Designing for Additive Manufacturing = DfAM + MfAM

What images pop into your mind when we use the term “DfAM”? Perhaps you see beautiful topology optimized structures, organically inspired bird bone constructs, and perfectly latticed shapes? Imagine this all too familiar scenario…

It’s late on a Friday. The production supervisor has allocated the senior technician’s afternoon to set up a Ti-6Al-4V build in the designated multi-laser powder bed fusion machine. A new machine of which only one technician is currently trained on. The build chamber is purged. All you need is Design Engineering to send you the topology optimized part planned for a test build. You have left an open spot in the nesting of the build you have prepared to run over the weekend. At 4:49 PM an email pings your inbox. There is an attachment. It is the light-weighted part… riddled with overhangs and disproportionate feature transitions. You know full well this will blow up your perfectly engineered build that is set to run smoothly over the weekend…

The Design Engineer’s creation (left) and the Production Engineer’s reality (right) (Image courtesy of Laura Ely)

All AM processes have a distinguishing moment when we go through this manufacturing scenario where we must merge the worlds of the manufacturable and the designable. We refer to this “make it work for the process situation”, whether it be binder jetting, directed energy or another form of AM, as Modifying for AM or more simply, MfAM. Similarly, when the organization is leaning in across the board to design parts that are not only highly optimized but also manufacturable with AM, this is DfAM. Combining the two, the entire equation for designing for AM can be defined as:

Eq 1.  Designing for AM = MfAM + DfAM

When combined, DfAM and MfAM provide a winning strategy to help anyone translate design requirements into a viable end-use part with AM. For instance, if designers have the freedom to “clean sheet” a new design, then they would apply DfAM followed by MfAM. Alternatively, if they are redesigning an existing part for AM, then they would apply MfAM followed by DfAM.  Iterations between the two would obviously follow in both cases, but with this thinking one can begin to tailor a process and structure approaches to support designing for AM.

Designing for AM combines both restrictive and opportunistic design

While many of us agree that topology optimization, generative design, biomimicry, and cellular/lattice structures are often used for DfAM, Production and Process Engineers will be quick to ask:  What about… Build orientation? Designing support structures? Build layouts? References to establish datums for the machinist? Fixturing the part during heat treatment? In some cases, helpful new tools assist, but intervention by an engineer, who understands the entire process workflow, is still necessary. As it turns out, many companies still adhere to the traditional split between design and manufacturing when it comes to AM. They subsequently struggle when a designer hands off a topology optimized part, for instance, to a process engineer only to find that it must be modified for AM in order to be printed successfully. This often adds weight and extensive post-processing, which adds cost and delays, decreasing the viability of AM and tarnishing its promise.

In a recent survey, fewer than 35% of industry considered a finished component as the end point of DfAM. As a result, as an industry we use the same terms, but we do not share the same meaning, specifically for DfAM. We need to reconcile the traditional Design for Manufacturing mindset with the “design and material freedoms” afforded by AM and embrace a more holistic view when designing for AM; one that considers every step from requirements definition to delivery of a finished component. Only then will companies establish viable pathways to industrialize AM.

In the words of Helen Keller, “Alone we can do so little, together we can do so much.” The same is true for creating successful products in AM. We will achieve the most success by taking a multi-skilled, team-based approach and embracing both the opportunities and restrictions of our AM processes with the notion that designing for AM equals DfAM + MfAM from the very beginning.