Local or Offshore Aluminium Casting? The Answer Changes With Your Part, Not Your Politics
A Decision Usually Made at the Wrong Level
Most organisations answer the local-versus-offshore question once, at company level, and then apply it to everything. Either "we source in Asia because it is cheaper" or "we buy domestically because imports are risky."
Both positions are wrong for roughly half the parts they cover.
Sourcing route is a per-part decision driven by four measurable variables: annual volume, part value density, demand volatility, and how much engineering iteration the component still needs. A high-volume Die Cast housing with frozen geometry and stable demand behaves completely differently from a low-volume structural bracket that changes twice a year.
The buyers who get this right run both routes simultaneously and allocate each part to the one that suits it.
Total Landed Cost: The Number That Should Drive the Decision
The ex-works quotation is the only number most comparisons use, and it typically represents around two thirds of what an imported casting actually costs to have on your shelf, ready to build.
A complete landed cost model contains eight elements:
| Element | Typical Share | Notes |
| Ex-works piece price | 60–70% | The only number in most quote comparisons |
| Ocean freight and handling | 5–15% | Volatile; rises sharply during capacity shortages |
| Import duty | 0–10% | Depends on HS classification and trade agreements |
| Inventory carrying cost | 3–8% | Longer pipelines require more stock, funded by you |
| Quality risk provision | 2–8% | Sorting, rework, scrap, and inspection labour |
| Tooling amortisation | Variable | Divide by realistic three-year volume, not year one |
| Engineering and travel | 1–3% | Audits, sample reviews, problem resolution |
| Expedite contingency | 0–5% | Air freight when the vessel schedule fails you |
Three of these are consistently underestimated.
Inventory carrying cost. A 45-day sea transit plus safety stock means you finance perhaps 60–90 days of inventory that a nearby supplier would deliver weekly. On a mid-value component at typical financing rates, that quietly erodes a large slice of the piece-price advantage.
Quality risk provision. Not a judgement about capability — a statement about distance. When a batch is wrong, a local supplier collects and replaces it within days. An imported batch means sorting the stock you already hold, air-freighting replacements, and waiting weeks for corrected production. Cost that in advance rather than absorbing it as an unplanned event.
Expedite contingency. Every long pipeline eventually needs one air freight shipment. Budget for it as a recurring probability, not an exception.
The practical test: build the model once for a representative part. If the landed gap is under roughly 15%, the offshore advantage is thin enough that lead time and responsiveness should decide. If it exceeds 30%, offshore is compelling provided the part is stable.
Pipeline Length Is a Design Constraint
Distance does not only add days. It adds inertia — the delay between deciding something must change and that change appearing at your goods-in door.
A local casting services pipeline runs production, road transport, goods-in. Perhaps two to four weeks end to end.
A long-distance pipeline runs production, consolidation, port booking, ocean transit, customs clearance, inland haulage, goods-in. Ten to sixteen weeks, plus safety stock at the end.
The consequences are structural rather than logistical:
Demand response lag. A 40% volume increase reaches you in twelve weeks offshore versus three weeks locally. In a growing product line, that gap is lost revenue, not just inconvenience.
Engineering change latency. A design revision must flush through everything already in the pipeline. With three months of stock in transit and in store, a change agreed today reaches customers next quarter. Parts still undergoing iteration are poor candidates for long pipelines for exactly this reason.
Quality escape exposure. If a defect is discovered at final assembly, everything produced since the root cause began is already made and largely already shipped. Distance converts a contained problem into a large one.
Working capital. Buffer stock is not free storage. It is your cash, sitting in a warehouse or a container, unavailable for anything else.
None of this argues against offshore sourcing. It argues for matching pipeline length to demand stability. Frozen geometry with predictable volumes tolerates a long pipeline comfortably. Anything volatile does not.
