University of Michigan professor doubles 3D printing speeds using vibration-mitigating algorithm
The Smart and Sustainable Automation Research Laboratory of Chinedum Okwudire, an Associate Professor of Mechanical Engineering at the University of Michigan, has developed a software algorithm called “FBS Vibration Compensation” that effectively doubles 3D printing speeds.
Consumer 3D printing has done something that few people would have thought possible a few decades ago: brought small-scale, computer-driven manufacturing to the home and office—and for the price of a new television.
But despite the wealth of desktop 3D printing options out there—within the FDM category and beyond—consumer-level 3D printing does have its limitations.
Speed is undoubtedly one of those limitations. Although desktop FDM 3D printers can print faster today than they could a few years back, there are still some fundamental characteristics of compact, tabletop 3D printers that put limitations on their speed.
Interestingly, it’s not just things like motor speed that keep desktop 3D printers lagging behind their industrial counterparts. Because while some desktop-standard hardware can theoretically operate at very high speeds, hard-to-control factors like vibration make those speeds impossible to implement in practice.
This is because consumer-grade 3D printers are made with lightweight components—something that works to their advantage in some regards, like portability and cost, but which means they aren’t well equipped to resist motion-induced vibration.
That means that overdoing 3D printing speeds can lead to errors caused by vibration around the print head.
Chinedum Okwudire’s lab at the University of Michigan is trying to combat this very problem—by developing a software compensation technique that anticipates and mitigates vibration before that vibration ever occurs, effectively blocking the problem at its source.
The technique is known as “filtered B-splines (FBS) Vibration Compensation,” and uses a priori knowledge of a 3D printer’s dynamics to mitigate vibration problems. This allows the machine to print faster without inducing errors, like surface waviness or ringing.
“Software compensation is not a new technique, but the challenge is how to employ it in a way that’s effective, robust, and versatile,” Okwudire explains.
And with the FBS algorithm working effectively in experiments, Okwudire and his team think this could be a highly cost-effective means of squeezing that extra bit of juice from a low-cost desktop 3D printer.
“The great thing about using software compensation is that it is, in a sense, free,” Okwudire says, contrasting the clever technique with other options like investing in higher-grade physical components.
One test conducted by the lab involved 3D printing a model of the Capitol Building, with and without the FBS algorithm.
First, the model was printed without FBS using “conservative acceleration levels” and at 60 mm/s speed—a simulation of how one might avoid vibration problems without having access to a vibration-mitigating algorithm. This meant that a printing time of 3 hours and 59 minutes was needed.
Next, the researchers attempted to print the model, still at 60 mm/s but with 10 times higher acceleration levels. They carried out this print both with and without the FBS algorithm, which required 2 hours and 6 minutes for each. While the FBS-assisted print turned out perfectly, the unsupported print failed with comically extreme layer shifting. The researchers even found they could push the acceleration to 20x without compromising print quality.
But while early testing shows a lot of promise for the FBS technique, the University of Michigan researchers want to fully get their heads around how vibration compensation works. Doing so, they say, will allow them to mitigate motion-induced vibration even further, ultimately helping them to further increase 3D printing speeds.
In their investigation into FBS, Okwudire’s lab will be joined by Mechanical Engineering Professor Emeritus Galip Ulsoy, also of the University of Michigan.
“We’re going deep into the math and fundamentals to better understand FBS and how to apply it even more effectively,” says Okwudire, who believes that this refinement will help the group better accommodate uncertainty in 3D printer vibration.
Excitingly, this potentially game-changing research is already being prepared for life outside of university. Okwudire is planning talks with desktop 3D printer manufacturers who may be interested in using the FBS method to improve the speed, accuracy, and reliability of their machines. He is currently working to introduce the FBS algorithm into firmware, like Marlin and Repetier, used on desktop 3D printers.
The Associate Professor also has a wider vision for the future of desktop 3D printers, machines that he says can “have a lot of educational and commercial uses.”
He envisages a future in which consumer-level 3D printers utilize artificial intelligence and generate and gather “big data” for performance enhancement. The aim would be for every 3D printer to learn from the experiences of other printers in order to improve itself.
Such intelligent 3D printers could also be designed to be more flexible and user-friendly, by equipping them with “apps” like smartphones. These apps would be designed by everyday users as well as the companies producing the machines, and it is this kind of user involvement that Okwudire thinks is so exciting about desktop 3D printing.
“What most appeals to me about desktop 3D printing is that it’s a grass-roots and, no pun intended, bottom-up technology,” he says. “So many ideas come from the broad base of enthusiastic users, which creates an incredible opportunity: to make 3D printers as versatile and easy to use as today’s smartphones.”
Okwudire’s has been frequently recognized for his contributions to research. His awards include the Young Investigator Award from the International Symposium on Flexible Automation, the Outstanding Young Manufacturing Engineer Award from the Society of Manufacturing Engineers, the Ralph Teetor Educational Award from SAE International, a University of Michigan Mechanical Engineering Department Achievement Award, and the MLK Spirit Award.