{"id":8137,"date":"2026-10-26T02:29:11","date_gmt":"2026-10-26T02:29:11","guid":{"rendered":"https:\/\/chinametalfoundry.com\/?p=8137"},"modified":"2026-10-27T02:22:47","modified_gmt":"2026-10-27T02:22:47","slug":"casting-part-consolidation","status":"publish","type":"post","link":"https:\/\/chinametalfoundry.com\/pt\/blog\/casting-part-consolidation\/","title":{"rendered":"Casting Part Consolidation: Turning an 8-Part Weldment Into One Casting"},"content":{"rendered":"
\n
\n

Casting part consolidation is sold as a design idea. On the shop floor it is a cost decision: you delete eight plate parts, ten bolts and five metres of weld, and you take on one pattern, one core box and one casting drawing. Whether that trade pays depends on a single number \u2014 how many units you build before the programme ends.<\/p>\n<\/div>\n

This guide gives you the arithmetic and the limits of casting part consolidation. Every figure below is a worked example built from real shop rates and standard solidification rules, not a rule of thumb. Where consolidation loses, we say so and show the number that kills it. For background on the process families referenced here, see the American Foundry Society<\/a>, the North American Die Casting Association<\/a> and the American Welding Society<\/a>, and start with our own metal casting process overview<\/a> if the process names are new to you.<\/p>\n

\"Casting
One casting on the left, the eight plates and fasteners it replaces on the right.<\/figcaption><\/figure>\n

What Casting Part Consolidation Actually Removes<\/h2>\n

A welded or bolted assembly carries costs that have nothing to do with the shape of the part. Cutting, bending, fit-up, welding, grinding, stress relieving, fastener purchase, torque checks and final assembly all disappear at once when the assembly becomes one piece. At the same time, casting part consolidation takes on costs a weldment never had: tooling, a longer qualification loop, and a set of solidification risks that only exist once you put all the geometry into one body.<\/p>\n\n\n\n\n\n\n\n\n\n
Disappears<\/th>\nAppears<\/th>\n<\/tr>\n<\/thead>\n
Cutting, bending and fit-up labour<\/td>\nPattern and core box tooling<\/td>\n<\/tr>\n
Weld metal, grinding and stress relief<\/td>\nDesign, FEA and drawing release work<\/td>\n<\/tr>\n
Fasteners, seals and torque records<\/td>\nSolidification risk at thick junctions<\/td>\n<\/tr>\n
Multiple part numbers and their inventory<\/td>\nLonger first-article cycle before SOP<\/td>\n<\/tr>\n
Leak paths and alignment stack-up<\/td>\nMinimum run quantity on every production batch<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Two of these lines decide most projects and they pull in opposite directions. Labour deletion saves money on every unit, forever. Tooling and engineering cost money once, before a single good part ships. That is why volume is the whole argument \u2014 and why our casting tooling cost breakdown<\/a> should be read next to this one, alongside the wider picture in casting total cost of ownership<\/a>.<\/p>\n

Casting Part Consolidation Math: The Breakeven Number<\/h2>\n

The test for casting part consolidation is one line:<\/p>\n

Breakeven units = (tooling + engineering + qualification + expected tooling changes) \u00f7 (weldment unit cost \u2212 casting unit cost)<\/strong><\/p>\n

If breakeven units is under roughly a third of your annual volume, consolidation pays back inside four months and the decision is easy. If it is more than about three times your annual volume, the payback stretches past three years and you are buying risk, not savings. Here is the full model for a real part.<\/p>\n

