A marine customer came to us with a familiar problem: a critical piece of deck hardware had been discontinued, the original moulds were long gone, and commissioning a new injection-moulded tool would take months and cost more than the boats were earning in that time.
The brief
- Replace an obsolete hardware component used across a fleet of working vessels.
- Survive salt water, constant UV and daily mechanical load.
- Beat the cost of cutting a new injection mould.
- Ship in weeks, not quarters.
The constraint that mattered most wasn't cost. Vessels out of service don't earn, and the fleet had a finite stock of usable spares — perhaps a season's worth. The clock was the brief.
Step 1 — Capture the geometry
There was no CAD, no drawing and no supplier. What there was: a worn original, and a second one that had already failed.
We 3D scanned both. The worn part gave us the true mounting geometry — the interfaces that had to match the boat exactly. The failed part told us where the load actually goes, which is information no drawing would have contained. It had cracked at a sharp internal corner where the wall stepped down, a classic stress riser.
Step 2 — Rebuild the CAD, then improve it
Scan data is a mesh, not a model. We rebuilt the part as clean parametric CAD, which is what makes later revisions and second sizes cheap.
Two deliberate changes went in:
- A generous radius replaced the sharp corner that had failed.
- Wall thickness was increased slightly in the loaded section, where there was clearance to do it without touching any mating surface.
Every mounting interface was left exactly as scanned. The part had to drop into an existing boat with no modification — a reprint that requires drilling is not a reprint.
Step 3 — Material trial
Three candidates went into test prints:
- ASA — UV-stable and inexpensive, but not stiff enough under the working load.
- PA12 nylon — tough and durable, slightly too flexible for this geometry.
- Glass-filled PA12 — stiffer, dimensionally stable, and still tough enough to take an impact without cracking.
Glass-filled nylon by SLS won on every axis that mattered: stiffness, saltwater resistance, UV behaviour after dyeing, and cost per unit at the fleet's quantity. Nylon does absorb moisture, so parts were dried and dimensionally checked after conditioning rather than measured straight out of the powder bed.
Step 4 — Short production run
Once the customer had fit-tested a first article on an actual vessel, we printed the batch.
- Parts were nested to fill the powder bed, which is what makes SLS cost-effective at batch quantity rather than one-off.
- Each part was depowdered, bead-blasted and dyed black — the dye adds meaningful UV protection to an otherwise vulnerable surface.
- Critical dimensions were measured on every unit before packing.
The batch shipped inside three weeks of first contact.
The result
- Lead time: 3 weeks, against an estimated 16+ weeks for traditional tooling.
- Unit cost: well under the injection-mould break-even at the fleet's volume.
- Field performance: the printed parts remain in service with no reported failures.
- Reorders: the CAD now exists, so the next batch is a print job rather than a project.
That last point is the quiet one. The expensive part of this job was never the printing — it was recovering the geometry of a part nobody had documented. Once that exists, the fleet has a supply line for a component the market abandoned.
Would this work for your parts?
The pattern generalises to any fleet, machine or product line with parts the original supplier has stopped making:
- there is at least one surviving sample, even a broken one,
- the part is plastic or nylon and not safety-critical,
- annual volume sits below the tooling break-even.
If that's your situation, the break-even maths is here, and our enterprise service covers production runs and repeat batches. To start with a single sample part, just get in touch.
Specifics in this case study are shared with the customer's permission; the component and vessel type are kept general at their request.
The precision suction fixture shown with this article isn't a marine part — client work like this is usually covered by an NDA — but it's the same category of small-batch, tight-tolerance hardware we print for boat fittings and deck hardware.
