Submarines that can be manufactured at sea, on demand?

I note with interest (courtesy of a link at Old NFO’s place) that the US Navy is looking at a 3D-printed, easy-to-assemble submersible vehicle.

A team of researchers from the U.S. Department of Energy and the U.S. Navy have created the military’s first 3D-printed submarine, an achievement that may have the potential to accelerate the defense R&D process.

The sub – called the Optionally Manned Technology Demonstrator – is a 30-foot submersible made of thermoplastic resin, and it closely resembles the covert infiltration mini-subs used by the Navy SEALs. The hulls for these subs currently take three to five months to build and about $600-800,000 each. But the Naval Surface Warfare Center Carderock Division’s Disruptive Technology Laboratory (DTL) partnered with Oak Ridge National Laboratory (ONRL) to bring down the expense: using ORNL’s Big Area Additive Manufacturing facility, they printed the hull in six sections at a cost in the tens of thousands. A contractor assembled the sections into the final product. The whole process took weeks rather than months.

There’s more at the link.

If this works, think of what it’ll mean for deployment.  Instead of having to ferry something like this (for example, for use by Navy SEAL teams in a combat zone), a 3D printing facility can be installed aboard something like a Landing Platform Dock, or LPD.  I’m sure the US Navy’s San Antonio class LPD’s could accommodate it.  The Navy could “print-on-demand” however many of these it needed, and simply sink or otherwise discard them after use.  They’d be cheap enough to be disposable.  The technology need not be limited to submersibles, either.  Boats, aircraft parts, even weapons housings could be produced in this way.  Think of the saving in shipping, through not having to get them all the way from rear bases to the front line.

This might indeed be a revolutionary development.



  1. Actually, DARPA was working on this concept back in the '80s, but the technology wasn't available to do much more than local CNC machining from archived data sent over a network.

    Nicholas Negroponte also dealt with it at some length in Being Digital, (1996) using the catch phrase "ship bits, not atoms."

    Glad to see tech is finally catching up. I'm betting it won't be long before it gets cheap enough for manufacturers to simply keep part description databases online and a local "resource provider" prints (or machines) the part. Not quite time yet to sell your UPS and FedEx stock, but it's coming.

  2. The Army has been working since at least 2013 (when I was assigned to ARCIC). Mostly as a way to simplify at least some parts of the logistics chain.

    Instead of having to ship various items from the States, and maintain a stockpile against anticipated need, simply set up 3D printers at forward support bases and make the items as needed. Shipping blocks of base material instead of finished items would both reduce and simplify the supply chain.

    Of course, being able to do the 3D printing in metal rather than plastic will be a big deal.

  3. I read about Boeing looking into this in the early 90's, except they were looking at CNC milling to make a variety of parts from a single common blank, so that they didn't have to stock and ship a bunch of specialized parts and instead make them on demand in several repair hubs.

    In this case, given the stresses involved (and the inspections required because of them) I'd be leery of doing it with a pressure hull, but I could see it done with small boats, especially for special forces use – a stock of used commercial pieces like outboards, gauges, etc could be assembled onto a hull made to look like a local boat – or into an imitation of a specific boat for infiltration or to make someone look bad. The same could be done to change the look of vehicles or even personnel.
    As Jeff above says, the difference from the past is doing it by computer instead of manually – I suspect that much of the manual ability has been lost, so much that it isn't an option anymore.

  4. If the materials are just a thermoplastic, I imagine they could simply grind them back down into feedstock once they are done. They might need a filtration/purification step to deal with contaminants, but it ought to be possible.

  5. Thanks Peter, and this particular one is carbon fiber… Which IS a step forward. Yes DARPA was involved in the early CNC revolution, looking at data transfer and 'local' manufacture. Back then it was also a 'bandwidth' issue.

Leave a comment

Your email address will not be published. Required fields are marked *