Case A \u2014 900 per year steel weldment<\/h3>\n
\n

A machinery bracket built from eight laser-cut S355 plates, four bends, 5.6 m of 6 mm fillet weld, ten M12 fasteners, then machined on two setups. Shop rates used: welding and fitting at $72\/h, machining at $95\/h, inspection at $72\/h. Deposition is taken at 2.0 kg\/h with a 30% arc-on factor, which is why 0.79 kg of filler metal still costs 1.32 h.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Cost line<\/th>\nWeldment<\/th>\nOne casting<\/th>\n<\/tr>\n<\/thead>\n
Material purchased<\/td>\n27 kg plate \u00d7 $1.35 = $36.45<\/td>\n14 kg casting \u00d7 $2.95 = $41.30<\/td>\n<\/tr>\n
Laser cut + nesting<\/td>\n18.4 m at 1.35 m\/min + 0.30 h = $50.08<\/td>\n\u2014<\/td>\n<\/tr>\n
Press brake, 4 bends<\/td>\n$38.00<\/td>\n\u2014<\/td>\n<\/tr>\n
Fit-up and tack<\/td>\n0.60 h = $43.20<\/td>\n\u2014<\/td>\n<\/tr>\n
Welding<\/td>\n0.79 kg filler, 1.32 h = $94.75<\/td>\n\u2014<\/td>\n<\/tr>\n
Grind and blend<\/td>\n0.70 h = $50.40<\/td>\n\u2014<\/td>\n<\/tr>\n
Stress relief<\/td>\n16 kg \u00d7 $0.35 = $5.60<\/td>\n\u2014<\/td>\n<\/tr>\n
Machining<\/td>\n0.90 h = $85.50<\/td>\n0.35 h = $33.25<\/td>\n<\/tr>\n
Inspection<\/td>\n0.30 h = $21.60<\/td>\n0.15 h = $10.80<\/td>\n<\/tr>\n
Fasteners<\/td>\n10 \u00d7 $1.85 = $18.50<\/td>\n\u2014<\/td>\n<\/tr>\n
Final assembly<\/td>\n0.40 h = $28.80<\/td>\n\u2014<\/td>\n<\/tr>\n
Scrap and rework<\/td>\n+4%<\/td>\n+2%<\/td>\n<\/tr>\n
Delivered unit cost<\/strong><\/td>\n$491.80<\/strong><\/td>\n$87.06<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Saving per unit: $404.74<\/strong>. Non-recurring cost: $6,400 pattern and core box, $5,800 design and FEA, $1,200 sample and first article inspection<\/a>, plus an expected tooling change of 0.12 \u00d7 $4,800 = $576 \u2014 $13,976 total<\/strong>. Breakeven is 34.5 units. At 900 per year that is under two weeks of production, and the first-year net is $364,266 \u2212 $13,976 \u2212 $609 of early-run scrap = $349,681<\/strong>. Purchased weight drops from 27 kg of plate to a 14 kg ductile iron casting<\/a>.<\/p>\n

Two things are worth noting before you copy these numbers. First, the weldment side is dominated by labour, not material \u2014 $412.33 of the $472.88 subtotal is labour and machine hours, which is why this comparison swings so hard towards casting. Second, the comparison is unfair to the weldment if your current assembly is already a cheap stamping design with spot welds instead of fillet welds; spot welding costs a fraction of arc welding per joint and closes most of the gap. If that is your baseline, read casting versus welded fabrication<\/a> before deciding.<\/p>\n

\"Casting
Ribs, bosses and cored passages replace what used to be separate plates and welds.<\/figcaption><\/figure>\n

Where Casting Part Consolidation Stops Working<\/h2>\n

Case A is the easy one. The two failure modes that break casting part consolidation are volume that never recovers the tooling, and geometry that cannot be cast at all. Both are detectable before you spend a dollar.<\/p>\n

Case B \u2014 40 per year frame: the tooling never pays back<\/h3>\n
\n

A 1.6 m frame currently welded from four plates: the weldment lands at $382.10 delivered. The consolidated casting weighs 26 kg and needs four cores, so it costs $76.70 + $24.80 of cores + $42.75 machining + $8.64 inspection, then a 5% allowance for casting scrap and core-shift rework = $160.53<\/strong>. Saving per unit is $221.57, which looks excellent until you put the tooling next to the volume: $18,400 tooling + $6,200 engineering + $1,400 qualification + $700 risk = $26,700, or 121 units to breakeven<\/strong>. At 40 per year that is three years, and the programme ends in three years. The honest answer is no \u2014 unless the same casting is shared with a sibling product at 55 per year, which pulls payback to 1.3 years and flips the decision. Sharing one consolidated casting across two products is the single most effective lever when volume is the blocker.<\/p>\n

Case C \u2014 400 per year aluminium: small parts, real tooling, still worth it<\/h3>\n
\n

Two aluminium parts machined from plate plus eight fasteners: material $16.32, chip credit \u2212$5.83, machining 0.84 h = $79.80, fasteners $8.80, assembly $25.20, inspection $14.40 \u2192 $142.85<\/strong> with 3% scrap. One fundi\u00e7\u00e3o \u00e0 cera perdida<\/a> at 0.62 kg \u00d7 $12.80\/kg plus 0.12 h of machining and a short inspection cycle lands at $24.13<\/strong>. Saving is $118.72 per unit; wax tool $4,900 plus fixture $3,000 plus $2,400 engineering = $10,300, so breakeven is 87 units. At 400 per year the first-year net is $37,187<\/strong>. Push the same part to 6,000 per year and a fundi\u00e7\u00e3o sob alta press\u00e3o<\/a> tool at $48,000 pays back in under 200 units instead \u2014 the process choice follows the volume, not the other way round.<\/p>\n\n\n\n\n\n\n\n
Case<\/th>\nSaving \/ unit<\/th>\nNRE<\/th>\nBreakeven<\/th>\nVolume<\/th>\nVerdict<\/th>\n<\/tr>\n<\/thead>\n
A \u2014 steel weldment<\/td>\n$404.74<\/td>\n$13,976<\/td>\n35<\/td>\n900\/yr<\/td>\nGo<\/td>\n<\/tr>\n
B \u2014 large frame<\/td>\n$221.57<\/td>\n$26,700<\/td>\n121<\/td>\n40\/yr<\/td>\nNo, unless shared<\/td>\n<\/tr>\n
C \u2014 aluminium pair<\/td>\n$118.72<\/td>\n$10,300<\/td>\n87<\/td>\n400\/yr<\/td>\nGo<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Prototype Before You Cut Production Tooling<\/h2>\n

