According to the consultancy MarketsandMarkets, aerospace firms will be using additive manufacturing (AM) to make more than $3 billion in parts by 2022, up from a bit over $700 million in 2017. Aerospace and AM have a mutually beneficial relationship. Aerospace firms often have the rigorous requirements in weight and performance and make the small volumes that AM is best suited for, so encourage AM development. AM in turn helps OEMs make much better products, boosting the sector’s growth.

But AM, also called 3D printing, is still young, and engineers have a lot to learn about the new process. In a webinar presented by the Institute of Electrical and Electronics Engineers, Denis Cormier, professor of industrial and systems engineering at the Rochester Institute of Technology, explained some basics.

Powder-bed laser printing systems are the majority of machines used now, but all the different laser and electron beam systems have their particular uses, Cormier advised. Each machine maker provides a guide for designing parts to be made with its AM process, including the surface roughness and ribbing to expect and orientation of support structures. “Surface finish is much better now, but you are not getting a machined surface yet,” Cormier said.

AM of lattice structures is increasingly popular, both to reduce weight and create heat sinks. The struts in AM lattices can be as small as a half millimeter and generally should not be smooth. But Cormier warned that the software code required for lattices can be “astronomically large,” and not every designer’s PC could handle it. However, some machine-makers are starting to enable trickle feed of code, like that used on computer numerically control machines.

Part unitization, or the combination of many small parts into an AM one, is another popular trend, as evidenced by the GE LEAP engine’s fuel nozzle. “You can throw the DFM [Design for Manufacturing] guidelines out the window,” Cormier joked. “There is no problem getting the part out of a mold.”

But very complex unitized parts may pose problems in getting extra powder out of internal channels. Most laser machines will allow powder to be jiggled out, but removal is tougher with electronic beam machines. Generally, if a liquid or gas is to flow through the internal channel, surface roughness is good. But for laminar flows of parallel layers, finishing the rough surfaces may be desirable.

AM can print exceedingly complex parts that may replace up to 20 parts made by conventional methods. But engineers must remember the surfaces that will have to be finished and seek to make this easier, for example by making unitized parts easy to hold in vises.

One big trend in AM is toward larger parts, up to 1-2 meters for laser powder bed or electronic beam processes. “But the cost of metal powder can be a challenge for large parts,” Cormier warned. A machine may have up to $100,000 worth of powder in it before it begins to make a large part. AM processes that use wire-feed may economize on bigger parts.

Another trend is toward simulation-based design. This can help engineers achieve uniform heating for uniform microstructure before parts are expensively printed. And OEMs are building machines with sensors and cameras that can measure the temperature of the powder beds to improve control.

The other trend, or wish, is for less expensive powders. Cormier hopes that competition will drive powder cost down or that wire and other feedstocks will come more cheaply. That would certainly spur adoption in aerospace and other industries.

For the moment though, most AM machine makers continue to require their own proprietary powders be used in their machines. That’s expensive, but may be necessary for a while longer to ensure quality results, especially in safety-critical aerospace uses.