Manufacturing a shell for deep dives at 1000m+



I am hoping someone can point me in the right direction to find out what it would take to build a housing for the OpenROV 2.8 that can withstand the pressure at depths below 1000 meters. I have seen other underwater vehicles that use titanium bodies or really thick acrylic.

My specific questions are these:

  1. What considerations does one need to take into account when building a housing that can survive at 1000m+? (materials, shapes, etc)

  2. Are there any resources that I can reference that will help me understand the science of my endeavor?


First off Welcome

In concept it can be done but there will be a number of challenges @Walt_Holm and @Darcy_Paulin are a good resource for pushing the depth limit’s of these unit

For the depth and materials it may be one of the simpler issues think smooth curved surfaces (sphere / Cylinder dome’s) rather than flat surfaces if you must use flat make them as small as possible and thick)

Attached is a spreadsheet I play with ripped off from a couple of different threads here in the past

Simple Pressure Vessel Failure Modes (1).xls (41 KB)

I would see the major issue being the drag especially if in the ocean (potentially less of an issue in a lake)

Given a tether of 1000m and 3mm cross section (neutrally buoyant) that’s equivalent to 3m2 surface area that any currents present will make it difficult for the unit to operate

Although for deeper dives clump weights have been used to explore a small area around the clump weight

Good luck and keep us all in the loop with how you are going



I have exactly ZERO experience with ROVs but I am SCUBA certified and have a degree in Physics (from 1991 and I am a software developer by trade but that is another story.) I just read the article in the May/June 2016 Popular Science and was intrigued by this project. When I read about the max depth I assumed that it was because of pressure causing leaks/crushing and immediately had an idea. I think it should work but you folks with real world experience may know a myriad of reasons why it won’t.

The biggest problem for manned submersibles is that the humans inside need to breathe air (a gas) which is massively compressible so the structure has to withstand the crushing pressure of the water outside while allowing the air pressure inside to remain “suitable for human consumption.”

Unmanned submersibles don’t have to worry about pesky “occupants” but they do still have to protect their contents (electronics, etc) from the nasty water outside.

My proposal, fill the “empty” spaces in the ROV with liquid. “Andrew, you knucklehead, then you are going to ruin the electronics before you even put the ROV in the water.” Ah, but what if you use the same dielectric (non-conductive) mineral oil that über-gamers recently started using to immerse their gaming PCs to cool them or that has been used in transformers (of the voltage conversion type, not the Saturday morning cartoon type) for years?

Anyone have experience with something like this and can explain why it wouldn’t work? The only issue would be that you would have to get as close as possible to every last bit of air out of the spaces because any bubble could lead to crushing. However, making the “skin” of the ROV slightly less brittle would allow it to flex without cracking to accommodate minuscule bubbles (think fractions of a grain of rice) that might hide in some of the components inside the ROV.

I searched for dielectric in these forums but only found references to using dielectric grease for connectors.


The main reasons mineral oil compensation won’t work well in this application is because it will affect the optic systems in the camera, the brushed DC motor in the servo and make the ROV negatively buoyant.

Here are some posts on this topic.


Ahhh, didn’t find other posts because I was searching for “dielectric.” Doh! So manned craft have pesky humans inside but ROVs other pesky objects like cameras and servos! LOL


Well, I actually read one of the links you provided and it looks like the only issue is the index of refraction of the liquid changing the focus characteristics of the lens. Other than that, the idea works. Of course, the camera is kind of a big deal and the problem can be compensated for but it is not a trivial matter.


Thank you all for your input.

I think that using mineral oil is an interesting idea and it raises another question.

Let’s assume that I fill the ROV up with dielectric mineral oil but only so that it occupies 50% of the interior (50% of the empty space is air and 50% is mineral oil). Because there is less air in the ROV now am I less vulnerable to pressure issues?

Disregard the odd design of this hypothetical ROV - I am just curious about the science behind pressure and how to overcome this obstacle.


Reducing the amount of air inside the tubes will result in a greater depth specification this effect is not linear at all. So reducing the amount of air inside the tube by 50% will not double the depth specification. Actually I would not recommend this design option as the mineral oil will only be at the bottom of the tube (filling it up to the half the height). This results in an uneven pressure distribution on the end caps and consequently probably reduces the maximum depth.