New Extrusion Machine Changes ShAPE

For nearly a decade, the Department of Energy’s (DOE’s) Pacific Northwest National Laboratory (PNNL) has advanced its patented, award-winning Shear Assisted Processing and Extrusion (ShAPE) technique. A pioneering technology for American manufacturing, ShAPE trades the indiscriminate heating typical of extrusion for a precise shearing action; the extreme deformation caused by the shearing can produce materials and components with superior properties while reducing costs and energy.

The Lab’s first two experimental ShAPE systems, ShAPE 1 (2017) and ShAPE 2 (2023), were designed and built to demonstrate the efficacy and scalability of the core ShAPE technique, showcasing how the innovative extrusion method could advance energy resilience and economic competitiveness for a range of critical energy sector use cases.

Now, a third ShAPE machine has arrived at PNNL: ShAPEshifter.

Unlike its predecessors, ShAPEshifter (built by Bond Technologies) is not a proof-of-concept. It was designed with a singular application in mind: extruding an especially brittle material called bismuth telluride at diameters as narrow as one millimeter.

For the ShAPE team at PNNL, working with such a delicate material meant that they needed to design a machine that is as flexible and reconfigurable as possible.

“From a force and scale perspective, this is ShAPE 0.5, not ShAPE 3,” explained Scott Taysom, a research engineer at PNNL. “But in many ways, ShAPEshifter is nimbler and more adaptable than its predecessors.”

Shaping up (literally)

All ShAPE machines, including ShAPEshifter, perform extrusion: forcing a “billet” of material through a die to produce the desired shape. However, performing thin extrusions of a very brittle material required a series of departures from the typical configurations for a ShAPE machine.

For instance, the pre-existing ShAPE systems extrude horizontally. But bismuth telluride is so fragile that it could sag and fracture when extruding horizontally, so researchers designed ShAPEshifter’s massive frame to pivot upward on command.

Once upright, stretching 14 feet into the air, it can extrude vertically.

“For particularly long extrusions, we suspect the vertical configuration might help the product keep its form,” Taysom said. “But we can see reasons that we might want vertical or horizontal orientations for these extrusions and are looking forward to exploring the scientific impacts of extrusion orientation.”

Direct impact

That same need for long extrusions necessitated another big change for ShAPEshifter. Previous ShAPE machines were primarily designed for indirect extrusion, which involves plunging the die into the billet of feedstock material. But there’s a drawback.

“With direct extrusion, the length of the extrusion is governed by the amount of material loaded into the machine,” explained Scott Whalen, co-developer of ShAPE and chief materials scientist at PNNL. “In commercial extrusion, they’re often making products 100 feet long or more. For that to work with indirect extrusion, your die would have to be incredibly long and complex.”

So the team ensured that ShAPEshifter was equally capable of direct extrusion, or pressing the billet into the die, rather than the other way around, to enable those longer extrusions.

A turn for the better

The adjustments don’t stop there. Working with delicate material posed yet another significant challenge. ShAPE 1 and 2 create localized heat on the billet by rotating the die, which rotates the extrusion itself as it emerges from the machine.

“Imagine you’re extruding several feet of brittle pencil lead,” Taysom said. “If that’s spinning multiple times a second, it’s probably going to shatter.”

“Ultimately, for an industrially viable process, you don’t want the die to rotate, because the extrusion will also rotate—which can be difficult to handle,” Whalen added. “Instead, the billet should rotate while the die remains stationary, which keeps the extrusion from spinning.”

But the heat from the rotation is still necessary. The researchers designed ShAPEshifter so that it could flip the script, keeping the die stationary while rotating both the canister that holds the billet and the ram that forces it through the die.

“Where the ShAPE 1 machine has a single rotational axis and a single translational axis, ShAPEshifter has two of each,” Taysom said. “Combined with the pivoting frame and ShAPE’s capacity for both direct and indirect extrusion, ShAPEshifter can achieve all sorts of configurations: this is rotating, this isn’t rotating; this is plunging, this isn’t plunging; this is horizontal, this is vertical. Basically, ShAPEshifter enables research on any tooling configuration we can dream up.”

Smooth operator

The last big difference in ShAPEshifter is the subtlest. The first two ShAPE machines use a hydraulic ram to press on the billet as it’s fed through the die—but again, ShAPEshifter’s specific needs demanded a new approach.

“Hydraulics are an economical way to get a lot of force,” Taysom explained. “But hydraulics work by essentially pulsing the liquid—which works great for ShAPE 1 and 2, where your extrusions are moving pretty fast.”

However, with ShAPEshifter, researchers plan to extrude at ram speeds as slow as one-tenth of a millimeter per minute. And at that rate, pulses from hydraulic systems would produce meaningful inconsistencies in the finished product.

“For ShAPEshifter, we switched to an electric servo-driven system to ensure smooth extrusion at incredibly slow speeds,” Taysom said.

A delicate matter

“ShAPEshifter is a purpose-built machine dedicated to performing research for national security solutions,” Whalen said. The machine is an extension of the Lab’s work with ShAPE 1 to extrude bismuth telluride—a brittle material that has remarkable thermoelectric efficiency—in a wire form.

Using traditional manufacturing methods to produce bismuth-telluride wire is difficult and costly.

“Normally, you would melt pellets of bismuth telluride into a glass cylinder to produce a cast billet,” Whalen explained. “Then, you would mill that billet into a superfine powder, densify that powder into a solid billet and extrude it.”

“With ShAPE, we can bypass all of that, extruding the bismuth-telluride castings directly into the finished product,” he continued. “And we’ve published research showing that those ShAPE extrusions are just as performant.”

And of course, even though ShAPEshifter is a purpose-built system, the research conducted on the versatile new system will advance a range of areas—and the ShAPE technique itself.

The research on ShAPEshifter is being conducted in partnership with Sandia National Laboratories. ShAPEshifter was delivered to PNNL in May 2025 and is currently undergoing acceptance testing.