Key takeaways
- A lot carries about 3,480 seconds of fixed time — programming, sheet load, first-part check and deburr setup — so five parts cost 786 seconds each and a thousand cost 93.5.
- One 300 × 200 mm bracket alone on a 1,250 × 2,500 mm sheet uses 1.9% of the steel; a production nest uses 92%, so prototype material is a sheet purchase rather than a per-part cost.
- Fixed time falls from 88% of the work content at five parts to 3.7% at a thousand, which is why the same drawing should be re-quoted at each volume step.
- Keep the prototype flat pattern under the fabricator's control until the bend allowance is proven, or the first production lot will be cut to a K-factor nobody validated.
One file, two very different prices
The DXF that cuts the first sample cuts the ten-thousandth. What changes between them is not the machine and not the program: it is how much fixed work each part has to carry, and how much inspection the customer needs to see. Laser cutting for prototypes is cheap per drawing and expensive per part; the same route at volume is the reverse, and the switch between the two happens on a quantity that can be calculated.
The calculation is worth doing before a design review, because it is the only way to answer the question every programme asks in month two: should this part still be laser cut?
Take a typical bracket program. Programming and nesting is 20 minutes, loading the sheet 3 minutes, the first-part dimensional check 25 minutes, and setting up deburring 10 minutes. That is 58 minutes, or 3,480 seconds, of work that happens once per lot regardless of how many parts follow. Per-part touch time, cutting, bending, inspecting and packing, is about 90 seconds.
Five parts: 450 seconds of touch time against 3,480 seconds of setup, so 786 seconds of work per part, or 13 minutes. A thousand parts: 90,000 seconds of touch time against the same 3,480, so 93.5 seconds per part. The fixed share falls from 88.6% of the work content to 3.7%.
What actually changes as the quantity rises
Four things, and buyers usually plan for only the first.
The nest. A prototype of one part is cut from a full sheet, and the part uses a small fraction of it. A production run cuts forty or fifty parts per sheet. Material per part falls by a factor of forty, which is why a prototype price often looks unrelated to a production price even on the same geometry.
The fixture or the tool. Up to a few hundred parts, a laser blank plus a press brake with a standard tool is the route, and the bend positions live in the program. At higher volume, a dedicated bend fixture or a checking gauge pays for itself in cycle time and in scrap.
The inspection level. A prototype gets a dimensional check on the features that matter. A production lot gets a first article inspection record, and a PPAP where the customer's quality system requires it.
The packaging. One part ships in a box, a thousand ship on a pallet with a batch label. Handling is a line on the quotation at both ends, but it is a different line.
Route by quantity band
| Quantity band | Typical route | Inspection | Ganesh limit or note |
|---|---|---|---|
| 1–10 parts | Laser cut, standard brake tooling, no fixture | Dimensional check against drawing | Plate cut to 20 mm on the laser; no tooling cost |
| 10–100 parts | Laser cut, standard press brake tooling | First part plus sample check | 200 T, 5-axis brake, 4,450 mm bed, to 10 mm |
| 100–2,000 parts | Laser cut, dedicated fixture, inspected lot | FAI with dimensional report, PPAP on request | Fixtures designed and built in-house |
| 2,000–10,000 parts | Laser or press, decided on the tool business case | FAI plus in-process checks | Tooling quoted separately from piece price |
| Above 10,000 parts | Press blanking or progressive die | PPAP, ongoing capability checks | In-house tool room for press tools and dies |
| Any quantity, mixed assembly | Laser cut plus welded assembly | Assembly-level check | Three CLOOS cells, 30+ manual stations |
Worked: the prototype sheet is not the production nest
A 300 × 200 mm bracket has a net area of 0.06 m². On a 1,250 × 2,500 mm sheet, that is 3.125 m² of steel, so a single prototype part uses 1.9% of the sheet and 98.1% of it is remnant.
If the remnant is scrapped, the prototype carries the whole 49.1 kg of 2 mm CRCA. If it is retained, the next prototype, or the next revision of the same part, uses it. At a prototype stage where the design changes three times in a month, the retained remnant is often the difference between a viable development budget and a painful one.
