A reverse-engineered spare part made by 3D printing is a component recreated from a physical sample — captured by 3D scanning, rebuilt as a clean CAD model, then printed — when no drawing, supplier, or catalogue part number survives. At Layer X, this is one of the most valuable services we run from our Ahmedabad facility, because Gujarat's textile mills operate looms and winding machines that are decades old and long out of manufacturer support. This case study walks through one such job: an obsolete drive gear on a sizing machine at a Narol-cluster textile mill, which we scanned, redesigned, and delivered in PA12 nylon by selective laser sintering (SLS).
The problem: a stopped machine and a part nobody sells
Textiles are the backbone of Ahmedabad's industrial economy — the city has spun and woven cloth since the nineteenth century, and bodies such as the Ahmedabad Textile Industry's Research Association (ATIRA) still anchor the sector locally. That heritage has a hidden cost: much of the running machinery predates the digital-drawing era, and when a small component fails, the original equipment manufacturer may no longer exist, let alone stock the part.
Our client ran a sizing machine whose secondary drive train relied on an intermediate polymer gear. The original gear had cracked across two teeth and was slipping under load, halting the line. The OEM had been absorbed into a larger group two decades earlier, the parts catalogue was gone, and the only reference the mill had was the broken gear itself — worn, greasy, and missing a fragment. There was no drawing, no material spec, and no part number. Machining a one-off metal replacement was quoted at several weeks' lead time, and the mill was losing production every shift the machine sat idle.
Why the original was a polymer gear — and why we kept it polymer
The failed gear was not metal by accident. In this position the polymer gear runs quietly, tolerates minor misalignment, needs little lubrication, and acts as a deliberate sacrificial element — it is cheaper to replace a worn gear than to damage the shafts it drives. Switching to steel would have transferred wear and shock load onto costlier components upstream. So the engineering brief was clear: reproduce the geometry faithfully and stay in a tough, fatigue-resistant polymer. That pointed directly at PA12 nylon and, given the tooth detail and the need for consistent properties in every direction, at SLS rather than a layer-bonded process.
Step one: 3D scanning the worn sample
We began by capturing the physical gear with structured-light 3D scanning. Because the part was worn and partially broken, scanning alone could not give us the correct part — it gave us the as-failed shape, which is not the same thing. That distinction matters: reverse engineering a functional component is never a straight scan-to-print copy. The scan is evidence, not a blueprint. We treat the point cloud as the starting reference and rebuild deliberately, a discipline we describe in our guide to dimensional inspection with CMM and optical scanning.
The scan gave us hub diameter, bore, keyway, overall width, pitch-circle diameter, and enough surviving tooth flanks to work back to the gear's defining parameters.
Step two: rebuilding the gear in CAD
From the surviving teeth we reconstructed the gear's fundamentals — module, tooth count, pressure angle, and the involute profile — rather than tracing the worn outline. A gear is defined by its generating parameters, so recovering those let us regenerate a mathematically correct set of teeth, including the two that had cracked away and the tips that field wear had rounded off. We rebuilt the involute flanks to standard geometry, restored the root fillets, and re-cut a clean keyway to the shaft dimensions.
Two engineering decisions went in at this stage. First, we applied Geometric Dimensioning and Tolerancing to the bore, keyway, and pitch circle per ASME Y14.5, so the fit onto the shaft was controlled rather than eyeballed from a worn sample. Second, we accounted for the difference between the OEM's original moulded material and our PA12 — small dimensional and stiffness differences that change how the tooth meshes under load. We also thickened the tooth root slightly, within mesh limits, because the original had failed there. Reverse engineering is the moment to fix a known weakness, not faithfully reproduce it.
Step three: why SLS in PA12 was the right process
We printed the gear in PA12 (Nylon 12) using our SLS nylon 3D printing service. SLS fuses nylon powder with a laser and uses no support structures, so every tooth flank — including the internal keyway and both faces — comes out with consistent surface and consistent properties. That gives isotropic strength: the part is not weaker across the layers, which is exactly what a gear tooth in continuous mesh demands. A fused-filament gear would have carried a weak plane between layers, and a gear tooth is the last place you want one.
