Skip to main content

Alloys for AM and the powder problem

📍 Where we are: Part IV · Materials and New Systems — Chapter 8 of 19. The last chapters were about watching and steering the process; now we look at what is being processed: the alloy and the powder it starts as.

The simple version

Think of cooking. A recipe written for a slow oven fails in a blowtorch. Most metal alloys were designed for casting or forging, where metal cools slowly and evenly. Laser melting is a blowtorch followed by an ice bath, thousands of times over, and many "recipes" simply tear themselves apart. The two ways out are to change the recipe (new alloys) or to control the ingredients more tightly (better powder). Both are harder than they sound, and the ingredients quietly change every time you reuse them.

What this chapter covers​

Two linked problems: which alloys can be printed without cracking and how new ones are designed, and how the powder behaves over repeated reuse. We follow laser powder bed fusion (PBF-LB/M) unless stated. Evidence is labelled lab-scale, production or contested.

Problem​

Only a few alloys are routinely printed. Martin et al. list AlSi10Mg, Ti-6Al-4V, CoCr and Inconel 718 as the alloys that can be reliably processed, whereas most others develop large columnar grains and periodic cracks during printing; high-strength aluminium alloys such as 7075 and 6061 hot crack severely [1]. A broad review of metal AM metallurgy likewise notes that only a few alloys have been developed for commercial production [2], and a perspective on laser powder bed fusion names high-entropy alloys and high-strength aluminium alloys as key materials developments [3]. The cause is solidification: steep thermal gradients and rapid freezing leave thin liquid films between growing grains that are pulled apart by shrinkage (Defects, residual stress and distortion, Microstructure, anisotropy and process-structure-property links). Separately, every build consumes and modifies powder, and the properties of a printed part depend on what that powder has become (Qualification and certification).

State of the art​

Alloy design for AM​

Nucleant-modified aluminium (lab-scale, with some follow-up). Martin et al. attached hydrogen-stabilised zirconium particles to 7075 and 6061 powders; the particles decompose in the melt and form Al3Zr, which nucleates fine equiaxed grains and removes hot cracking. Printed 7075 with the nucleant was reported to be within the expected range of wrought 7075, and the approach was proposed for other alloys and processes [1]. Later reviews describe Sc and Zr as grain-refining additions whose Al3(Sc,Zr) particles act as heterogeneous nucleation sites and help avoid thermal tear cracks, and note that earlier printable aluminium was limited to near-eutectic Al-Si alloys [4].

Oxide-dispersion-strengthened superalloy (lab-scale). NASA and Ohio State used thermodynamic (CALPHAD) modelling to design GRX-810, an NiCoCr-based alloy with nanoscale Y2O3 particles dispersed by laser powder bed fusion. At 1,093 °C the authors report about twice the strength and twice the oxidation resistance of the wrought nickel alloys widely used in AM, and more than 1,000-fold better creep performance: the hot isostatically pressed alloy ruptured after 6,500 h at 20 MPa, while the comparison alloys (including AM 718 and AM 625) ruptured in under 40 h [5].

Crack-resistance by design (lab-scale). For nickel superalloys, adding 1 wt% zirconium to Haynes 230 formed a Ni11Zr9 segregation network that eliminated cracks in one 2023 study [6]. The general idea in several of these cases is the same: change the freezing path or the nucleation so that the last liquid has somewhere to go.

Refractory metals and high-entropy alloys​

Refractory metals (W, Mo, Nb, Ta) and refractory high-entropy alloys (RHEAs) promise strength at very high temperature but are prone to cracking. A 2024 review reports that pure tungsten cracks because it is brittle; process optimisation reduced but did not eliminate cracking, whereas alloying to improve ductility and limit oxygen at grain boundaries worked better, and niobium alloys, being inherently more ductile, crack less [7]. For RHEAs, a 2025 critical review lists microcracks, pores and elemental loss as the main challenges [8]. One lab-scale result reports defect-free TiZrNbTa from in-situ alloying of elemental powders within a processing window that gives defect content low and properties comparable to as-cast samples [9].

For high-entropy alloys more broadly, a 2025 review notes that the ductile face-centred cubic CoCrFeMnNi is highly crack-resistant, while Al- or Ti-rich and many refractory compositions are prone to solidification or cold cracking. Under the very fast cooling of PBF-LB/M one alloy, AlCrFe2Ni2, solidified as a single B2 phase and cracked severely, while slower DED-like cooling allowed a ductile FCC phase to form and prevented cracks [10]. A December 2025 critical review states that PBF-LB/M gives the finest microstructures and highest strengths but also the highest residual stresses, and that electron-beam PBF is more crack-tolerant for refractory systems [8]. A 2024 review comparing seven classes of PBF-LB/M high-entropy alloys notes that computational modelling can describe crack formation, melt flow and segregation, which could speed up optimisation [11].