Which Parts Belong Where
Four characteristics decide the allocation. Score a part against each and the answer usually becomes obvious.
| Characteristic | Favours Local | Favours Offshore |
| Annual volume | Under 5,000 | Above 20,000 |
| Value density (£/kg) | Low — freight dominates | High — freight is marginal |
| Demand volatility | High or seasonal | Stable and forecastable |
| Design maturity | Still iterating | Frozen, PPAP complete |
| Cosmetic criticality | Class A visible surfaces | Functional or hidden |
| Assembly complexity | Multi-component sub-assembly | Single part, simple pack |
| Regulatory burden | Aerospace, defence, medical | General industrial |
Value density deserves particular attention and is routinely ignored. A heavy, low-cost casting carries freight cost that can approach its own manufacturing cost. A compact, high-value aluminium casting carries freight as a rounding error. Two parts from the same drawing package can genuinely belong on different continents for this reason alone.
Two allocations that are almost always correct:
Low-volume structural work stays local. Weldments, frames and brackets produced by structural fabrication are bulky, low in value density, and frequently subject to late change. Shipping air inside a container to save labour cost rarely computes.
High-volume, small, detailed components go offshore. Especially Zinc Casting parts, where tool life reaches 500,000–1,000,000 shots, cycle times are short, and the value-to-weight ratio is favourable. This is the clearest case in the whole matrix.
Running Two Sources Without Doubling the Work
Dual sourcing is usually described as insurance. Its more valuable function is leverage — a supplier who knows a qualified alternative exists behaves differently in every subsequent negotiation and every quality conversation.
It is also frequently mismanaged. Three principles keep the cost proportionate.
Do not dual-source everything. Qualify a second source only for parts where a supply interruption stops your line, or where annual spend is large enough that price leverage matters. For everything else, single-source and accept the exposure.
Qualify before you need it. Second-source qualification conducted under pressure — after a failure, with production stopped — costs more and produces a worse outcome. Run it as a planned project when nothing is on fire.
Compare parts, not certificates. Run identical Die Casting Parts from both sources through the same CMM programme and produce a side-by-side dimensional comparison. Two suppliers holding the same ISO 9001 certificate can produce measurably different components, and only measurement reveals it.
Tooling Portability
The single practical blocker to dual sourcing is tooling. A die designed for one machine and one plant's process window may not transfer cleanly. Four provisions make it possible:
Standard die base and shot sleeve dimensions
- rather than plant-specific fittings
The full 3D die model
- including cooling circuits, as your property
Documented qualified shot parameters
- — velocities, intensification pressure, die temperature — which are the process knowledge that makes the tool work
A written release clause
- obliging the current holder to release the die within a defined period without preconditions
Without the third item, a transferred tool arrives at the new supplier as a lump of steel with no process history, and the receiving foundry rediscovers the window from scratch — typically several weeks of trials.
A pragmatic middle path: run the primary tool offshore for volume, and hold a bridge tool locally for emergency production at low output. It costs a fraction of a second production tool and eliminates the total-stoppage scenario.
Hybrid Allocation Within a Single Assembly
The most efficient structure is often not one route for the whole assembly.
A typical enclosure product might comprise a complex die cast housing with thin walls and integrated fins, a machined interface plate, several fabricated mounting brackets, and fasteners. Forcing all of it through one route is rarely optimal.
Housing offshore
- — high tooling amortisation benefit, stable geometry, good value density
Brackets local
- — bulky, low value density, still subject to change, and
custom metal fabrication
needs no tooling
Final assembly local
- — near the customer, so configuration changes and regional variants are handled at the last possible moment
This structure has a further advantage. It delays commitment. Components arrive as inventory that can be built into several product variants, rather than as finished assemblies committed to one configuration months before the order exists.
The requirement it creates is interface discipline. When cast and fabricated components meet at assembly and originate from different suppliers on different continents, tolerance stack-up must be analysed properly and datums must be unambiguous. Two suppliers each working within their own tolerance can still produce parts that will not fit each other.
Managing an Offshore Route Properly
If a part is allocated offshore, four controls determine whether it performs as modelled.
