This helps the resin to reach all small detail in your design. Depending on how you made your design, it remains to be seen if this is something you actually want...

changing plans to make something

In the meantime, my fellow student Paula showed how she was working on tiles for her fan. And I really liked her idea! Tiles, walls and epoxy allow for a whole range of creative ideas. Starting with fancy coloured tiles with pieces of fruit entrapped in them, to full-fledges networked walls with LEDs!

So I am now going to experiment with making a simple tile with a piece of orange and on with a leave from my garden in them.

On friday, just before the weekend came, me and Paula quickly milled two small tiles out of a piece of PIR that was laying around and I brought it home as I had some epoxy at home.

on the left you see the mold and on the right the two things I want to put in.

The part where I want to put the leave from the garden in has a thick layer of vaseline. The part with the orange is getting the cling foil.

One of the issues is that the epoxy I have at home is really really fast drying. It will be getting strong within 5 minutes, the package claims...

So I prepare everything so I can move really fast. While preparing I notice that it is really hard to convince the cling foil it should cling to the edges of my mold and stay there.

So I decide to remove the bottom part of the PIR, cover that in cling foil as well and use that as a top lid to press the composite in with the help of a glue clamp. I also used to pieces of wood I had laying around to equally divide the pressure over the surface.

This creates a big mess.

And also some questions. As what is this green part doing here?!

After opening the mold there were numerous lessons to be learned and observations to be observed.

The weird green part was my leave that, by the pressure of the clamp, had bee pushed out of the mold. Probably because there was way too much epoxy in there, the leave left together with the excess of epoxy flowing off.

When you press fresh fruit, you get fresh juice. Also when the fruit is soaked in epoxy resin. So it is a feature when you want to color your epoxy, otherwise it is a bug...

Lastly, the epoxy ate some of the PIR. Despite the layer of vaseline, you can clearly see that the surface has become nice and smooth, whereas the left, cling foil, part still has the bumpy surface of the shopbot.

I am not too certain about the type of epoxy I used, nor if this is something to worry about. In the datasheet of the epoxy it states clearly that "R&G 5 minute epoxy is free of solvent so it does not attack sensitive foams like foamed polystyrene, etc." 🤷🏻‍♂️

Next spiral: milling the PIR

As I am not happy with my result from the networking week, I hope to make a network of tiles and put my little student boards inside the epoxy. let's see how far we get tomorrow!

I make a design in Fusion.

The design consists of hexagons (that is a thing in our class) with little boxes next to it. The little boxes are there because I want to have copper wires extending from the edge of the hexagons. The wires should protrude on the side, so I need to prevent the resin from overflowing. The little part between the hexagon and the square should function as a dam. I plan to make tiny cuts with a knife in there and run the wires through the cuts. Hopefully the material qualities of PIR will close back around the wires and prevent the epoxy from also running through those cracks.

I export the design as STL. This STL I import in VCarve following the guide from the Moulding and casting week. But then I run into some errors in VCarve about the material depth and drill depth.

Then it dawns that a 3D design is not needed here: these are pockets in a plate: so I can also design as 2D. I quickly remake the design in VCarve itself. Then you do not choose the roughen path, but the pocket path tool. And export this path to a shopbot file.

I get a plate of PIR and put double sided tape on the bottom to glue it to the table

I also add wood on all four sides to keep it in place.

I zero the X, Y and Z of the shopbot. This time I miss-zeroed the Z on purpose and did a dry run first. After that turned out fine, I re-zeroed the Z on the actual pane using the alligator clip.

The milling started at 12:40 and was done at 13:10. As far as I am concerned this is promising. Most of the time is spend these days on setting up the machine. So with some more practice the time to make a mold can be greatly reduced.

But then we also need to come up with a plan to reduce the sanding that needs to be done afterwards...

Putting boards in Epoxy (and cotton?)

Now that I have my epoxy mold, I can pour the epoxy in. Before I try any of this on my own carefully crafted boards, I try it on an old defunct PCB that was lying around in the lab. Also, I figured, it would be nice to have bare copper wires in the epoxy as that will look better then the insulated ones.

So I first strip a few jumper wires and solder the end together and onto the edge of a test PCB.

I then put a lot of wax in two pockets of the pir and fill them up with epoxy.

Fun fact; the epoxy I used came from different containers. I first used up the old that was here, before I opened a new one. The eopxy consists of two parts: part A and part B. You have to mix them in the ration 100:41. When I started with the old epoxy, I somehow reversed the amounts. So I added 100 parts of B to 41 parts of A. I used this substance to put over the carefully prepared PCB.

When I used fresh epoxy, I noticed I messed up and this time I did use the right amounts. I put some straints of cotton in the pocket and also added some coffee beans for fun. Lastly I added a little LED: just to see how it would shine.

Now we wait.

The next day, upon my return, indeed the epoxy had not become hard in the pocket with the wrong mix. The other one was very hard, and filled with bubbles. Despite me degassing the epoxy and popping bubbles with a lighter...

I was unable to get the bean/hard one out without breaking the mold. This was unexpected as in the group assignment it had been proven that vaseline is a good release-agent.

The piece with the PCB was indeed still somewhat fluid. I made a little canal in the side of the pocket and poured the excess resin into a paper cup. The bottom part was still somewhat hard and I decided to let the rest sit there and wait until the end of the week if it still hardens out. Also because it can then be easier thrown out.

As for the LED tile: the LED works. But I am not really happy with the bottom of the tile and all the bubbles in there. I think I should make another attempt maybe where I use a vacuum bag to press really hard and use more fibres and less epoxy. This result is rather boring..

You can have a look at Paulas documentation for some more interesting results and projects with tiles.