Our standard SLS specification holds ±0.2mm dimensional tolerance with a 4–6 day lead time, which for a bore, keyway, and gear mesh of this size is well inside the running clearance the assembly needs. Because SLS needs no supports and no tooling, we could go from approved CAD to a physical, functional gear in days — not the weeks metal machining would have taken. For readers weighing the powder-bed route in detail, our post on SLS design rules for PA12 nylon covers the wall thickness and clearance limits we worked to here.
Where PA12-GF and PA12-CF would change the answer
Standard PA12 was correct for this gear because the design intent was a quiet, slightly sacrificial element. Had the application demanded higher stiffness or lower creep under sustained load — a structural drive gear rather than an intermediate one — we would have moved to glass-filled PA12-GF or carbon-filled PA12-CF. The trade-offs between those grades are set out in our guide to glass-filled versus carbon-fibre nylon PA12. Matching material to duty is half of doing reverse engineering properly.
Process comparison: how we could have made this part
Before committing to SLS we weigh every viable route against lead time, cost, and fitness for the duty. The table below sets out the options we considered for this specific gear.
| Route | Tolerance | Typical lead time | Fit for this gear? |
|---|---|---|---|
| SLS — PA12 nylon | ±0.2mm | 4–6 days | Best fit — isotropic, no supports, keeps the sacrificial polymer design |
| FDM — Nylon PA12 filament | ±0.3mm | 3–5 days | Weak inter-layer plane through the teeth; rejected for a loaded gear |
| DMLS — metal | ±0.1mm | Longer, higher cost | Over-specified; shifts wear onto costlier shafts |
| CNC machining a blank | Tight | Several weeks (quoted) | Kept the machine idle far too long |
Tolerances for our metal and filament routes are quoted from our DMLS metal 3D printing and FDM specifications respectively; the SLS figures are from our nylon service page.
Validation before the machine ran again
A reverse-engineered part earns trust by measurement, not assertion. We verified the printed gear against the reconstructed CAD before it left us: bore and keyway to the shaft dimensions, pitch-circle diameter, tooth thickness, and overall width. We referenced ISO/ASTM 52900 — the international standard that defines additive-manufacturing terminology and process categories — so the SLS process was documented correctly in the job record under our ISO 9001:2015 quality system. Every reverse-engineering job at Layer X ships with a dimensional record, because a mill needs to know the replacement is right before it trusts it in a running line.
The gear went in, meshed cleanly, and the sizing machine returned to production the same shift it was fitted. The mill also now has a clean CAD model and a documented process on file with us — which means the next replacement is a same-day reprint at a known price, not another archaeology project. That is the real dividend of doing reverse engineering properly the first time: the second part, and every part after it, is trivial.
What this case tells other Gujarat manufacturers
Obsolete-spare reverse engineering is not a niche curiosity in a manufacturing state like Gujarat — it is a standing need across textiles, packaging, pharma machinery, and general engineering. The pattern repeats: an old machine, a failed low-cost part, no drawing, and an OEM that has vanished. The lesson from this job is threefold:
- A scan is evidence, not a blueprint. A worn sample must be rebuilt to correct geometry, not copied outline-for-outline.
- Reverse engineering is the moment to fix known failures — we strengthened the tooth root that had cracked, rather than reproducing the weakness.
- Match process and material to the original design intent. A sacrificial polymer gear stays a polymer gear; SLS PA12 delivers that with isotropic strength and no tooling.
This same workflow — scan, rebuild, print — recreates cams, bushes, guides, sprockets, housings, and end-effectors across industrial machinery. For a related functional-nylon example, see our SLS nylon gripper end-effectors case study.
Have an obsolete part with no drawing? Send us the broken component or a photograph, or upload your CAD file for a 24-hour quote — we reverse-engineer and print functional spares in PA12 nylon from our Ahmedabad facility, shipped pan-India.