The production nest is the same geometry: 48 parts on a 4 mm web, 92.2% yield, 1.02 kg of steel per part. The gap between 49.1 kg and 1.02 kg per part is not a machine improvement; it is a change in how the sheet is used.
That is why the nesting question and the quantity question are the same question. How nesting reduces sheet metal cutting cost works the yield arithmetic in full, and it explains why a prototype quotation that does not mention the remnant is incomplete.
The drawing should not change; the tolerance should
The most common prototype-to-production failure is not a process failure. It is a drawing that was dimensioned loosely for a sample and never tightened for a lot.
At prototype, ISO 2768-1 class m general tolerances are adequate because the part is being fitted by hand and adjusted on the bench. At production, the same part goes into a fixture and the assembly rejects anything that moves, so the functional dimensions need explicit tolerances and the non-functional ones can stay general.
The geometry should be frozen at the same revision that passed the first article. If it moves, the first article is void and the record belongs to the old revision. That sounds bureaucratic until the first warranty claim, at which point the revision number on the drawing is the only thing that settles it.
Where the part is welded after cutting, the assembly tolerance is not the sum of the cut tolerances; it is set by the fixture. That is covered in welded assembly tolerances, and it is worth reading before a prototype is approved on cut dimensions alone.
What laser cutting for prototypes has to prove
A prototype is not a small production batch. It is an experiment with three questions in it, and a first sample that answers all three saves a production lot of rework.
The first is the flat pattern. The bend deduction has to be confirmed against the tool that will form the part, because the K-factor in a CAD library is an assumption until a part comes off the brake and measures correctly. Ask the fabricator which K-factor was used, and check the formed part, not the flat blank.
The second is the feature behaviour around the bend. A hole 3 mm from a bend line either distorts or it does not, and the answer depends on the radius, the thickness and the grain direction. A prototype is the cheapest place to find out.
The third is the finish. Powder over a laser edge looks different from powder over a deburred edge, and a coated prototype shows the customer what they will actually receive. A bare-metal sample passes a design review and fails a visual standard.
Re-quote at each volume step
A price is only valid for the quantity band it was built for. A quotation for 25 parts assumes a shared sheet, a standard brake tool and a sample-level inspection; a quotation for 2,500 assumes a fixture, a lot inspection and a nest built for yield.
The practical arrangement is to ask for two prices on one drawing: the prototype quantity and the annual quantity, both stated in the same quote. That single document shows the customer what the volume step is worth, and it lets the buyer compare a laser route and a tooled route on the same basis instead of comparing a tool-free price with a price that quietly excluded the tool.
Where the annual quantity is a forecast rather than a purchase order, say so. A price break that depends on a forecast is an assumption, and it should be visible in the quotation rather than buried in a unit rate.
Packaging, from a box to a pallet
One part leaves in a box with a drawing number on it. A thousand leave on a pallet, strapped, with a batch label that ties the lot to the material certificate and the inspection record.
That transition is where parts get damaged rather than made badly. A coated panel stacked flat with no interleaving shows marks at the corners, and a long panel shipped upright in a rack needs a support at the middle of its span, not just at the ends. Prototype packaging hides these problems; production packaging exposes them.
Specify the packing method at the same time as the finish, because the two are decided together. The finished coated assemblies capability covers the interleaving and the packing used on coated work, and it is worth reading before a prototype is signed off as complete.
Inspection at each stage
A prototype needs a dimensional check, and the useful version of that check measures the features the next operation depends on rather than every dimension on the drawing. Hole positions relative to the bend line, the overall envelope, and the flatness of any face that will be welded.
The first production lot needs a first article inspection record: every dimension on the drawing, measured and recorded, with the material certificate and the process settings attached. PPAP documentation is available where the programme requires it, and it is prepared from the production fixture rather than the prototype bench.
In-process checks then carry the lot. On a laser-cut part the stable features are the cut profile and the hole pattern; the variable ones are the bend angles and the coating film. Check the variable ones on a sample per shift, and the stable ones on the first part of each nest.
Material traceability runs to mill certificates where the project asks for it. That matters more at production than at prototype, because a prototype made from stock is rarely audited and a production lot always is.