Every casting part consolidation project carries one probability nobody writes down: the chance that the first tooling needs a correction. On our own work that sits around 12%, and a correction costs about $4,800 and four to six weeks. A printed sand mould set costs roughly $1,250\u2013$2,400 and turns around in two to five days instead of four to six weeks for a pattern.<\/p>\n

Run the expected value honestly. If a printed mould programme cuts the correction probability from 12% to 4%, it saves 0.08 \u00d7 $4,800 = $384 of expected tooling<\/strong>. Against a $1,850 prototype that is a loss \u2014 on tooling risk alone the prototype does not pay. It pays on schedule, and schedule is the reason most casting part consolidation projects get a prototype at all. If the programme has a fixed SOP date and each week of slip is worth $9,000 in expediting, air freight or lost line time, then removing eight points of risk across a five-week correction is worth 0.08 \u00d7 5 \u00d7 $9,000 = $3,600 of expected schedule exposure<\/strong>. That is the number that justifies the prototype, and it is a schedule argument, not a tooling argument.<\/p>\n

There is a second benefit that is easy to miss: printed moulds and prototype castings can be poured in the production alloy, so the mechanical and dimensional data you collect is usable in the FAI package<\/a> instead of being thrown away. See Prot\u00f3tipos impressos em 3D<\/a>, metal prototype casting<\/a> e mould and pattern making<\/a> for the routes we run before cutting production tooling, and weigh them against 3D printing versus casting<\/a> for volume parts.<\/p>\n

Casting Rules That Break a Consolidated Design<\/h2>\n

Casting part consolidation concentrates geometry, and concentrated geometry creates thick junctions, long cores and walls that are too thin to fill. Four rules catch almost every problem before tooling is cut.<\/p>\n

Minimum wall thickness by process<\/h3>\n
\n

The wall you can cast is set by the process and the alloy, not by the load case. Use this table as the first filter on any consolidated design.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n
Process and alloy<\/th>\nPractical minimum wall<\/th>\nTypical draft<\/th>\n<\/tr>\n<\/thead>\n
Green sand, grey or ductile iron<\/td>\n4\u20136 mm<\/td>\n1\u20132\u00b0<\/td>\n<\/tr>\n
Resin sand, carbon and low-alloy steel<\/td>\n6\u20138 mm<\/td>\n1.5\u20133\u00b0<\/td>\n<\/tr>\n
Resin sand, aluminium<\/td>\n4\u20135 mm<\/td>\n1\u20132\u00b0<\/td>\n<\/tr>\n
Shell mould, iron<\/td>\n3\u20134 mm<\/td>\n0.5\u20131\u00b0<\/td>\n<\/tr>\n
Permanent mould, aluminium<\/td>\n3.5\u20135 mm<\/td>\n2\u20133\u00b0<\/td>\n<\/tr>\n
High pressure die casting, aluminium<\/td>\n2.0\u20133.0 mm<\/td>\n1\u20132\u00b0<\/td>\n<\/tr>\n
High pressure die casting, zinc<\/td>\n1.0\u20131.5 mm<\/td>\n0.5\u20131\u00b0<\/td>\n<\/tr>\n
Investment casting, stainless steel<\/td>\n2.0\u20133.0 mm<\/td>\nNone needed<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Bosses, ribs and the modulus check<\/h3>\n
\n

A solid boss is the classic shrinkage trap in a consolidated part. Take a \u00d830 \u00d7 20 mm boss sitting on a 10 mm wall. Volume is 14,137 mm\u00b3 and the cooling surface is the side plus the top only, because the base merges into the wall \u2014 2,592 mm\u00b2. Chvorinov modulus is 5.45 mm<\/strong> against 5.00 mm<\/strong> for the wall itself (half the wall thickness). Solidification time scales with M\u00b2, so the boss freezes 1.19 times later than the wall it sits on and will draw metal back out of it. On a machined face that shows as a 0.5\u20131.5 mm sink after facing. Core a \u00d814 mm hole through the same boss and the modulus drops to 3.33 mm \u2014 now the boss freezes in 0.44 of the wall time and the wall feeds it instead. That is why the rule is boss OD at least 2\u00d7 hole ID, and why our casting riser design<\/a> guide treats feeding as the first check, not the last.<\/p>\n

