Skip to main content

Open-source metal and large-format printing

📍 Where we are: Part V · Beyond the Desktop — Chapter 14 of 16. Chapter 10 closed a small plastic loop. Here the open-source idea is pushed to its edges: metal, and very large parts.

The simple version

A desktop printer is like a hot glue gun on rails. Two extensions are natural. Swap the glue for a welder, and you can lay down steel. Swap the glue stick for a bucket of plastic pellets and a big gantry, and you can print furniture. Both exist as open designs. The reason they are not in every garage is the same reason you do not weld in the living room: they are more dangerous, more finicky and less forgiving of mistakes.

What this chapter covers​

We survey three families: printers built around a welding torch (GMAW), open laser powder bed fusion research machines, and pellet-fed large-format printers. For each we say what is published, what has been demonstrated, and how mature it is. The maturity labels are mine, based on the sources cited; they are judgements, not official ratings.

Welding-based metal printing (GMAW)​

GMAW (gas metal arc welding, often called MIG) feeds a metal wire into an electric arc and melts it onto the part. If you move the torch or the table under computer control, you build a part bead by bead. This is a form of directed energy deposition (see Book 2); the arc-and-wire variant is usually called WAAM. A foundational review describes WAAM as using arc welding tools and wire to make large components (over 10 kg) with high deposition rates and low material and equipment costs, aimed at low-to-medium complexity parts [1]. It is coarse compared with laser powder bed, but cheap, because a commercial welder does the hard work.

Michigan Tech's open-hardware group published the first design in 2013: a delta-style RepRap carrying a welder torch, controlled by open electronics [2]. A 2018 paper converted a CNC router kit and a MIG welder into a similar machine and demonstrated applications such as repairing a part and making tool-like objects; it is open access and links its design files [3]. Supporting open pieces include a voltage and current monitor for the weld [4]. The 2013 paper provides a bill of materials and construction and operating procedures; I did not verify a total cost, so check the paper for the version you intend to build.

Maturity: research and advanced hobby. Parts are rough, surfaces need machining, and the heat, UV light and fumes require welding-grade safety. The WAAM literature stresses that residual stress, distortion and defects such as porosity must be managed to get reliable parts [1] [5] [6]. I found no evidence of qualified production use of these open machines. The open GMAW papers [3] [4] and the 2013 printer [2] are the Michigan Tech group's own publications and are modestly cited; they are kept because they are the primary sources for these designs, and the established WAAM reviews above carry the general claims.

Open powder-bed research rigs​

Industrial laser powder bed fusion (LPBF) machines are closed: you cannot change the laser scan strategy or read raw sensors freely. That matters because moving LPBF from prototypes to critical parts is hindered by a lack of confidence in part quality, which the authors of a widely cited review say will come from understanding the process physics through modelling and simulation [7]. A broader review of metal AM treats process, structure and properties together [8]. Two open efforts address the access problem.

DTU OpenAM. A team at the Technical University of Denmark published an open LPBF system with a design paper [9] and a hardware repository under the permissive CERN-OHL-P v2 licence. The README gives a 250 by 150 mm build area, a 300 W laser at 1080 nm and a closed-loop gas flow [10]. The paper's abstract states that it is functional for aluminium, tool steel, titanium and stainless steel and that the system has 1,500+ print hours; these are the authors' own statements, and I have not seen logs. The README says the paper describes version 1 and that version 2 supersedes it, so the hours do not automatically transfer to the v2 design in the repository. The same README states plainly that it is an experimental R&D setup whose safety aspects have not been approved by a certified official, that the electrical documentation is still to come, and that the control software lives in other repositories. A high-power laser and fine metal powder are serious hazards; this is not a kit for a home workshop.

reAM250. A group around TUM presented a modular, open-architecture research platform for process monitoring and control in laser PBF [11]. The paper reviewed existing self-built research systems to formulate requirements, and aims to ease sensor integration and give full control over process parameters; its stated motive is that commercial machines limit access, that the effort of building such systems is an entry barrier for new researchers, and that the lack of standardization limits the transferability of results. It is a 2024 conference paper, so treat it as recent and lightly cited; like the DTU design paper, it is the project's own publication.

Maturity ladder for open-source metal and large-format printing: GMAW printers, pellet-fed gantries, and open powder-bed research rigs, each marked with what is published and what is not yet demonstrated. Three open routes beyond the desktop, grouped by what is published; none is certified for safety-critical production. Original diagram by the authors, created with AI assistance.

Pellet-fed and large-format printing​

Filament is a convenient but expensive and small-volume form of plastic. Fused granular fabrication (FGF), also called fused particle fabrication, feeds plastic pellets or ground granules straight into a screw extruder, which is cheaper per kilogram and works with ground recycled plastic. Pearce's group published a 3-D printable pelletizer chopper that makes pellets of several sizes for FGF machines; it reports a US$185 build and 0.5 kg/h with one motor [12].

A concrete large-format example is a Hangprinter-based machine: a cable-driven printer (cables replace the rigid frame, so it scales to big volumes) with a pellet extruder and a hopper that holds up to 5 kg. A HardwareX paper describes the design, electronics (a Duet3 controller running RepRapFirmware) and files on OSF, and quantifies accuracy and strength [13]. For context, a 2023 review of polymer large-format AM covers the system configurations, controls, materials and application sectors [14]. The paper lists the hardware cost as CAD 2,125 (US$1,638). It reports a tensile strength of about 37 MPa for its printed parts and judges that within the range of other systems, while noting feedstock and printhead effects; read the paper for the full table.

Maturity: working research and maker prototypes. Pellet extruders themselves are commercially common, but open, well-documented, calibrated large machines are fewer.

Open DED and WAAM controllers​

I looked for open controllers for wire-arc additive manufacturing (WAAM) and DED. What I could verify is the Michigan Tech stack around GMAW printing (open firmware, a modified slicer and a weld monitor) in [3] and [4]. I could not verify a widely adopted, maintained, general open WAAM controller, so I do not claim one exists.

Honest assessment​

RoutePublishedDemonstratedMain gap
Open GMAW printerPaper, design filesSimple steel parts, repairsSurface quality, safety, skill
Pellet or Hangprinter FGFPaper, filesLarge plastic printsCalibration, feedstock consistency
Open LPBF (DTU)Repo, paperMulti-material research buildsSafety approval, electronics docs
Open research LPBF (reAM250)Paper, software, CADMonitoring and control researchNot a production machine

Why it matters​

Open metal rigs matter most to researchers who need to change what the machine does. A closed industrial machine is a good tool; an open one is a good instrument. Large-format open printers matter where shipping a part is costly.

In the real world​

Expect these machines in universities and labs, not on factory floors. The most useful thing a non-expert can do is read the build documents and their stated safety caveats before spending money.

Key terms​

  • GMAW — gas metal arc welding (MIG); wire melted by an arc, usable as a deposition source.
  • DED — directed energy deposition: material is fed and melted at the point of deposition.
  • WAAM — wire arc additive manufacturing: DED using an arc welder and wire.
  • LPBF — laser powder bed fusion: a laser melts metal powder layer by layer.
  • FGF — fused granular fabrication: printing from pellets or granules instead of filament.
  • Hangprinter — an open cable-driven printer design whose size is set by its cables, not a rigid frame.
  • CERN-OHL-P — a permissive open hardware licence from CERN.
  • Research rig — a machine built to be modified and instrumented rather than to produce parts.

Where this leads​

Research machines need research tools. Next: Open tools for research: data, simulation and monitoring.