Verify the production site, not the trading entity. Confirm that certificates name the actual foundry rather than an affiliated sales company, and verify them with the issuing registrar. Ask specifically which operations happen on that site and which are subcontracted — machining and finishing are frequently outsourced, adding lead time and a second party to any quality claim.
Specify the document package as a delivery condition. Material certificates per heat, full dimensional reports with actual values, treatment records and NDT results. Written into the purchase order, not requested afterwards.
Specify packaging against destination climate. A 45-day sea transit produces daily condensation cycles. Machined aluminium faces need VCI film and desiccant sized to enclosed air volume. Arriving with corrosion converts an on-time shipment into a late one.
Agree corrective action terms in advance. Acknowledgement within 24 hours, containment within 48, a documented 8D within 5–10 working days, and supplier-funded freight in both directions when the fault is theirs. Negotiating this while a line is stopped puts you in the weakest possible position.
Frequently Asked Questions
Q1: Should I source aluminium castings locally or overseas?
Decide per part, not per company. Offshore favours high volume above roughly 20,000 units per year, high value density, stable forecastable demand and frozen design. Local favours volumes under 5,000, bulky low-value-density parts where freight approaches manufacturing cost, volatile or seasonal demand, and components still undergoing engineering change. Build a total landed cost model for a representative part: if the gap is under 15%, responsiveness should decide; above 30%, offshore is compelling provided the design is stable.
Q2: How do I calculate total landed cost for imported castings?
Add eight elements to the ex-works price: ocean freight and handling, import duty per HS classification, inventory carrying cost for pipeline and safety stock, a quality risk provision for sorting and rework, tooling amortisation divided by realistic three-year volume, engineering and audit travel, and an expedite contingency for the air freight every long pipeline eventually needs. Ex-works typically represents only around two thirds of the real figure, which is why quote-to-quote comparisons systematically favour distant suppliers.
Q3: How much buffer stock does an offshore casting supply need?
Enough to cover the full replenishment pipeline plus demand variability. A route running production, consolidation, ocean transit, customs and inland haulage typically spans 10–16 weeks, so buffer stock commonly reaches 60–90 days against perhaps 10–15 days for a local supplier. That inventory is your working capital sitting idle, and it should be costed into the landed model rather than treated as a warehouse detail.
Q4: How do I dual-source a die casting without doubling my costs?
Be selective and be early. Qualify a second source only for parts where interruption stops your line or where annual spend justifies price leverage. Run the qualification as a planned project rather than an emergency response. Compare actual parts from both sources through the same CMM programme instead of comparing certificates. Where a full second tool is not justified, a low-cost bridge tool held locally provides emergency capacity at a fraction of the cost.
Q5: Can I move my tooling from one supplier to another?
Only if you planned for it. You need four things: ownership stated explicitly in the contract, a written release clause obliging the holder to release the die within a defined period without preconditions tied to commercial disputes, the full 3D die model including cooling circuits as your property, and the documented qualified shot parameters. Without the parameters, the receiving foundry must rediscover the process window through trials, typically costing several weeks. A tool without its process data is worth considerably less than a tool with it.
Q6: Which parts should stay local even when offshore is cheaper per piece?
Bulky low-value-density parts where freight approaches manufacturing cost, particularly weldments and frames from structural fabrication. Parts still undergoing design iteration, because a long pipeline means changes take a quarter to reach customers. Components with Class A cosmetic surfaces, where a colour or finish dispute is far cheaper to resolve locally. Anything with heavy regulatory documentation burden. And any part where demand is volatile enough that response speed outweighs unit cost.
Q7: Does splitting an assembly across two routes create quality problems?
It creates a specific engineering requirement rather than an inherent problem. When cast and fabricated components originate from different suppliers and meet at assembly, tolerance stack-up must be analysed formally and datums must be unambiguous on every drawing. Two suppliers each working correctly within their own stated tolerances can still produce parts that will not assemble. Resolve this at DFM with a stack-up analysis, and specify datum features that are stable cast or machined surfaces rather than parting lines.