Rib thickness should sit at 0.6\u20130.8 of the wall, rib height below 5\u00d7 rib thickness, and junction fillets at 0.5\u20131.0 of wall thickness. A consolidated design that keeps every wall within a 2:1 thickness range will almost always feed cleanly; one that jumps from 6 mm to 25 mm will not, no matter how good the solidification simulation<\/a> looks. Shrinkage in these junctions is the single most common defect we see on first shots \u2014 see defeitos comuns de fundi\u00e7\u00e3o<\/a> for how it presents on the fracture face.<\/p>\n

Cores have a length limit<\/h3>\n
\n

Internal passages are what make casting part consolidation attractive, and they are also what kills it. A sand core supported at one end only behaves like a cantilever in liquid metal.<\/p>\n\n\n\n\n\n\n\n\n\n
Core length \u00f7 diameter<\/th>\nWhat it needs<\/th>\n<\/tr>\n<\/thead>\n
Up to 3<\/td>\nSingle core print is enough<\/td>\n<\/tr>\n
3\u20135<\/td>\nLonger print, or a second print opposite<\/td>\n<\/tr>\n
5\u20138<\/td>\nPrints at both ends<\/td>\n<\/tr>\n
8\u201312<\/td>\nCore supports or chaplets, with shift risk<\/td>\n<\/tr>\n
Above 12<\/td>\nNot castable in one piece \u2014 split it or drill it<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

A 280 mm oil gallery on a 20 mm core is 14:1. It will not hold itself. Either the part becomes two castings with a bolted joint, or the gallery is drilled afterwards. Both are valid; neither is the saving you budgeted. Leave openings of at least 25 mm so core sand can actually be shaken out, and remember that every cored hole adds a machining allowance<\/a> decision on the faces that get finished.<\/p>\n

One part now means one alloy<\/h3>\n
\n

An assembly lets each piece have its own material: a wear plate here, a machinable boss there, a cheap cover somewhere else. One casting has one chemistry and one heat treatment, and casting part consolidation quietly deletes that freedom. If a bracket needs a 55 HRC wearing face and 12% elongation in the arm, a single ductile iron grade cannot give you both. The honest options are a bolt-on wear plate, a localised heat treatment<\/a> or a surface engineering route, or accepting a two-piece design. Check alloy capability early against sele\u00e7\u00e3o da liga de fundi\u00e7\u00e3o<\/a>, grey versus ductile iron<\/a>, cast steel<\/a> e a\u00e7o inoxid\u00e1vel<\/a> before the geometry is frozen.<\/p>\n

A 20-Minute Screening Test Before You Redesign<\/h2>\n

Run these eight checks on the current assembly drawing before anyone opens CAD. Two or more failures means casting part consolidation is probably the wrong project for this part, and no amount of quoting will change that.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n
Check<\/th>\nPass condition<\/th>\n<\/tr>\n<\/thead>\n
Volume<\/td>\nAnnual units at least 3\u00d7 breakeven units<\/td>\n<\/tr>\n
Envelope<\/td>\nFits the flask or machine with 80 mm of sand cover<\/td>\n<\/tr>\n
Internal passages<\/td>\nEvery core L\/D at or below 12<\/td>\n<\/tr>\n
Single alloy<\/td>\nOne grade covers every function on the part<\/td>\n<\/tr>\n
Wall range<\/td>\nThickest wall no more than 2\u00d7 thinnest<\/td>\n<\/tr>\n
Machining<\/td>\nAll datums reachable in two setups or fewer<\/td>\n<\/tr>\n
Shake-out<\/td>\nOpenings of at least 25 mm for core removal<\/td>\n<\/tr>\n
Programme life<\/td>\nThree or more years of production after SOP<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

In practice, casting part consolidation pays best where the assembly is bolted and sealed: pump and valve bodies<\/a> that used to be three pieces and two gaskets, automotive brackets<\/a> built from five stampings, and agricultural frames<\/a> where every bolt is a field-service failure point. It pays worst on sealed-for-life designs with a two-year model cycle.<\/p>\n

\"Casting
Case C: two machined parts and eight fasteners become one casting.<\/figcaption><\/figure>\n

How to Write the RFQ for a Consolidated Casting<\/h2>\n

Casting part consolidation produces a new part, not a revised one. Quote requests that treat it as a small change get vague prices back. Six things make the difference:<\/p>\n