My current project has been making a set of Buoyancy floats for an Underwater Camera rig. The camera rig is about 25lbs out of water and as much as 5lbs negative in the water. Typically, buoyancy floats made of non-compressible foam are used to get the camera to around 1 pound negative or a bit less. I reconfigured my camera recently with the new HF-1 strobes and it got somewhat heavier and more negative, so I needed to rebalance it with new floats. I figured why buy something when I can make my own?

 

The requirements are:

* Minimize dry weight / maximize buoyancy

* Watertight and pressure resistant to at a least 100 feet deep

* Be competitive with the fairly lightweight foam blocks

 

I ended up with a two-part float sandwiched together around a 12" ULCS arm with dimensions of 225mm, 65mm, 65mm. Weight was 200g vs about 120g for the Foam blocks. Buoyancy is higher than what the foam blocks provided by about 20% as I increased the volume a bit more than the foam. I was getting about 325g from 4 standard Stix floats. Jumbo floats would be about 725g. These printed floats are about 800g of buoyancy per arm.

 

I tried a number of solutions, but the one that has worked so far is printing the floats with ABS, 1.6mm walls, top, bottom, 8% grid infill with a couple of internal 5mm wide "ribs" at 30%, and Acetone Painting of the entire print surface. The acetone painting melts the outer surface slightly and seems to be doing the trick of sealing the parts effectively at about 4 atmospheres of pressure.

 

During test dives this week I took the floats down to about 95 feet max and spent more than an hour at over 80 feet of depth. I will need to test them to 130 feet before calling it done.

 

A previous attempt printed with ABS, 5% Gyroid infill, and painted with epoxy ended up imploding at 75 feet. The print failed along layer lines with the infill collapsing. I changed the orientation of the print (horizonal rather than vertical), used Grid infill along the axis of maximum surface area, and add the internal 30% infill ribs. Successful PLA versions with epoxy were much heavier than the ABS solution.

 

Now that I have tested these to work, I will probably add a layer of carbon fiber/epoxy to the outside of the floats. This will add durability and some improved aesthetic to the final product.

 

 

 

 

 

Edited July 20, 20241 yr by Dave_Hicks
All text was lost!

@Dave_Hicks That's very cool Dave! Keep us posted.

From a design standpoint you should try to configure with cylinders. Would have better compression strength. Outside shape and internal(next to arms). You could put groves in to accept tie wraps to hold the halfs together.

2 hours ago, Tobyone said:

From a design standpoint you should try to configure with cylinders. Would have better compression strength. Outside shape and internal(next to arms). You could put groves in to accept tie wraps to hold the halfs together.

 

The problem with that is that I would have to print vertically, and I found out experimentally that it does not work well. You end with concentric rings on the long access with no cross-hatching from layer to layer. That model ended up imploding with a crack across layer lines and the (vertical) gyroid infill collapsing. My Fusion CAD model actually has a cylinder and rounded-rect option in it.

 

The rounded rectangle design can be printed on its side and the long top/bottom surfaces print with lines at 90degs between layers which results in a stronger print. I also changed to grid infill vs gyroid (8%) that supports those long surfaces. Lots of opinions say that gyroid is better, but not under 4atm of pressure apparently. I further added two internal ribs 5mm wide @ 30% on the thicker half of the float. The other half ends up better supported due to the inset grove for the ULCS arm.

 

I considered adding groves or loops on the side for zip ties but decided against it. I plan to add a carbon fiber wrap and groves would not work with that option. Electrical tape holds them together just fine and will last for years of diving.

Edited July 21, 20241 yr by Dave_Hicks