Home/Guides/Replacing a discontinued part
Obsolescence & spares

Replacing a discontinued part: which path wins

When a part is discontinued, obsolete, or broken and there is no CAD, you have three routes: reverse engineer and reprint, bridge-buy or stockpile the remaining stock, or redesign and re-source. The right one depends on how many you need, how long you can wait, and whether you will need the part again.

Scan & reprint

Recreate the part from the part itself

  • No drawings, no CAD, no supplier — just the worn or broken part
  • You need a few, fast, and possibly again later
  • A functional engineering polymer is fit for the job
  • You want a permanent digital master you control
Bridge-buy / stockpile

Buy the remaining stock while it lasts

  • The exact original part is still available somewhere
  • It must be a certified or identical OEM component
  • You can predict how many you will ever need
  • It buys time to engineer a proper replacement
Redesign & re-source

Re-engineer it for ongoing supply

  • You need the part in volume, indefinitely
  • The original design was a weak point worth fixing
  • A modern equivalent or material upgrade exists
  • Tooling pays off across the quantity you need
01 · The fork

How do I replace a part with no CAD?

When a part is discontinued and there is no drawing or model to send a supplier, the decision splits three ways. Reverse engineer and reprint recreates the part from the physical object — scan or measure it, rebuild it in CAD, and print a functional replacement in days. Bridge-buy or stockpile means securing the remaining original stock before it disappears, buying time rather than solving the problem. Redesign and re-source re-engineers the part for ongoing volume supply, usually moving to moulding once quantities justify a tool.

These are not mutually exclusive. A common sequence is to bridge-buy a handful to keep running today, reverse engineer the part so you are never stranded again, and scale to re-sourcing only if the volume turns out to need it. The CAD created in step two carries straight into step three — reverse engineering a part once gives you both an immediate spare and a path to production.

02 · Signals

Which path wins, and when

Match the route to your real constraint — availability, quantity, and urgency decide it more than the part itself. The signals below point to the path that usually fits.

Reverse engineer & reprint
  • The part is discontinued or the supplier is gone
  • You have the part, a photo, or a sketch — but no CAD
  • You need one to a few, and need them soon
  • You will likely need the part again
  • A functional polymer meets the load and environment
Bridge-buy / stockpile
  • The original is still in stock somewhere
  • Certification or an exact OEM match is mandatory
  • Lifetime demand is small and predictable
  • You need breathing room to do it properly
  • The part is cheap relative to engineering it
Redesign & re-source
  • You need the part in volume, on an ongoing basis
  • The original kept failing and is worth improving
  • A modern equivalent or better material exists
  • Quantities justify tooling and a moulded run
  • Long-term unit cost matters more than speed now
03 · Scan & reprint

Reverse engineering a broken or obsolete part

This is the default route for legacy parts, because it needs nothing but the part. Send us the broken or worn original, a photo, or a sketch; we 3D-scan or measure it, rebuild it as a clean parametric CAD model, and print a functional replacement — usually within a few days. You walk away with both the part you needed and the CAD you never had.

A worn or broken sample is not a dead end. As long as enough geometry survives to infer the original shape, we reconstruct it — mirroring intact features, recovering dimensions from the undamaged side, and correcting the wear or fracture in CAD rather than copying it. Reverse engineering is a chance to fix the flaw that broke the part in the first place: thicken a thin web, add a fillet to a sharp corner, or move to a tougher material. Picking that material is its own decision — our PETG vs ABS vs ASA guide walks through it.

Best of all, the CAD model is permanent. Once a part is reverse engineered, it never goes obsolete again: reprint a spare on demand, send the file anywhere, or scale the same model to moulding later. One reverse-engineering job ends the dependence on a supplier who has moved on.

04 · Buy vs re-source

When buying or re-sourcing beats reprinting

Bridge-buy / stockpile wins when

  • The exact original is still available to buy
  • A certified or identical OEM part is mandatory
  • Lifetime demand is tiny and you can name the number
  • You just need time to engineer a real fix
  • Per-unit price is trivial next to design cost

Redesign & re-source wins when

  • You need the part in volume, indefinitely
  • The original was a recurring failure point
  • A modern off-the-shelf equivalent can be designed in
  • Quantities clear the crossover into moulding
  • You want one supply chain, not a hunt each time

Stockpiling only postpones the problem: the shelf eventually empties and you are back where you started, now under time pressure. Re-sourcing solves it permanently but earns its cost only at volume — the same volume-versus-tooling logic as our FDM vs injection moulding guide. The reverse-engineered CAD is what makes the jump cheap: validate the part as a print first, then move to a moulded run through our associated manufacturing company with no redesign in between.

05 · Accuracy

How accurate is a reprinted part?

