We had two major problems to overcome:

1. There was no previous documentation for the fabrication process of the Smartfin.

2. The past Smartfin designs had a tendency to break either in half or near the pins, resulting in the complete loss of the fin.

Our goal was to both document a process for the fin's construction and to strengthen the fin to prevent it from breaking in the water.

In order to create the documentation for the fin, we had to decide on a variety of materials for all the major parts of the process:

The Release Agent

We needed a reliable release agent to ensure that our resin wouldn't bond our molds shut. We looked at two different release agents: beeswax and PVA. We covered some aluminum stock in either beeswax or PVA and set them in curing resin.

Ultimately, both worked for our purposes. The beeswax made it noticeably easier to remove the aluminum from the resin, but it also left more residue and was much harder to apply in even layers. It was harder to remove the aluminum covered in PVA, but it left almost no residue. We went with PVA so we could ensure there would be minimal residue that could affect the finish of the fin.

The Mold

With the mold, our main goal was to use a material that is both robust and inexpensive. While we looked at materials like aluminum from the start due to its strength, we knew its manufacturing process would be long and difficult. For this reason, we ended up looking into 3D printed materials such as VeroClear and silicone as well:

Our tests narrowed down our mold material choice to silicone and aluminum. While silicone was much easier to make, it also was more difficult to seal the mold to prevent resin leaks. Additionally, we had to clamp the mold between aluminum plates to ensure it wouldn't bow outwards when the resin was poured in. If we clamped them too hard, then the mold would bow inwards and ruin the finish of the fin. If we clamped it too loosely, then we would risk resin leaking out of the bottom.

For these reasons, we ended up choosing aluminum as our final mold material. When we compared the costs, two aluminum test plates cost approximately $56 from Industrial Metal Supply with a 15% student discount and 15 minutes of milling per plate. The silicone test mold cost about $48 in materials, but our student privileges meant we didn't have to pay for overhead or the worker's time so this price could be much higher for general manufacturing. It took about 2.5 hours to print the entire test mold.

The Fin Fabrication Method

In order to strengthen the fin itself, we decided to incorporate fiberglass into its composite. We looked at three different methods:

Method 1: Epoxy Core

A mixture of epoxy resin and fiberglass strands was poured into the mold and left to cure for at least 24 hours. The mold halves were then separated, and the resulting fin was removed and post-processed via sanding and polishing.

Method 2: Die-Cut Fiberglass

Sheets of fiberglass were cut into the cross-sectional shape of the fin. These sheets were then layered with resin in the two halves of the mold. The two halves were clamped together, and the resin left to cure for 24 hours before the fin was removed and post-processed.

Method 3: Combination

This method is a combination of Methods 1 and 2. Essentially, one layer of fin-shaped fiberglass sheets were placed into each of the mold halves and kept in place with resin. Fiberglass strands were spread across the surface of the fin in one mold half and covered with resin to hold them in place. The mold halves were then clamped together, and resin was poured into the mold between the fiberglass sheets. This created an epoxy-fiberglass core surrounded by a shell of fiberglass sheets.

We cast three test fins using each method: 

Visually, we could see that the epoxy core method had an uneven distribution of fiberglass strands which wasn't ideal for our purposes. The die-cut method left voids and had an excess of air bubbles. The combined method somewhat eliminated both of these problems. To get a more in depth look at the fiberglass distribution throughout the entirety of each fin, we took X-rays. The whiter areas depict areas of high fiberglass density, and the darker areas represent a lack of material. 

The X-rays show the distribution of fiberglass throughout both the face and depth of the fin. The epoxy core method had very noticeable areas that lacked fiberglass, but a fairly even distribution across the depth. The die-cut method had noticeable gaps throughout its depth but was fairly even across its face. The combination method had a fairly even distribution across its face and better distribution throughout its depth than the die cut method.

We also did 3 point bending tests on three samples per method. From the provided data, we eliminated the epoxy core method as a viable method as its strength was significantly lower than that of the other methods.

With these factors taken into account, we chose Method 3, the combination method, for our final fins.

Our final molds were made out of 12"x12"x1" aluminum plates and milled with a HAAS Pre NGC milling machine. In order to properly align the mold halves, we inserted 1" diameter steel pins into Mold Half B and cut through-holes into Mold Half A. Additionally, we included relief slots on both sides of both mold halves; these slots provide an area to pry the mold apart using a pry bar after the fin finishes curing.