Prototypes and production lots on the same machines
Prototype and production parts run on the same 12 kW fibre laser with the 40 ft × 16 ft bed, 12,192 × 4,877 mm, supported by two 6 kW and two 3 kW machines. The bed size matters at both ends: a single long prototype panel and a full production nest are both single-sheet jobs.
Forming is on the 200 T, 5-axis CNC press brake with a 4,450 mm bed, Delem control, to 10 mm. Welding uses three CLOOS robotic MIG cells and more than 30 manual MIG/MAG, spot and SPM stations, with fixtures designed and built in-house, which is what makes the 100-to-2,000 band a fixture job rather than a repeat of the prototype.
Finishing runs through a 16 ft × 16 ft shot blast chamber, seven-tank pretreatment, an 18 ft × 14 ft powder booth, a wet paint line and a curing oven. Programmes run under ISO 9001:2015 and ISO 14001:2015. Published customer programmes include Mahindra, Mahindra Powerol, Godrej & Boyce, TKE, Eros Elevators, Minda, Gunnebo and Indian Railways, and the custom parts route covers the one-off end of the range.
What to send at each stage
- Prototype: STEP file, material and thickness, the dimensions that must be right, and the date the sample is needed.
- Pilot: the same file plus the assembly it fits into, and the tolerance that the assembly actually needs.
- Production: the released drawing with GD&T on functional features, the finish specification, the inspection level, and the annual quantity with the release pattern.
The laser cutting and fabrication hub collects the guides for each stage, and prototype to serial production for sheet metal parts covers the handover in more detail. The laser cutting cost estimator gives a first-pass split at any quantity, and the RFQ checklist lists the fields that change a price between a sample and a lot.
When laser cutting stops being the answer
Laser cutting is the right route while the geometry is moving, while the quantity is small, or while the part is large enough that a tool would be impractical. None of those conditions survives a stable design at high annual volume.
The switch point is a business case, not a rule. A blanking tool removes the pierce cost and recovers the strip more efficiently, but it costs money and time, and it locks the geometry. Stamping versus laser cutting by annual quantity works the crossover in detail, and how to choose laser cutting for a new part covers the first decision that precedes it.
We do not cut above 20 mm on the laser, we do not run a tube laser, and we do not produce formed features in a cut profile. Where a volume ramp needs any of those, the route changes to another process, and the sooner that is said the cheaper the change is.
Frequently asked questions
Is laser cutting economical for a single prototype part?
It is the cheapest route for one part, because nothing has to be tooled. You pay for the whole sheet unless the remnant is kept, so a single bracket can look expensive per piece while the order still costs less than a press tool. Ask for the remnant to be retained against the next release.
When should a laser-cut part move to a press tool?
When the geometry has stopped moving and the annual quantity is in the thousands. The crossover depends on the tool and the part, but a stable bracket with no formed features usually justifies a blanking tool somewhere between two and five thousand pieces a year.
What changes on the drawing between prototype and production?
Tolerances tighten to the ones the process can hold, the flat pattern moves to the fabricator, GD&T replaces coordinate dimensioning on the functional features, and the finish and inspection requirements become explicit. The geometry should not change at all.
Do prototypes need a first article inspection?
A prototype needs a dimensional check against the drawing, not a full PPAP. The full first article record belongs to the first production lot, because that is the first part made with the production fixture, nest and process settings.
Can the prototype supplier be different from the production supplier?
It can, but then the production supplier has to re-prove the flat pattern and the bend allowance, and any tolerance that was quietly relaxed in the prototype is discovered again. Moving the part costs one first article, and it is worth budgeting for it explicitly.
Sources
- ISO 9013:2017 — Thermal cutting: classification of thermal cuts, geometrical product specification and quality tolerances · 3 Nov 2026
- ISO 2768-1:1989 — General tolerances, Part 1: linear and angular dimensions without individual tolerance indications · 3 Nov 2026
- ISO 9001:2015 — Quality management systems: requirements · 3 Nov 2026
Quote the same drawing at two volumes
Send the STEP file with a prototype quantity and an annual quantity. We will price both routes on the same drawing, so the cost of the volume step is visible before you commit to a tool.
Request a quote