Ben's Website

Serious musings

Electron Beam FreeForm Fabrication (EBF3)

There are many cool things happening in 3D printing these days, but the technique I'm most excited about, electron beam freeform fabrication (EBF3), has received very little coverage.  So in this and following posts, I want to describe the basics of this technique and some of the cases where I think it is the ideal manufacturing technology.

Printing in plastic is easy.  Heat some PLA or ABS to 300-400F and squirt it out of a small nozzel while tracing the outlines of your part.  Alternately, selectively shine a UV light source on some UV-cure epoxy and you have a stereolithography machine.  These two techniques, finally free from patent protection, are responsible for virtually all of the media buzz in 3D printing.

While these technologies accomplish the basic aim of converting a CAD design into a dimensional prototype, few of these additively-produced prototypes can withstand loadings similar to those a traditionally-machined part (even when machined from the same plastic, let alone metal versus printed plastic).  Not every application needs this durability, but it is the greatest limitation of every 3D printer you've probably heard of.

Printing in metal is expensive; in contrast to the great variety of Kickstarted $300-3,000 consumer/prosumer printers, MatterFab made news this past summer with the announcement of a metal-printer targeted at $100,000.  This printer, and it's million-dollar-plus competitors, uses a kilowatt-class laser to melt particles in a metal powder together, forming a solid part.  Depending on the scan speed, laser intensity, and material addition rate, this method (referred to as laser-engineered net shaping – LENS – and metal laser sintering) can produce fully-dense parts with material properties similar to those of cast or annealed parts.  Since melting the metallic powder depends on the relationship between the laser wavelength and intensity and the powder's melting point and absorbtivity, machine cost and material selection are closely related.  Common configurations have difficulty producing aluminum, titanium-aluminide, tungsten, magnetic alloys, and others.   These difficulties are easily explained by considering the reflectivity of some common metals versus common laser wavelengths:

Reflectivity of common metals and common laser wavelengths. HPLD stands for high-powered laser diode. Chart courtesy Kennedy, Byrne, and Collins, 2004.

Similar to LENS, Electron Beam Freeform Fabrication (EBF3) directly melts metallic materials to form a fully dense part, though using an electron beam rather than a laser. EBF3 commonly uses a stationary electron beam and a multi-degree-of-freedom positioning system to build parts layer-by-layer. As shown below, the electron beam is focused at a particular point, melting any co-located materials. Introducing new material into this region – by a wire feeder – increases the volume of this pool. Indexing the positioning system causes the pool to move, leaving behind newly deposited material. Adding a second wire feeder enables in-pool alloying and the production of functional gradients (varying the alloy along the part). Most EBF3 systems operate inside a vacuum chamber to both prevent the surrounding environment from attenuating the electron beam, which also eliminate the prospect of part contamination.

Left: Schematic representation of electron beam freeform fabrication (EBF3), courtesy Taminger & Hafley, 2008. Right: An EBF3 machine at NASA Langley, courtesy of Bird & Hibberd, 2009.

Along with the prospect of metal-agnostic (or more so than LENS), studies from an EBF3 research group at NASA Langley indicate that resulting parts are stronger than wrought and tempered alloys:

Comparison of EBF3-produced Al 2219 to Al sheet and plate from Taminger & Hafley, 2008. ‘Typical’ refers to conventionally-produced wrought and tempered sheet and plate properties. Of note, the ‘As-deposited’ specimen has greater strength than the wrought and a T62-tempered EBF3 deposit outperforms a conventional T62 alloy.

In addition to producing parts with commendable material strength, EBF3 is a fast process. Able to trade resolution for speed, EBF3 has been demonstrated at deposition rates of 178 to 594 cm3/hr (11-36 in3/hr) in Al 2219 and 434 cm3/hr (26.5 in3/hr) in Ti-6-4 [Taminger & Hafley, 2008]. As a point of comparison, a representative laser-based system deposits at 8 to 33 cm3/hr (0.5 – 2 in3/hr) [Taminger & Hafley, 2010].  The electron beam is also more efficient at delivering energy to melt pool, at approximately 95%, than a laser process, which might see 10% efficiency due to losses in the laser, beam transmission losses, and the naturally high reflectivity of most metals [Taminger & Hafley, 2010].

According to Lori Garver (NASA Deputy Administrator through 2013), EBF3 is used in fabricating the titanium spars for use in the F-35 Joint Strike Fighter; some more mundane results are below:

Parts produced in Taminger & Hafley, 2008: a) a TI-6-4 wind tunnel model, b) a square box of Al 2219, c) an Al 2219 airfoil, d) an Al 2219 mixer nozzle, e) an Al 2219 converging/diverging nozzle, f) a Ti-6-4 guy wire fitting, g) a Ti-6-4 inlet duct, and h) a Ti-6-4 truss node.

The significant disadvantage of EBF3 is poorer control of the part surface quality than plastic and LENS printers. EBF3 part resolution is essentially limited by the feed wire diameter, but this diameter dependence has not been demonstrated in the literature.  Given the commercial availability of LENS techniques, the majority of the community has focused on understanding EBF3 and its unique alloying ability.  EBF3's selling point of printing with high strength alloys places the focus on accurate alloy production; applications demanding these alloys are sufficiently advanced (and costly) to delay interest in higher resolution.

EBF3 also requires an evacuated build environment, on the order of 1x10-4 Torr, adding an appreciable degree of complexity to any EBF3 (terrestrial) system [Taminger & Hafley, 2008].  Davé’s original 1995 description mentions that use of a high-energy electron beam (>500keV) can eliminate the need for vacuum, though such a device will be accompanied by its own complexities in generating large potentials. The literature has apparently not yet considered this variation.

Producing spars for the F35 is nice, but to me the killer application for EBF3 is not terrestrial, but in-space.  In the next post I'll lay out why I think EBF3 is the ideal in-space manufacturing technology.