Design for additive manufacturing (DfAM)
📍 Where we are: Part III · Design for AM — Chapter 9 of 15. Part II showed what each process can make. Now we flip the question: how should you shape a part so that a given process makes it well?
Imagine two cooks. One gets a recipe designed for a different kitchen and has to tweak it so it works on their stove: less salt, lower heat, a different pan. The other starts from scratch and invents a dish that only their unusual oven can make. Both are doing "design for additive manufacturing". The first is restrictive: adapt a design to the rules of the machine. The second is opportunistic: use what the machine uniquely allows.
What this chapter covers
We separate the two attitudes of DfAM and then go through the practical rules that dominate powder bed metal work: overhang angle, minimum features and wall thickness, holes, powder removal and heat. The numbers you will see are examples from specific machine, alloy and process combinations; they are not universal. Your running example, the lightweight bracket, is redesigned lightly at the end.
Two attitudes: restrictive and opportunistic
The review by Thompson and co-authors describes DfAM as the design side that lagged behind process development, covering both opportunities (new freedoms) and constraints (new limits) [1]. The standard ISO/ASTM 52910 gives general requirements, guidelines and "red flag" warnings about situations that often cause trouble, and notes that AM considerations are unusually interdependent; it does not give process-specific data [2].
- Restrictive DfAM takes a design (often one made for machining or casting) and adjusts it so it can be printed: add fillets, change angles, add drain holes. It lowers cost and risk, but it keeps the old shape.
- Opportunistic DfAM starts from what AM makes easy: complexity that costs nothing extra, merged assemblies, internal channels, graded lattices (see topology optimization and lattices). It needs more design effort and testing but can change what the part is.
A practical rule: use opportunistic thinking when the function justifies it, and restrictive rules for everything else.
Overhangs: the 45° rule and its limits
In powder bed fusion, each layer is melted on top of the one below. A downward-facing surface that sits over loose powder has nothing solid beneath it, so it can sag, warp or turn rough. A common rule of thumb is that surfaces tilted less than about 45° from the build plate need supports [3][4]. A review of overhang fabrication reports a minimum reliable overhang angle of 45° in one 316L study, with warping and poor underside roughness below that. It also stresses that smaller laser spot sizes make low-angle overhangs harder [3]. Such limits apply only to the machine, material and parameters of each study. So 45° is a design default, not a law; check your supplier's own data.
Three everyday powder bed rules (overhang angle, hole shape, powder escape) are restrictive DfAM; the freedom AM adds is the opportunistic side.
Original diagram by the authors, created with AI assistance.
Practical fixes: tilt the surface, chamfer a horizontal ceiling into a 45° roof, or make a round horizontal hole a teardrop (pointed on top) so that its roof is self-supporting. Better still, choose a build orientation that avoids the problem instead of relying on supports, which cost material and time and must be removed later [5]. Orientation is the subject of the last chapter in this Part.
Minimum features and wall thickness
A laser or electron beam has a finite spot size, and powder has a finite particle size, so there is a smallest thing the process can make reliably. A guideline study on a titanium alloy tested thin walls, bars and bore holes of varying size in several orientations to find the process limits [4]. Such limits are specific to the machine, material and parameters; treat them as an illustration of the kind of numbers a guide contains, not as values for your machine.
Other recurring rules in design guides [4]:
- Keep wall thickness reasonably uniform, because abrupt changes from thin to thick add stress and warping risk.
- Add fillets to inside corners instead of sharp edges.
- Leave a gap between neighbouring features so the powder between them is not fused together.
- Remember that as-built surfaces are rough, and that accuracy and roughness depend on orientation and position on the plate [4]; leave machining stock where a surface must be precise.
Powder removal
Metal powder trapped in a closed cavity is not just a nuisance: it adds mass, it may loosen later, and it hides the inside from inspection. Design guidance commonly recommends more than one escape hole per hollow region (so powder can leave and air can enter); minimum hole sizes are machine- and powder-specific, and bore-hole limits are among the features tested in the guideline study [4]. Channels should be short and straight where possible, and bends should be gentle. In electron beam melting, the sintered cake makes small passages even harder to clear (see electron beam melting).
Thermal considerations
Metal AM builds heat in and then cools it quickly, so heat flow matters. Thin features connected to large masses, tall thin walls, and big flat areas all concentrate stress and can warp or crack. The designer can help by keeping sections similar in thickness, avoiding long flat down-facing areas, and thinking about where heat can escape. More on distortion is in orientation, supports and build preparation.
Design for a specific process
Rules from one process do not carry over. Powder bed metal needs support and powder escape; binder jetting and sintering need shrinkage and sag control (see metal binder jetting); material extrusion has its own bridging and anisotropy rules. Start with your process's design guide and test coupons before committing a design [2][6].
The bracket test
Our lightweight bracket arrives as a machined design. Restrictive changes: round the corners, make the mounting holes teardrops or orient them so the holes are vertical, thin the heavy block, and give any internal pocket two escape holes. Opportunistic change: replace solid areas by a ribbed, load-following form, which is the subject of the next chapter. Either way, the mounting faces keep machining allowance.
Why it matters
A design that ignores the process leads to supports that are hard to remove, powder trapped inside, warped parts and failed builds. A design that uses the process well can be lighter, consolidated and cheaper. DfAM is how that difference is captured.
In the real world
Research groups and machine makers publish process-specific design guides, and academic studies such as the titanium-alloy guideline study derive limits for a single combination of material and machine [4]. A framework paper organises DfAM methods into process selection, redesign for function and optimisation for the chosen process [7]. Treat any rule you read as a starting point to be checked on your own equipment.
Key terms
- DfAM — design for additive manufacturing: shaping a part to suit (or exploit) an AM process.
- Restrictive DfAM — adapting an existing design so it can be printed.
- Opportunistic DfAM — using AM's freedoms to create designs that other processes cannot make.
- Overhang angle — the tilt of a downward-facing surface from the build plate; shallow angles usually need support.
- Teardrop hole — a round hole with a pointed top that can be printed without supports.
- Minimum feature size — the smallest wall, hole or gap a process makes reliably.
- Powder escape hole — an opening that lets loose powder leave a hollow part.
- Red flag — a situation named in standards as often causing problems in AM design.
Where this leads
The most striking opportunistic designs come from letting software find the shape: topology optimization and lattices.