3D-printed parts are increasingly finding their way into airplanes and operating rooms.

Garter experts predict  that, by 2021, 75% of new commercial and military aircraft will feature engine and airframe components made through additive manufacturing.

Similarly, the use of 3D-printed medical implants  are set to increase by 20 percent over the next decade.

As the role of additive manufacturing emerges in a variety of industries, engineers will need to verify that their 3D-printed part is genuine and works as designed.

A team at NYU Tandon School of Engineering has found a new way to prove the provenance of a part.

Led by Nikhil Gupta, an NYU associate professor of mechanical engineering, the university researchers divided up a QR code and hid the hundreds of pieces throughout the layers of the 3D-printed part.

“By converting a relatively simple two-dimensional tag into a complex 3D feature comprising hundreds of tiny elements dispersed within the printed component, we are able to create many ‘false faces,’ which lets us hide the correct QR code from anyone who doesn’t know where to look,” Gupta said in a press release this week .

Gupta spoke with Tech Briefs about why “exploding” the code makes counterfeiting practically impossible.

Tech Briefs: What kinds of counterfeiting problems exist in the 3D-printing field today?

Prof. Nikhil Gupta: The counterfeiting problem is pretty big in manufacturing in general. In the aircraft industry, which is highly regulated, you still find a lot of counterfeited parts. There are several reports that show that a missile defense system was supplied with counterfeit parts . These are the places where we don’t assume there will be counterfeit parts.

Now, with additive manufacturing, people can reverse-engineer the product just by taking the scans; going to a high-quality print shop, printing the high-quality part, and applying it somewhere.

I see the problem as two-fold. First, you can make the counterfeit parts using 3D printers and try to sell it in the supply chain. Second, you can take a part, reverse engineer it, and then try to sell this reverse-engineered part.

A 3D printed part featuring an embedded, “exploded” QR cloud. (Image Credit: Gupta)

Tech Briefs: What is your approach to address counterfeiting?

Prof. Gupta: There’s an identification tag that we put on to the genuine part. This additive-manufacturing process makes everything layer by layer – unlike casting, where you pull this whole chunk of metal into a mold, and everything is cast in one go.

In this layer-by-layer manufacturing process, you can put certain identification information inside. Then, you can do CT scan or thermal imaging to retrieve that identification information.

Tech Briefs: What is specifically being placed into the part to identify it?

Prof. Gupta: Technically, you can do anything that helps you identify it, but in our case, we took a QR code – the ubiquitous quick-response code that you can find on all types of products.

We didn't just put this whole QR code in one layer during the manufacturing; we exploded this QR code into hundreds of parts –100 to 500 parts – and then we put each of these parts in different layers of manufacturing.

Tech Briefs: Why break up the QR code?

Prof. Gupta: If you put this whole code in one place, there’s a better chance of finding it or reverse engineering it. Also, if you put this whole code in one place, then the part may be weaker because the code acts like a defect.

Once we explode it into 500 parts, each part is really small. It does not impact structural integrity, because each of these hundreds of parts are so small that they are below the threshold of making any change to the strength of the part. Also, you can’t reverse engineer a part with 500 small features embedded inside. So, at that point, it’s very easy to separate the genuine part from the counterfeit part.

Tech Briefs: If I’m on the receiving end of all this, how do I verify and read this kind of distributed QR code?

Prof. Gupta: You have to read the QR code from a particular angle so that you can get this image perfectly. From any other angle, you will just get an image that won’t make any sense. The encryption and decryption information has to come from the supplier, to tell you where exactly to read this from. To read it, you need a CT scanner or thermal imaging equipment, or ultrasonic imaging; these tests are done in the aerospace and defense sector already. These imaging devices are available for quality assurance in most places.

Gupta and collaborators exploited the layer-by-layer AM printing process to “explode” QR codes within computer-assisted design (CAD) files so that they present several false faces — dummy QR tags — to a micro-CT scanner or other scanning device. (Image Credit: Gupta)

Tech Briefs: How those are placed in the 3D-printed part, layer by layer?

Prof. Gupta: Once we explode this code into, say, 100 parts, then we have to place this code into the CAD model.

Most people think that you take a CAD file, you put the CAD file into a 3D printer, and that’s how the printing works. But actually, what happens is that there are more steps in between.

You convert this CAD model into a STL, or stereolithography, format, which only gives the surface information and forgets the volume information inside. Then, that part is taken and “sliced” into two -dimensional pieces. Finally, those 2D slices are converted into a tool part.

Once we are at the level of creating these two-dimensional slices, we embed these hundreds of small QR cubes, and then let the printer print it with them.

Once the part is printed, the code is hidden inside. Each layer has only a small part of the QR code. Once you look at it from a specific angle, though, you are able to see the whole image of the QR code, from the CT scanner.

Tech Briefs: What is innovative about your approach to authentication?

Prof. Gupta: This is the first time we exploded this code and put it inside a part so it doesn’t affect the strength of the part. It’s very, very difficult to reverse engineer, because finding those 500 pieces and coding them into a part for reverse engineering is practically impossible.

The study, additionally funded by The Office of Naval Research, is now available online .

What do you think? Share your comments and questions below.