Plan around the standard FDM accuracy band, and design critical fits rather than relying on as-printed dimensions. The numbers below are what a reverse-engineered, 3D-printed replacement holds in practice.

±0.2–0.3mm
Standard features
Typical FDM dimensional accuracy — or ±0.2–0.5% of length at large sizes, whichever is greater.
Critical fits: designed-in allowances + selective machining
2–5days
Scan to printed part
From a sample, photo, or sketch to a finished functional replacement — against weeks to re-source an obsolete part.
vs 2–4+ week tooling for a moulded run
60–90%
Z vs XY strength
A well-tuned, dried, hot-printed PETG, ABS, or PA part along the layer axis — orient the part to the load.

The honest limit: a reprinted part matches the original's fit and function, not necessarily its exact material or a regulated certification. Where a part must be identical for compliance reasons, that points back toward bridge-buying the OEM original — but for the vast majority of mechanical spares, a measured-and-rebuilt replacement in the right polymer does the job, often better than the part it replaces.

06 · The process

How reverse engineering works at PartForm

Five ordered steps take a legacy part from a broken object with no documentation to a reproducible, printable component you control.

Send the part, a photo, or a sketch
Send the broken or worn original, a photo, a sketch, or rough dimensions. No drawing or CAD file is required to start the conversation.
Capture the geometry
We 3D-scan or hand-measure the part to capture its geometry, recovering dimensions from a worn or damaged sample where the surfaces allow it.
Rebuild it in CAD
We rebuild the part as a clean, parametric CAD model — correcting wear and damage, and strengthening weak features rather than copying flaws.
Print and validate
We print the part in a suitable engineering polymer and check form, fit, and function against the real application before committing to a batch.
Reprint on demand or scale to moulding
The CAD model becomes a permanent digital master — reprint a spare any time, or scale the same model to moulding through our associated manufacturing company once volume justifies a tool.
07 · At a glance

Decision table

FactorScan & reprintBridge-buy / stockpileRedesign & re-source
Best quantityOne to a few, repeatableSmall, fixed, predictableVolume, ongoing
Speed to first partDaysImmediate, while stock lastsWeeks (tooling)
Needs original CAD?No — we create itNoNo — rebuilt from the part
Future-proof?Yes — permanent digital masterNo — stock runs outYes — engineered for supply
Lets you improve the partYes — fix the failureNo — same flawYes — full re-engineer
Best whenDiscontinued, no CAD, need it nowStill available, certification-boundHigh volume, recurring need
08 · FAQ

Common questions

How do I get a spare part for a discontinued machine?
Send the broken or worn part, a photo, or a sketch — no CAD or drawing is needed. We 3D-scan or measure it, rebuild it in CAD, and print a functional replacement in a suitable engineering polymer, usually within a few days. The CAD model is kept as a permanent master, so you can reprint the part any time and never depend on the original supplier again.
Can you reverse engineer a part if I have no drawings or CAD files?
Yes — that is the normal case. Most legacy and obsolete parts have no surviving documentation. We work from the physical part, a photo, a sketch, or rough dimensions, capture the geometry by 3D scanning or hand measurement, and rebuild a clean parametric CAD model from it. You end up with the CAD you never had, plus a printable part.
Can you reverse engineer a broken part?
Usually, yes. As long as enough of the geometry survives to infer the original shape, we can reconstruct a broken or worn part — mirroring intact features, recovering dimensions from the undamaged side, and correcting the wear or fracture in CAD rather than copying it. A second reference part or an assembly photo helps where the damage is severe.
How accurate is a reverse-engineered, 3D-printed replacement?
Plan around ±0.2–0.3 mm on standard features, or ±0.2–0.5% of length on large parts, whichever is greater — typical FDM dimensional accuracy. Critical fits get designed-in allowances and selective machining rather than relying on as-printed accuracy, so mating dimensions and bearing seats land where they need to.
Is it cheaper to reverse engineer a part or buy a replacement?
It depends on availability and quantity. If the part is still stocked and you need a few, buying is usually cheaper. Reverse engineering wins when the part is discontinued, the lead time to re-source is long, the minimum order is far more than you need, or you will need the part repeatedly — because the one-time CAD work then becomes a reusable digital master you can reprint on demand.
The honest trade-off

No single path fits every legacy part

Reverse engineering is the most flexible route and the one that future-proofs you, but it is not always the cheapest in the moment — if the original is still on a shelf and you need only a handful, buying them is simpler. And where a part must carry a regulated certification, a measured reprint matches the fit, not the paperwork.

The right path follows your real constraint — availability, quantity, and how long you can wait — and the three routes combine well. Send us the part, a photo, or a sketch and we will tell you honestly which one fits.

Discuss your part →