Molding and casting

From the product page on the Smooth-on website, we found the Technical Bulletin and Safety datasheet.

This material is very safe to use. According to the safety datasheet: "Ingestion: rinse mouth". The technical bulletin says nothing about heat generated by the reaction of the components, so this seems to be a very easy material to apply. The only notes about heat are as a means to speed up curing.

We used clear plastic cups with insulation tape on the side at exactly the same height to prepare equal volumes for both parts.

After stirring, I dropped a surprise into a paper cup. Heleen poured in the mix, and we let it cure for 30 minutes. We had 6 minutes to do this before curing starts, according to the datasheet (pot life).

The surprise was easy to remove and left a nice negative in the puck of silicon.

Together with Henk, we created "bio-silicon". It's not really silicon, but it has similar hardness properties.

Ingredients (by volume):

• 1 part glycerin (99%)
• 1 part gelatin powder
• 2 parts water

Put the glycerin and water in a container and heat in the microwave. Add the gelatin powder while stirring. Once fully mixed, reheat and stir again. Pour the mixture into a mold and cool it down.

We poured the mixture into a machinable wax mold (of a fertility statue) made by someone in a previous year.

It did not come out very cleanly.

In a second attempt, we applied some sunflower oil to the mold and cast it again.

Next, happy with the result, we cast the remainder of the experiment Heleen did with a solid Smooth-on plastic into our new the glycerin / gelatin mold and removed it after 30 minutes of curing.

The result was quite good!

We cast Smooth-on Sorta-Clear-18 into the still usable mold, and the result was (after a weekend of curing) very good.

Henk: "We're not doing that; we'll only use the ShopBot this week."

My guess:

Filament printing

Removing the layer lines is hard because most filaments do not sand very well or require toxic chemicals to smooth.

SLA printing

Will probably yield the best results.

Machinable Wax milling

Wax is soft and have be post process pretty easily.

Dropping the bio experiments, I decided to go for a simple tealight candle from beeswax. Henk had prepared blocks of machinable wax, so I designed toward that, keeping in mind the examples here in the lab. The examples CNC mill a "positive" mold to cast a "negative" mold to cast the "positive" end result.

In the OpenSCAD program below, I created a "shape" and some support objects; "wick", "pouring" and "venting". These are subtracted from the "base" and that is "sliced" and negated to get the mold (of the mold). It was a bit of a puzzle but OpenSCAD made this manageable.

base_x = 70;
base_y = 90;
base_z = 40;

register_d = 10;
shape_d = 30;

$fn = 50;

module shape() {
  w = shape_d;
  translate([0, 0, w / -4]) difference() {
    cylinder(h = w / 2, d1 = w, d2 = w * .8);
    translate([0, 0, w / 2]) sphere(w / 10);
  }
}

module wick() {
  w = shape_d;
  d = 2;
  translate([0, 0, w / 2 - w / 4]) cylinder(h = shape_d * 2, d = d, center = true);
}

module pouring() {
  w = shape_d;
  d = 6;
  translate([-shape_d * .8 / 2 + d / 2, 0, w / 2 - w / 4]) cylinder(h = shape_d * 2, d1 = d, d2 = 10);
}

module venting() {
  w = shape_d;
  d = 3;
  translate([shape_d * .8 / 2 - d / 2, 0, w / 2 - w / 4]) cylinder(h = shape_d * 2, d = d);
}

module base() {
  cube([base_x, base_y, base_z], center = true);
}

module slice() {
  sep = 5;

  module register() {
    union() {
      difference() {
        cube([base_x, base_y, base_z / 2], center = true);
        cube([base_x - register_d * 2, base_y - register_d * 2, base_z / 2], center = true);
      }
    }
    translate([base_x / 2 - register_d, 0, 0]) cylinder(h = register_d, d = register_d, $fn = 200);
  }

  // bottom
  translate([base_x / 2 + sep / 2, 0, 0]) union() {
    difference() {
      children();
      m = base_x;
      translate([base_x / -2, base_y / -2, 0]) cube([base_x, base_y, base_z]);
    }
    register();
  }

  // top
  translate([(base_x / -2) - sep / 2, 0, 0]) rotate([180, 0, 0]) difference() {
    children();
    m = base_x;
    translate([base_x / -2, base_y / -2, -base_z]) cube([base_x, base_y, base_z]);
    register();
  }
}

module mold() {
  rotate([180, 0, 0]) difference() {
    cube([(base_x + register_d) * 2, base_y + register_d, base_z], center = true);

    slice() difference() {
      base();
      rotate([90, 0, 0]) {
        shape();
        wick();
        pouring();
        venting();
      }
    }
  }
}

mold();

A not dangerous material at hand at the Waag is Smooth-on Sorta-Clear 18 (after ingestion: "rinse mouth"), so I decided to try using that for my beeswax candle mold. To find out how much material I needed, I measured the volume of the mold of the mold (yes, confusing!) using water.

I needed 175ml.

Sorta-Clear has a density of 25.6 in³/lb and water has 27.68 in³/lb, so that's close enough for our purpose. As we are in a civilized region of the world, I measured in milliliters. One gram of water is 1 milliliter, so I'll be fine with 175 grams of Sorta-Clear for this job.

Mixture (100A:10B by weight):

• Part A : (100 / 110) * 175 = 159 grams
• Part B : (10 / 110) * 175 = 15.9 grams

No material left over!

After curing for two days (it needs 24 hours), it was tightly packed into the mold, and it took some strength to take it out. Fortunately, the cured Sorta-Clear is pretty strong.

Oops, I discovered there was no entry for the pouring hole.

I used a box cutter to provide access to the pouring and venting holes.

Henk has beeswax here, so I prepared the beeswax by heating it. It melts at 80°C.

The cotton wick needs to be primed to avoid it from burning without "wicking" the beeswax as fuel.

For casting, I clamped it between two pieces of triplex with a screw clamp, to ensure it does not come apart when casting the beeswax.

After cooling it in the fridge for a while, I took it out. To my surprise, it was (almost) perfect!

But I applied the clamp too tightly, so the candle is not round but an oval. The material used (Sorta-Clear) is very flexible (18A), so it might be a too flexible for this purpose. Although, the candle is very small and delicate, so the flexibility of the mold does help keep it intact.

The tool paths are still visible, so I kinda failed on that this week. But it works!