CNC 6061-T6, 4140 steel
As part of the much larger project to design and fabricate a solar powered hydrofoil, propulsion had to be optimised.
I started looking for commercially available propellers, with high pitch model airplane propellers being the best. However, in this research, i discovered a few tools available for designing and modelling propellers. The program i found most useful is called JavaProp; i used it to (very roughly) model and predict the properties of the various off-the-shelf propellers available.
Eventually i designed a couple custom propellers in JavaProp, and figured out a process to CNC machine them on the 3-axis Haas i have access to. The key with the process was in the use of smart cutting operations, leaving as much stiffness as possible as late in the roughing operations as possible, yet taking a relatively deep cut on the finishing profile. Additionally, fixturing with Ridigax tooling wax was a confidence boost and allowed every surface to be cut by the mill, leaving the part totally separated from the rest of the stock, waiting to get melted free.
The second, slightly larger prop was machined from 4140 steel, with the aim of allowing this to be heat-treated to truly knife sharpness and hardness. A scary propeller indeed! This process was largely the same, though i did discover that the stock allowable deviation from geometery of 0.0004" left noticeable surface irregularities. Decreseing this to 0.0001" reduced this to a significantly better degree. The curved blades still didn't require that many line segments, the code was manageable, and our non-SS mill didn't data starve during the cuts. I was quite surprised by the need to decrease this tolerance!
Casting/Molding
FEA in Fusion360 revealed that this 6061-T6 propeller had a safety factor of around 2.5, which should be adequate (just) for competition. Additionally, untreated aluminum is not particularly corrosion resistant in certain electrolytic or alkaline environments. I was worried about corrosion.
Once i had a CNC'd master from 6061-T6, i decided to investigate making molds to cast wax into, for using with the lost wax investment casting process (for a CCSF class), and sand casting. I wanted to cast aluminum bronze (C955 or similar), which is heat treatable, and quite strong.
The first molds were Smooth-On silicone rubber, backed up with GF reinforced UltraCal plaster mother molds for stiffness.
From these molds i cast a carbon fiber epoxy propeller, with unidirectional fibers running tip to tip. At the time of writing, i've only had a chance to cast one, but an improved mold (with core pin for central bore) will be made and new CF propellers should come from this. These should be very stiff and corrosion resistant!
Further mold-making explorations resulting in two part molds made from UltraCal for both sides, and Tooling Epoxy backed by glass fiber fabric (red two part molds in pictures).
Eventually i'll CNC molds directly, but there are software limitations in Fusion and Inventor with creating constant draft from 3d curves. The leading and trailing edges of the blades used as parting lines can't be used since they have changes in all three axes. Rhino will allow this, but i lose a fully parametric workflow.