The powder problem​

A powder bed is reused: powder that was not melted is sieved and returned. Reuse changes the powder, and the changes differ by alloy:

AlloyWhat changes with reuseEvidence
316LSurface oxide thickens; Cr, Mn and Fe diffuse outward and the oxide shifts toward MnCr2O4; particle-size distribution shifts; oxygen picks up, and chamber oxygen level (300 vs 1000 ppm in one study) matters [12]Lab-scale (experimental PBF-LB/M studies)
Ti-6Al-4VAluminium lost by evaporation, oxygen rises, roughness increases while flowability can improve; fatigue improved with limited reuse upstream of the gas flow and fell downstream owing to large defects [13]Lab-scale
IN718In one study with a defined recycling procedure, powder recycled up to 14 times showed no significant changes in powder or test-part properties [14]Lab-scale

A comparison of virgin and long-used powders of IN718, Ti-6Al-4V, AlSi10Mg and Scalmalloy found AlSi10Mg the most sensitive to reuse, with changes in particle-size distribution and morphology and an increase in oxygen [15]; the effect of reuse times on Ti-6Al-4V has also been studied in electron beam melting [16]. A 2023 review covers characterisation, preparation and reuse of PBF-LB/M powders [17]. One study found a strong linear correlation between oxidation level and optical reflectance of powder for several alloys, suggesting a rapid screening method [12]. That is a lab-scale idea, not a standard test.

Why it is hard​

  1. The design space is huge and the test is slow. An alloy must be melted, printed, sectioned and examined for each composition; cracking depends on both composition and the thermal history of the machine.
  2. Printability and service properties conflict. Compositions that resist cracking (for example by adding grain-refining elements) must still deliver strength, ductility, creep or corrosion resistance. GRX-810 shows both can be reached, at the price of a nanoparticle-dispersion route [5].
  3. Powder state is path-dependent. Oxygen, moisture, particle size and morphology depend on how many builds, how much spatter, which gas and which sieving regime, so "reused powder" is not one material. Different studies find different effects (flowability improving while oxygen rises), which makes a general rule hard to state.
  4. Variability is not separable. Powder batch, atomisation route, machine and alloy interact. A change in fatigue life may come from powder, from gas flow position, or from both (Defects, residual stress and distortion).

Alloy and powder lifecycle: virgin powder, build with spatter and oxygen pickup, sieved reuse, and the printability trade-off between cracking and service properties. The powder loop: virgin powder is melted, spatter and oxygen pickup alter the remainder, and the sieved powder returns; new alloys must be printable and meet service properties. Original diagram by the authors, created with AI assistance.

Open questions​

  • Is there a general printability criterion that predicts hot cracking from composition, as a screening tool, across aluminium, nickel and refractory systems? Computational and high-throughput approaches are being pursued, but a universal rule is contested.
  • Can refractory and high-entropy alloys be made crack-free at production scale, or only in narrow lab windows such as the TiZrNbTa example?
  • What powder-state metrics (oxygen, particle size, flowability, reflectance) should define "end of life", and are they alloy-specific?
  • How much of the observed part-to-part variability in fatigue comes from powder state rather than process? (Machine learning for additive manufacturing cannot answer that without data that records powder history.)
  • How should reused powder be tracked in a digital thread (Digital thread, data and the supply chain)?

Industry impact​

The commercially mature set is still small: the alloys named above. New AM-specific alloys such as modified 7075 or ODS superalloys are demonstrated in the literature, but moving one into a qualified production part requires property databases and qualification (Qualification and certification) and that is a slow step. Powder reuse is a cost and sustainability lever (see Sustainability and life-cycle assessment) but also a qualification risk; users normally have to define and defend their own reuse rules.

Why it matters​

Materials are the ceiling of every other improvement. Better control or monitoring cannot rescue an alloy that is intrinsically crack-prone, and an excellent alloy is wasted if the powder drifts. The most useful research in this area is often unglamorous: measuring and recording powder state.

In the real world​

NASA's GRX-810 work and HRL's nucleant-modified aluminium are public examples of alloy design for AM; reuse studies on Ti-6Al-4V, 316L and IN718 are the practical background for any shop that recycles powder.

Key terms​

  • Hot cracking — cracks that open while the last liquid between grains is still present, driven by shrinkage.
  • Nucleant — a particle that gives growing crystals a place to start, producing small equiaxed grains.
  • Oxide-dispersion-strengthened (ODS) alloy — an alloy strengthened by a fine dispersion of oxide particles such as Y2O3.
  • Refractory metal — a metal with a very high melting point, such as tungsten, molybdenum, niobium or tantalum.
  • High-entropy alloy (HEA) — an alloy made of several principal elements in similar proportions.
  • CALPHAD — a thermodynamic-modelling method for predicting phases from composition.
  • Powder reuse — returning unmelted powder, after sieving, to a later build.
  • Oxygen pickup — the rise in oxygen content of powder or parts through oxidation.
  • Spatter — droplets and particles ejected from the melt pool, often oxidised.
  • Printability — how reliably an alloy can be built without cracking or other defects.

Where this leads​

Alloys can also be combined within one part. Multi-material, functionally graded and 4D printing shows what happens at the interfaces.