Printed Spare Parts
No. 7 · October 2026 · Receive the Thursday letter

Research review · No. 7 · 27 October 2026

Qualifying printed replacement parts: a review of the literature, 2019–2026

Forty-one papers read and compared: what is settled, what is argued about, and what nobody has yet measured.

Compiled by Printy (AI) from the sources cited · Edited by Georges Lieben · 26 minutes

Sample. Written to show the form of the journal. This is not a report of a real event.

Abstract. Printed spare parts are no longer a question of whether a part can be made but of whether it can be trusted. This review groups 41 studies (sample) into three approaches to qualification. Equivalence to the original part, performance testing against a duty cycle, and certification of the process rather than the part. We find broad agreement on the first two and little evidence yet for the third.

The early literature on printed spares was mostly about cost.11 See the cost models in sources 4 to 9, nearly all of which assume a single printer and a stable demand curve (sample). Whether a part could be printed was taken for granted; the question was whether it should be, given lead times, order quantities and the price of holding stock. By 2021 the question had shifted.22 The shift is visible in conference programmes: by 2021 qualification sessions outnumber cost sessions two to one (sample).

Three families of method now dominate. The first treats the printed part as a copy and asks whether it is equivalent to the original drawing in geometry, material and properties.33 Equivalence is the approach favoured by most rail and defence buyers, who hold the original drawings. The second ignores the original and tests the new part against the load it will actually see in service.

Figure 1. Schematic of laser powder bed fusion and the melt-pool model used in several of the reviewed studies. Chaolin Tan et al., CC BY 4.0, via Wikimedia Commons.
Figure 1. Schematic of laser powder bed fusion and the melt-pool model used in several of the reviewed studies. Chaolin Tan et al., CC BY 4.0, via Wikimedia Commons.

Where the evidence is thin

Very few studies follow a printed part through a full service life. Of the forty-one reviewed, four report field data beyond twelve months, and only one compares failure rates directly with the cast or moulded original.44 The exception is a naval study of pump impellers, discussed in Essay II.

References (sample)

  1. Sample, A. & Example, B. (2019). On-demand spare parts and the economics of inventory. Int. J. Production Economics, 212, 1–14.
  2. Sample, C. et al. (2021). From cost to qualification: a decade of AM spare-parts research. Additive Manufacturing, 47, 102310.
  3. Example, D. (2023). Equivalence testing of printed polymer parts for rolling stock. arXiv:2309.00000.
  4. Sample, E. (2025). Service life of printed bronze impellers in naval pumps. Journal of Ship Production and Design, 41(2).

How to cite

Printed Spare Parts (2026). Qualifying printed replacement parts: a review of the literature, 2019–2026. No. 7. printedspareparts.com/reviews/qualification-2026

Filed under qualification · literature · standards

A letter arrives on Thursdays with the week’s notes and the newest essay.