3D-printed parts get ‘smart’ for an Internet of Things world

Published in

Purdue Engineering Review

4 min readJan 12, 2024

3D printing is a type of additive manufacturing that produces a part by depositing a thin strip, or in some cases sheets of material, layer by layer, until the part is built up from a three-dimensional digital model. 3D printing is valuable because it has a low cost of entry (especially for most consumer-grade printers), is easy to learn, and allows people to rapidly produce prototypes before advancing to more expensive and labor-intensive manufacturing methods.

While 3D printing has come a long way, it still has a long way to go in terms of functional smart-product production. Currently, printed parts are less geometrically precise and mechanically weaker than those produced by traditional manufacturing methods. Plastic 3D-printed parts also lack functionality. That is, the printer can produce forms and shapes, but the printed part doesn’t do anything — for example, it doesn’t conduct electricity, change shape or respond to stimuli.

This is a crucial limitation in today’s Internet of Things (IoT) and Internet of Industrial Things (IIoT) environment, where sensing and “intelligence” increasingly are being built into products and components so they can communicate vital information about status and functionality across a smart, data-rich, analytic environment.

Our work focuses on plastic parts, specifically in the area of developing novel polymeric materials that have electrical and sensing functionality. This enables the printed parts to conduct electricity, and due to the material’s piezoresistive properties (i.e., strain-dependent conductivity), they can monitor changes stemming from deformation, material breakage and other factors.

Thus, rather than adding sensors after the part is printed, the printed part is its own sensor. This has lots of benefits. We get more sensory information (if the entire part is a sensor, it’s like having a sensor literally everywhere on the part). We obtain information on the inside of the component (traditional sensors are only applied on the outer surface). And we don’t weaken the component by bonding secondary materials or leaving voids for embedded sensors.

Perhaps the most common type of 3D printing is fused filament fabrication (FFF), in which a plastic filament is heated and pushed through a nozzle that traces out the shape of a part, layer by layer, from bottom to top. For these printable plastics to be electrically functional, we need to disperse conductive fillers into them.

Extremely-high-aspect ratio fillers (fillers that are long and skinny, like hairs) are best suited for this use because it takes far fewer of them to make the material conductive. And when it comes to these fillers, nanotechnology is hard to beat. We use carbon nanofibers (CNFs) rather than carbon nanotubes (CNTs) because CNFs are electrically comparable but cost much less, which helps to make this product commercially viable.

The challenge with mixing conductive fillers into printable plastics is dispersion — the fillers need to be spread out evenly and not bunched up to get good and consistent conductivity. Many people approach the process via so-called dry mixing, in which the fillers and the plastic (in pellet form) are mixed in a heated hopper and forced through a nozzle while being mixed by an extrusion screw. Unfortunately, this method doesn’t disperse the fillers well, giving rise to a final product with poor and inconsistent properties.

We developed a novel wet-mixing method to ensure uniformity across the mix of conductive filler and the base polymer prior to extrusion. This uniform mix improves the consistency of the electrical and mechanical properties of the end product, enabling sensing to be embedded throughout the part for a more functional product that monitors itself.

Diagram of Purdue researchers’ wet-mixing method. (Purdue University image/Brittany A. Newell)

The applications for this patent-pending method are endless! If we could embed sensors throughout full structures, the wealth of information would be astronomical, and align perfectly with IoT and IIoT. Imagine not only monitoring common failure points of a system’s structure but also the entire structure itself. You could assess how variations in environmental conditions or mechanical loads impact a structure, and make real-time predictions of the system life span and impending failure.

More and more products and components are being manufactured from plastic materials. With the communication protocols, big data analysis tools, and machine learning techniques currently being developed for IoT, the predictive capabilities of these parts would be immense.