NASA 3D prints rocket parts — with steel, not plastic
By John Hewitt on November 15, 2012 at 10:47 am12 Comments
NASA’s Marshall Space Flight Center in Huntsville, Alabama, has 3D printed nickel alloy rocket engine parts using a fabrication technique called selective laser melting, or SLM. The part will be used on the J-2x engine for the largest rocket ever built, known simply as the Space Launch System. 3D printing (see: What is 3D printing?) has become popular for fabricating parts from plastic, but using the technique with metals requires equipment that is a bit more extreme. Will 3D printing of hard materials become part of a general, growing trend, or will these exotic fabrication technologies be viable only for elite, niche markets?
SLM evolved from an older method known as selective laser sintering, or SLS. In the traditional sintering process, a part is first molded from ceramic or metal powder and pressed into the desired shape. The “green,” as it is called at this point, is then fired in a oven to bond it. The oven was later replaced with a laser which provides greater precision and eliminates the need to handle the fragile, green part. Since full melting of the powder would destroy the part in the process of fusing it, sintered parts are not as strong as cast or machined parts, but they do retain intrinsically desirable properties like resistance to corrosion and temperature.
It was later discovered that full melting of powdered metal particles could be achieved by a technique called electron beam melting, or EBM. Since electrons tend to scatter off of gas molecules, expensive and inconvenient vacuum chambers are required for this process to work. As more affordable laser systems with higher power, more accurate beamsteering, and better focusing optics were developed, SLM was born. Like EBM, the powder material is fully melted during fusion, but the SLM laser does not require vacuum to function. It only requires that an inert argon or nitrogen atmosphere is used in the work area to prevent oxidation.
A part such as this J-2x manifold would not be milled on a CNC machine because the forces required to remove metal would warp and destroy a part this thin. Before SLM, it would have to be fabricated as a weldment from its component parts. Its curved and flaring nozzles would first be stitched together from sheet metal, than bent, and tacked to the main body. Welding introduces nonuniform stress points or “heat affected zones” at the weld sites, and makes failure modes less predictable. It would be beyond human skill to manually reproduce parts like this to the required tolerances. It would therefore have to be made using automated bending machines and robotic welders, which take considerable time to program and set up.
If one can afford the M2 Cusing SLM machine on which this part was printed (pictured right), the main concern is probably not the price of a few weldments. In many complex governmental projects like the Space Launch System, a majority of efforts are spent setting and revising project timelines. The primary driver for continued funding is demonstrating the ability to get the parts in hand by the time stated in the proposal.
Owning an M2 Cusing SLM machine goes a long way towards meeting deadlines, but with 3D machines that print an ever increasing variety of materials rapidly proliferating, it is important to clearly understand when plastic will suffice for a part and when something different is needed for the job. (See: The world’s first 3D printed gun.)
Most plastics are relatively soft materials with low compressive and tensile strengths, low melting points, and poor chemical resistance. Even the more expensive formulations like polyimide or polyether ether ketone (PEEK) give only modest improvements relative to metals. Plastics also become brittle when cold, and are quickly aged by exposure to UV light from the sun. Their lack of hardness also means that fine details cannot be rendered by traditional methods of manufacture since they do not hold up to the forces required to create them. Fine detail is also quickly degraded by repetitive use.
Many of our everyday products depend upon the electric or magnetic properties of metals. Recently, engineered plastics have been made which have conductivity approaching that of steel. Copper, aluminum and precious metals are still in a class by themselves and will remain so for some time. Plastics can, however, be embedded with other materials to yield unique properties in a way not readily done with metals. The high temperature necessary to process metals, and their lack of transparency to many forms of energy, make them incompatible with many materials that be used in plastics for expanding their capabilities. Metals can be made insulating, like plastics, by adding a protective oxide coating through the process of anodization. Metals can also be readily labeled by laser etching, and they stand up to a variety of coatings which expand their range of functions.