Can embedded electronics be combined with 3D printing in a high temperature environment to produce a super heat-resistant drone? Ido Elyon and Stanley Leung of Stratasys Asia Pacific approached PhD student Phillip Keane, who had already successfully launched a CubeSat company, to try to answer this question.
The battery cavity which was embedded in the 3D printed quadcopter, produced on a Stratasys Fortus 450mc 3D Printer
Keane is researching applications of ULTEM 9085, Stratasys’ traceable, aerospace-grade, high-strength FDM 3D printing material, at the Singapore Centre for 3D Printing at NTU (Nanyang Technological University) Singapore. The drone that he designed, a quadcopter, was 3D printed with embedded electronics. Embedded electronics are not a first, but the temperatures involved were very high; when 3D printing ULTEM 9085, the material requires a print chamber temperature of a minimum of 160°C and an extruder temperature in the region of 300°C.
The end result was an incredibly tough quadcopter that can, in principle, survive in temperatures that exceed the limits of commercially available drones. Additionally, the project has determined some best practices to be employed if engineers should wish to embed electronics hardware mid-print.
Michael Swack at Stratasys Blog: What were the goals of the project?
Phillip Keane: We wanted to demonstrate that we could install off-the-shelf electronics in a high- temperature environment using an FAA-certified thermoplastic, and document the best practices for future reference. A few companies are looking at printing electronics and FDM plastics in a combined process, and there is a belief that in the future, 3D printers may be combined with other automated processes in order to reduce the steps required for manual assembly. So we wanted to demonstrate that it is possible to combine PCB installation and 3D printing in one stage.
One particular challenge of this project was to design the drone so that it featured no internal support material. This meant designing the interior with 45° angles to provide self-supporting angles. For cavities such as the battery cavity, we 3D printed a flat plate out of ULTEM separately, and installed the plate during the print process, to allow us to create the void without using support materials (which would have been very difficult to remove). So another goal was to try and maintain a nice curved exterior while having an angular interior. The Stratasys Insight software allowed us to fill the surplus areas with a honeycomb-like structure.
Also, as ULTEM is very strong, we decided to see how strong we could make the drone.
Michael Swack at SB: And how strong is it?
Phillip Keane: I was interested to see how FEA simulations of the drone structure compared to real life tests. So I designed a test piece of one of the drone arms and 3D printed multiple copies with a variety of different air gap parameters (resulting in different weights and different strengths). Upon testing, we showed a difference of around 15% between the simulations and the actual tests. In other words, each drone arm could carry 20kg before failing in simulation, compared to 17kg in real life. There are four drone arms, so the drone is capable of supporting a mass of around 68kg as a distributed load. Of course, as we are using small propellers and motors, the drone itself could never actually generate so much thrust as to lift that weight, but it’s still nice to know that our drone is rugged!
The difference between simulation and real-life is an interesting topic of research at the moment. FEA simulations are fairly accurate where it comes to monolithic (molded) parts, but because of the anisotropic nature of 3D printing, and due to other parameters, such as air gap, there are errors within the simulation.
Michael Swack at SB: How long did it take to produce with 3D printing?
Phillip Keane: The drone was completed under 14 hours. During the 3D printing, there were just three pauses for the electronics to be placed within the chassis.
Michael Swack at SB: Was the 3D printed quadcopter a successful project?
Phillip Keane: The question of whether we could 3D print a drone that was flight-ready was answered very early on in the project. Yes, if we used high-temperature batteries and 3D printed the drone upside down (to prevent the motor shaft colliding), we could, in principle, have a drone that would fly right out of the 3D printer. So with that question answered, we extended the scope of the project.
One issue we found was with the geometry of the brushless motors. In order to 3D print over a piece of hardware, the hardware in question must be completely flat. The motors have the prop shaft protruding above the print layer, which would have collided with the extruder. For that reason, we fitted the motors afterwards.
Everything else, the ESCs, the wiring harnesses, the receiver and the flight controller were fitted during the print, and were sealed within the airframe by the 3D printer. There were certain heatproofing measures taken with the electronic hardware. Some components had to be removed from the PCBs and replaced with high temperature equivalents.
Michael Swack at SB: How much heat can it take?
Phillip Keane, NTU (Nanyang Technological University) Singapore, examining his Stratasys 3D printed quadcopter, in front of the Fortus 450mc
Phillip Keane: The main structure and electronics can survive over 150°C very comfortably, and for several hours, as demonstrated by the 3D printing process. Our motors use high-temperature rated neodymium magnets, so they are good up to 180°C before they start to lose too much power.
The system overall is only as heatproof as the component with the lowest operating temperature. In this case, the battery is the weak link in the chain. Our battery is good up to around 100°C for short periods of time, so the overall operating temperature of this drone is in the region of 100°C.
Bear in mind, this was just a prototype intended to demonstrate the process of installing the electronics mid-print. It wasn’t our goal to design a heat-resistant drone. It just happened as a result of the process. Future iterations can be improved now that we have a new goal to aim for. We have found some high-temperature batteries that will allow temperatures approaching 200°C. It’s fairly heavy too. At just over 1.5kg (including the battery), it could do with losing a few hundred grams from the non-load bearing areas.
Michael Swack at SB: What are the applications for this drone?
Phillip Keane: As mentioned, this was just a prototype. But a more refined version could possibly be used by firefighters to investigate dangerous or hard to reach places. Also, National Geographic has made some nice videos where they have used commercially available drones to map the terrain around volcanoes. Apparently, their drones were not very heatproof, and they have lost a few of them in the volcano. So our drone can reach places that other commercial drones dare not fly.
Also, with regards to the process itself, I think there is a certain amount of value to be gained from embedding hardware directly into plastic structures. If you embed a piece of hardware directly, then you reduce the need for additional hardware mounting fixtures. This can potentially benefit applications where weight reduction is important, such as aerospace or motorsports. The process also allows us to completely seal a piece of hardware inside the structure. This offers protection from the environment, be it a wet or dusty environment, or even an environment with high and low pressure, without the need for excessive mechanical seals.
The 400mm class drone 3D printed using Stratasys ULTEM 9085 material. Transmitter shown for scale.
Stratasys and the University of Texas have recently demonstrated this by embedding RF components inside a piece of printed plastic (with Stratasys PolyJet 3D printing technology). So there is definitely interest in this area of research. I think we can safely say that we are the first to embed multiple systems within a single piece of ULTEM 9085 and have it fly afterwards.
Michael Swack at SB: What’s next for you and ULTEM?
Phillip Keane: I plan to finish my research at NTU first. And I’d quite like to carry on designing things with ULTEM 9085. Maybe I’d like to focus on building an underwater drone using the same process that we developed on this project. It’s a very cool material. I’m interested to see how far we can push it.
Michael Swack at SB: One last question: Why is the drone blue?
Phillip Keane: We had it painted. This is the same shade of blue as my first car which I owned, some 16 years ago.
Stratasys guidelines on embedding mechanical hardware can be found at this link, and a white paper that explains how to design strong, lightweight structures for FDM 3D printing can be downloaded here.