Don’t Panic, No One is 3D Printing a Nuclear Bomb

Additive manufacturing (AM), or 3D printing as it is colloquially known, has become something of a bogeyman in the security community because of the unique impact it has or will have on production and manufacturing. “3D Printing the Bomb? The Nuclear Nonproliferation Challenge,” a recent article by Matthew Kroenig and Tristan Volpe explores how AM could enable proliferation of nuclear weapons by 3D printing critical components. But how realistic is this? Like many new technologies, the novel concept of AM has led to some hyperbolic promises on just what it can deliver. While AM is a unique process with many new applications for manufacturing, it may not be the revolution that we have been promised. Thus, it is important to take a sober look at the implications of this technology.

The core assertion for the promise and dangers of AM lies in how it will completely revolutionize logistics by streamlining the chain of production. Kroenig and Volpe argue that AM scales back on extensive production lines and requires less material, therefore making it easier to hide clandestine production operations. Moreover, since all AM designs are digital and many are available in the public domain, the expertise typically needed to design components is reduced and the manufacturing process is made even simpler. Kroenig and Volpe acknowledge that some areas of nuclear weapon development would remain untouched in the near-term, such as the acquisition of fissile material, which would remain difficult to acquire in the correct amounts or enrichment level. However, the basic fear is that AM could be used to establish a centrifuge production operation in a garage: The international community would be none the wiser and enriched fissile materials would be that much easier to obtain.

The problem with this hypothetical is that it at best overstates the capabilities of AM and at worst, completely misconstrues them. First of all, much as we wish it, 3D printers are not something out of science fiction. We have not created Star Trek-style replicators that can produce food or objects out of molecules. 3D printers still require inputs in the form of powdered metals or plastics, sometimes in significant amounts. In the centrifuge production example, this would mean having a large supply of powdered maraging steel, the material most often used in centrifuges and a controlled item in its own right for its use in nuclear and missile applications.

This raises other problems since many materials require extensive post-production finishing to protect it from corrosion or strain. Maraging steel is among those materials that require post-production finishing and whose powdered form rarely matches its conventional alternative in terms of quality. If it is not properly tempered, the steel could shatter under the incredible pressure of the high speed rotations of a centrifuge. This is compounded by the problem that the AM process itself is not conducive to forming strong materials. The AM process fuses successive layers of metal powder, creating a lower-than-average tensile strength in the final product. If there are any pockets or flaws in the layering it could compromise the density of the final piece even further. This means it is not as simple as uploading a blueprint to the 3D printer and letting it go to work on the material. Working with complex materials like maraging steel, there is still a relatively high barrier of technical knowledge necessary to make such materials viable for use.

All of this is compounded on the fact that industrial 3D printers are not simple machines – they are highly complex mechanisms using high-powered lasers to superheat powdered metals. While the spread of AM technology is often cited, the vast majority of those printers are small, recreational printers that basically make plastic junk. The top of the line machines that would be capable of manufacturing centrifuges or other nuclear components in turn require top of the line operation and maintenance. Again, the bar for technical expertise in AM is much higher than most people might think it is, at least if you intend to print more than little plastic knick-knacks.

There are also significant costs associated with Additive Manufacturing, which may sound strange considering that reduction in overall costs through time and materials is one of the biggest arguments for AM. The fact of the matter is, however, it is still not cheap. Industrial 3D printers generally cost upward of $100,000 USD, while the very best can easily cost more than $1 million. This does not take into account the cost of materials, which can cost hundreds of dollars per kilogram. In many cases, powdered metals often cost far more than their non-powdered counterpart. Finally, the greatest costs come from the lack of economies of scale. The nature of AM is such that the first item you produce will cost the exact same as the 10,000th item you produce. This contrasts with traditional manufacturing where high initial costs give way to costs per unit that are only a fraction of the start as operational efficiency increases and fixed costs are spread out over more units. In a situation like the example of centrifuge manufacturing, AM just does not really make sense. A state would not want only one or two centrifuges, they would want tens of thousands in multiple running cascades and that is just not what AM is designed for.

All of a sudden, the allegedly small logistics chain for a centrifuge production operation now requires veritable tons of export controlled powdered metal, extensive post-production structural and finishing expertise, high investment costs for the machinery, materials, and labor, and no foreseeable gains for costs over time since there are no economies of scale. This process does not seem that much more accessible or sensible than traditional manufacturing in most instances of production, much less in the pursuit of nuclear weapons. It is not clear that AM has made any significant contribution to the proliferation of nuclear weapons or components. When you look at the core of what AM is, it looks to be more smoke than fire.

Overall, the case for a radical departure from traditional manufacturing that may enhance proliferation is just not there. AM is new and different, which can be scary, but it is not the wild, game-changer it has been hyped to be. Kroenig and Volpe realize that large areas of nuclear development will remain unchanged, such as obtaining fissile materials, but even the case for their limited area of application is overstated. AM is an emerging area that should be carefully watched, but it is just that – emerging. There is simply no present threat to nonproliferation.

Are there things we can do to prevent AM from becoming more dangerous in the meantime? Yes, especially with export controls. These are not entirely new technologies. Already the Commerce Control List has a category specifically for lasers and sensors, machine tools, and the category that controls building materials already controls powdered metals. It would not require any radical changes to include AM lasers, common dual-use powdered metals, or 3D printers in whole in these categories. It would only require determining where the threshold for control should be. Let’s take steps to regulate the most troubling aspects of AM using export controls, but let’s remain calm while we do it.

Photo is public domain. 



One comment

  • Evan, this a great analysis! I agree that AM’s threat is overinflated due to its novelty. I have some additional thoughts and comments below

    As you mentioned, the availability of quality powdered maraging steel is a steep barrier. The steel powder will require a BIS license, which complicates (but does not preclude) would-be proliferators. There have been export enforcement cases where American metal suppliers cooperated with US Government agencies to uncover a proliferation attempt to illegally re-export controlled aluminum from Belgium to Malaysia and ultimately Iran. A similar sting could be seen in a scenario with an unusual request for powdered steel (i.e. Commerce’s Red Flags of suspicious customers).

    There is also the questionable quality and performance of AM-sourced goods. While there have been improvements, the technology is not advanced enough to reliably produce centrifuge products. Last September saw reports of aerospace companies using 3D printing to manufacture smaller items, such as nozzles. These smaller components are the “low hanging fruit” of AM in aerospace. More complicated products and assemblies will remain outside of AM’s reach until the production quality and reproducibility are improved. I cannot see a major aerospace supplier relying on AM for the items most likely to be placed under stresses similar to a centrifuge (turbine components, landing gear subassemblies, etc.). As an open source analog, I would suggest following AM’s role in aerospace for additional insight into its proliferation potential. If AM’s

    The emphasis on AM’s complexity and post-production gap strongly suggest the printers’ manufacturers would play a large role in ensuring a proliferator’s 3D printer ran smoothly. There is the possibility that DDTC may consider the maintaince/service of a maraging steel-capable printer to constitute a defense service. A defense service requires authorization from State (including a lengthy application process), further challenging the would-be proliferators’ goals.

    The amount of suppliers who manufacture 3D printers sophisticated enough for nuclear/aerospace applications may be small enough for the US government to provide targeted export compliance training. A more educated supplier base adds another obstacle, requiring falsification of end use/end user

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s