Who this is for
Metal-fabrication job-shops, OEM component suppliers, structural-steel producers, and engineering businesses planning a new cell or a capacity expansion in Africa. If your question is "what do I import, what do I build or buy locally, and how do I size the machines?", this is the framework.
The core principle: import precision, localise bulk
Metal fabrication is one of the clearest cases of the hybrid sourcing logic. The precision machines — laser, press-brake, CNC — are mature global product classes where imported units offer real performance-per-rand that local fabrication cannot match. The bulk, structural, and service-sensitive scope — handling, cranes, frames, extraction, MCC — is heavy, low value-per-kilogram, and far cheaper and faster to source locally.
Get this split right and you buy world-class capability at the cutting edge while keeping the surrounding scope cheap, fast, and serviceable. Get it wrong — importing the steelwork or localising the laser — and you pay too much for the wrong things. The underlying decision framework is in Buy from China or fabricate locally?.
Decision rule: if it is a catalogued precision machine with a global field record (laser, press-brake, CNC, turret punch), import it. If it is heavy steel, handling, enclosures, or a service-sensitive utility, build or buy it locally.
The scope, split by what to import vs localise
| Cell element | Default sourcing | Why |
|---|---|---|
| Fibre laser cutter | Import | Specialist; mature class; performance/price unmatched locally |
| CNC press-brake | Import | Specialist; CNC backgauge and tonnage precision |
| CNC machining centre / turret punch | Import | Specialist precision; global product class |
| Welding power sources | Local dealer (named brand) | Local service network and consumables; brand support |
| Welding cells / positioners / jigs | Mixed | Robotic cells imported; jigs and fixtures local |
| Sheet / plate handling, cranes | Local | Heavy steel, low value/kg, freight-uneconomic, local service |
| Dust / fume extraction and filtration | Local specialist | Local service; component spares locally supported |
| Compressed air / nitrogen / gas supply | Local dealer | Local service and supply contracts |
| MCC, distribution, surge protection | Local panel builder | Local components; maintenance team can work on it |
| Cell PLC / overarching controls | Local SI (Siemens/AB) | Local skill base for ongoing changes |
| CAM / nesting software seats | Local + OEM support | Embedded in operator team; not vendor-locked |
| Tooling (press dies, laser optics buffer) | Mixed + local buffer | Pre-positioned spares for first 12 months |
Our metal-fabrication cell case study is a worked example of exactly this split — imported laser, press-brake, and welding cell; locally fabricated handling, crane, and MCC.
Sizing the precision machines
Fibre laser cutter — power and bed size
Two numbers decide the laser: source power (kW) and bed size.
- Power sets the maximum thickness you can cut at productive speed. Indicative: 1–2 kW for thin sheet; 3–4 kW for general fabrication (most African job-shops); 6 kW+ for thicker plate and higher speed. Spec for the thickest material you will cut regularly, not the occasional outlier.
- Bed size (e.g. 3 m × 1.5 m, 4 m × 2 m, 6 m × 2 m) sets the largest sheet you can process. Match it to your stock sheet sizes.
Don't over-buy power: a 6 kW laser costs and consumes more than a 3 kW for thin-to-medium work it will never need. Don't under-buy bed: a small bed forces wasteful sheet cutting and limits the parts you can nest.
CNC press-brake — tonnage and length
Two numbers again: bending tonnage and working length.
- Tonnage depends on material thickness, bend length, and die opening. A rough rule: bending mild steel needs roughly proportional tonnage to thickness and length — a 3 m bend in 6 mm mild steel needs on the order of 150–200 t, depending on the die. Spec for your heaviest regular bend.
- Working length must cover your longest part. A 3 m brake cannot bend a 4 m part.
A CNC backgauge (multi-axis) is worth it for repeatable multi-bend parts — it is a major part of why imported press-brakes earn their place over manual local equivalents.
Failure mode: sizing the laser and brake for the contract on the table, then turning down the next three jobs because the machine can't cut the thickness or bend the length. Spec for the 5-year product mix, with margin on the parts you expect to grow into.
The controls and software decision
Imported machines arrive with their own controllers and CAM/nesting software. Two things matter for African operations:
- Local SI support for the cell-level controls. Keep any overarching cell PLC on a platform (Siemens, Allen-Bradley) that your local system-integrator pool can service and extend. Avoid closed, proprietary cell controls that lock you into a single offshore vendor.
- Embedded CAM/nesting skills. The CAM seat and nesting know-how must live in your team, not only with the OEM. This is the difference between a cell you can program for new work and one that stalls when the OEM is slow to respond.
Spares and consumables — plan, don't improvise
The precision machines have consumables and wear parts a continent away from the OEM: laser optics and nozzles, press-brake tooling, welding consumables. A buffer kit specified at contract stage — sized for the first 12 months — avoids the 14-day air-freight emergency that stops the cell. See Spare-parts strategy for an imported Chinese line for the full discipline.
Utilities and site readiness
- Power and surge protection — lasers in particular are sensitive to grid quality. Budget surge protection and a clean-restart strategy for the African grid reality.
- Cutting gas / nitrogen — laser cutting needs reliable gas supply; twin-tank changeover and a service contract avoid mid-cut stoppages.
- Floor flatness and foundations — press-brakes and machining centres have level requirements; specify the floor before the machine lands.
- Extraction and fume management — a safety and compliance requirement, best sourced and serviced locally.
B-BBEE and local content
Because so much of a fabrication cell is legitimately local (handling, MCC, extraction, engineering services, jigs), a well-structured cell can carry a strong local-content share — useful where tender position depends on it. See B-BBEE and local-content scoring. Don't force the precision machines local to chase the scorecard, though — that produces an inferior cell.
How to scope a cell — a short sequence
- List your 5-year product mix — materials, thicknesses, lengths, tolerances, volumes.
- Size the precision machines for the most demanding part in that mix, with margin.
- Split the scope — import precision, localise bulk and service-sensitive scope.
- Fix the controls platform on a locally supportable standard; embed CAM skills.
- Plan spares, consumables, utilities, and floor as core scope.
- Calculate local-content share if it affects your tender position.
What CISH does in this part of the process
For metal-fabrication projects, we size the precision machines for your real product mix, structure the import/local split, keep controls on a serviceable platform, and plan spares and utilities. See Metal fabrication production lines, our cell case study, and the supplier audit checklist for vetting an imported-machine supplier.
Frequently asked questions
Can a local fabricator replicate an imported press-brake or laser?
Generally not at the precision, repeatability, and reliability the work needs. These are mature global product classes. Local fabrication shines on handling, structures, jigs, and enclosures — not on the precision machines themselves.
What laser power do most African job-shops need?
3–4 kW covers most general fabrication well. Go higher only if you regularly cut thick plate or need higher speed at volume; go lower only for dedicated thin-sheet work. Match bed size to your stock sheets.
How much press-brake tonnage and length do I need?
Size for your heaviest regular bend (thickness × length × die) with margin, and a working length covering your longest part. A CNC backgauge pays off for repeatable multi-bend work.
Should the cell controls be imported or local?
The machines come with their own controllers, but keep any overarching cell control on a locally supportable platform (Siemens/AB) and embed CAM/nesting skills in your team. Avoid closed proprietary cell controls.
How do I avoid laser/press downtime from spares?
Specify a 12-month buffer of optics, nozzles, tooling, and consumables at contract stage, held locally. The emergency air-freight of a critical consumable costs far more than buffering it.