Comparing AM Processes


Material Compatibility by Process

There is a broad spectrum of AM processes, materials, and related technologies. Understanding the key principles of each mainstream AM process is essential to design parts, products, and business strategies that leverage AM. More concretely, this information enables you to select AM processes for specific applications, and design integrated operations (e.g., including printing and post-processing) to meet application-specific needs.

That said, the AM process landscape is wide. Combined with the high rate of news announcements related to AM, it can become difficult to differentiate innovations from incremental advancements. Moreover, information on the quantitative performance of AM, such as material properties, build rates, and dimensional quality, is difficult to gather. Moreover, it is perhaps even more challenging to compare these quantitative metrics on an equivalent basis across different processes. Therefore, a strong foundation in the key principles of AM processes is essential to make these evaluations in your work.

The categorization of AM processes, established by the ASTM International Committee F42, is often shared as a standard classification to assist with navigating various process technologies and their permutations. Processes differ in the feedstock format, i.e. a solid filament, sheet, powder, or liquid. The material chemistry and energetics of transformation determine the resultant thermal trajectory of the process and influence the amount of energy required. While some processes command greater markets and applications than others at present, each process draws on important industrial needs.

A convenient means of summarizing the process spectrum, and identifying candidate processes for applications, is according to their materials compatibility, which is displayed in the table below.

In the table:

  • A full circle means that the commercially available machines using the noted process are capable of processing the noted material.
  • A hollow circle means the process-material combination has been demonstrated in research or pre-release commercial announcements, but machines are not yet shipping to customers.
  • Circles shaded black represent direct processes, i.e., the printing process produces a part with the desired final density and dimensions. Circles shaded blue are indirect processes, meaning that a densification step such as sintering is needed to give the part its final density and dimensions. That said, every AM process requires steps after printing.
  • The number notations for some process-material combinations refer to the notes at the bottom of this page, as these fall somewhat beyond scope.


  1. Massivit3D’s gel dispensing printing technology, which extrudes a gel followed by photopolymerization.
  2. Admatec’s stereolithography process that uses a metal-filled resin slurry.
  3. Lithoz, Admatec, and others, which use ceramic-filled resins and ceramic precursor resins.
  4. XJet’s technology inkjets metal nanoparticles in solvent, while direct metal jetting systems are in development by several companies.
  5. Ceramic LOM was previously commercially available; it is no longer on the market.
  6. See EnvisionTec and Impossible Objects.

Comparing Process Performance

The following charts compare the specifications of commercially available AM machines for the polymer AM processes of FFF/FDM, SLA, SLS, and Polyjet; and for the metal processes of SLM, EBM, BJ, BMD, and DED. Comparisons are drawn between machine cost (list sales price), machine build volume, estimated build rate, and estimated resolution. You may navigate the charts on your own, for instance by zooming, and selecting/deselecting data (when your mouse hovers over the chart, a toolbar will appear in the upper right corner).

You may choose to compare:

    • The relationships between build rate and machine cost for polymer and metal AM processes. Here, it is clear why some processes are more attractive for industrial uses, which can bear the large capital investment in exchange for much greater build rate. In addition, the build rate of machines within each process category generally scales with cost (and vice versa). For prototyping, desktop machines are much more cost-effective, albeit sometimes much slower than industrial systems.
    • The scaling of machine cost with build volume, showing universally that larger machines are more costly, because larger machines require more substantial structures and larger, more precise motion systems.
    • The comparative resolution values for each process, compared to its build rate. In process design or selection one may choose to tradeoff rate against resolution.

Source: MIT

Source: MIT

Source: MIT

Notes on the data collection and analysis:

  • Machine specifications are gathered from manufacturer websites, and costs are as listed in the 2018 Wohlers report.
  • Build rates are provided directly as the volumetric rate specified by the manufacturer (e.g., in units of cm3/hr), or converted from a linear vertical build rate (e.g., in cm/hr). In the latter case, the volumetric build rate is approximated from the build area (Abuild) and the vertical build rate (Rvertical) using the equation Rvol=0.25 x Abuild Rvertical.
  • Average layer thickness is used as an estimate of the process resolution; when a range of layer thickness values are provided in the machine specifications, the average of the minimum and maximum values is plotted above.