Six Times the Design Rules Contradict Each Other

Every design guide published by Die Casting Services firms ends the same way. Ten rules, cleanly numbered, each defensible. Keep walls uniform. Avoid undercuts. Break sharp edges. Keep datums in one die half. Mind your ejector pins.
Then you sit down with an actual part and discover something the guides never mention: you cannot obey all of them at once. Uniform walls tell you to keep the section at 2.5 mm. Ejector pin practice tells you the pin needs something substantial to push against. On this boss, in this corner, those two instructions point in opposite directions, and no numbered list tells you which one to break.
That decision is the entire job. What follows is how six of the most common collisions actually get resolved — not in theory, but at the bench, where somebody has to cut steel by Friday.

 
Collision 1 — Uniform wall thickness vs. ejector pin support
Where it bites. A 2.5 mm wall is exactly right for filling and for avoiding shrinkage. Then the tool designer marks eight ejector locations and three of them land on flat expanses of that same wall.
Why it happens. Ejection is a shoving match. Pins push the casting off the ejector half while it is still hot and comparatively soft. A pin bearing on an unsupported thin wall does not eject the part — it dimples it, distorts it, and eventually punches through.
The ruling: ejection wins. Add a local thickened pad or land the pin on a rib or boss. Yes, this creates a deliberate thick spot in a wall you were told to keep uniform. That is correct practice, not a compromise.
The bench test. A pad of roughly twice the nominal wall, blended in with generous radii, supports the pin without becoming a shrinkage site. What creates porosity is a large heavy section, not a small local pad.

Collision 2 — Eliminate the undercut vs. keep datums in one die half
Where it bites. There is a side port. A slide core would form it, but slides cost money, so somebody reorients the part to open the geometry along the draw. The undercut disappears — and the parting line now runs straight between the two bosses your assembly locates from.
Why it happens. Tolerance within one die half is governed by machining accuracy of the cavity, and holds around ±0.10 mm per 25 mm. The moment a dimension spans the split, it inherits die-shift variation and opens to roughly ±0.25 mm.
The ruling: if the dimension is functional, keep the slide. Tool cost is paid once, at the start. A tolerance that cannot meet the assembly is paid on every part, forever, usually as machining nobody quoted — which is how an OEM casting programme discovers a hidden per-piece cost in year two.
The bench test. Before eliminating an undercut, ask which dimensions the assembly actually depends on and check whether the reorientation moves any of them across the split. If it does not, delete the slide with enthusiasm. If it does, price the slide against a lifetime of secondary operations.

Collision 3 — Generous fillets vs. thin walls
Where it bites. Sharp internal corners are correctly forbidden: they crack the die steel and choke metal flow. So the designer applies a healthy radius everywhere. On thin-walled Die Casting Parts, that radius quietly builds a ring of extra metal at every wall junction.
Why it happens. A fillet adds material precisely where two walls already meet — the hottest, last-to-freeze region. Oversized radii on a 2 mm wall create a local heavy section that solidifies after the gate has frozen, leaving microshrinkage exactly where the part is most highly stressed.
The ruling: radius the corner, then core out behind it. Keep the fillet, and remove the mass it created by adding a scallop or metal saver on the reverse face.
The bench test. A useful starting point is a fillet radius between half and one times the nominal wall — enough to relieve the stress riser in the steel, not enough to build a hot spot.
 
Collision 4 — Minimum machining stock vs. casting variation
Where it bites. Guides rightly say to remove as little material as possible, because machining through the dense as-cast skin exposes the more porous core beneath. So somebody specifies 0.3 mm of stock — and half the batch cleans up while the other half still shows as-cast surface after the cut.
Why it happens. The stock allowance has to cover the casting's own dimensional spread. If the feature varies by ±0.25 mm and you allow 0.3 mm, some parts have almost nothing to remove and others have twice the intended depth.
The ruling: stock allowance must exceed the casting variation on that feature, not the machinist's ideal. Typically 0.5 to 1.0 mm on a cast surface. Then reduce the number of machined faces rather than the depth on each.
The bench test. The real saving is never in shaving stock. It is in asking which faces need machining at all — usually far fewer than the drawing says.
Bench note. The as-cast skin is the strongest material in the part, a dense fine-grained layer roughly 0.3 to 0.5 mm deep produced by rapid solidification against the die. Machining it away removes the best metal and exposes what lies beneath. Unnecessary machining does not just add cost; it can make cast metal weaker.
 
Collision 5 — Cooling fin density vs. thin steel conditions
Where it bites. A heat sink wants many tall closely-spaced fins. Every gap between two fins is a thin blade of die steel standing alone in the cavity.
Why it happens. Thin steel blades heat and cool faster than the surrounding block, so they heat-check early, wear quickly and occasionally break. The fin pattern that looks best in thermal simulation is often the one that destroys the tool.
The ruling: this is the one collision where thermal performance and tool life must be traded numerically, not by principle. Fewer, taller, thicker fins with more draft usually deliver comparable dissipation at a fraction of the maintenance burden.
The bench test. Ask your aluminum die casting supplier which fin gaps on the model create steel thinner than about 3 mm, and what maintenance interval those areas will drive. If the answer is a shrug, the quotation has not been engineered.
 
Collision 6 — Avoid internal features vs. the part genuinely needs one
Where it bites. Internal passages, closed channels and enclosed cavities cannot be formed in high pressure custom die casting. But the part needs a sealed coolant path.
Why it happens. Anything that would trap the die steel inside the solidified part cannot be withdrawn. No slide geometry solves a fully enclosed void.
The ruling: split the assembly, or change the process. Two castings joined with a machined face and a gasket is standard practice. Where a monolithic passage is mandatory, the part belongs in sand casting or gravity aluminum casting — both use expendable or removable cores, and most metal casting services offering only high pressure will decline the geometry outright.
The bench test. If a supplier accepts an enclosed internal passage in a die cast housing without comment, they have not read the model. That is worth knowing before tooling.
 
The Order Behind All Six Rulings
Every ruling above follows the same hierarchy, whether or not anyone says so out loud.

Rank    Question    Why it outranks the next
1    Will it eject?    A part that will not release from the die has no other properties worth discussing
2    Will it fill?    Flow length and wall thickness determine whether the geometry exists at all
3    Will it function?    Protect the datums the assembly depends on, not every dimension on the drawing
4    Will the die survive?    Thin steel and sharp corners set the maintenance bill for the programme's life
5    What does it cost?    Optimise only after the four above are satisfied
Read it downward. The higher rule wins. Most bad castings come from solving tier five before tier one — deleting a slide to save tooling cost, then discovering the tolerance no longer meets the assembly.
 
Questions From the Floor
Our guide says keep walls uniform, but our supplier added bosses under the ejector pins. Who is right?
Both. Uniform wall thickness is shorthand for avoiding heavy sections that cannot be fed after the gate freezes. A small local pad supporting an ejector pin is not that. Without it the pin distorts or punches through a hot, soft casting during ejection. Local thickening at pin and slide positions is correct practice on aluminium casting parts, not a violation of the rule.
How much does one slide core really cost compared with machining afterwards?
A slide typically adds a four-figure sum to the tool and a small increment to cycle time. Secondary machining adds a fixture, a setup and an inspection step to every single part, forever. The crossover is usually a few thousand pieces. Below that, machining may genuinely be cheaper; above it, the slide almost always wins — and a slide holds the feature relative to the same die half, which machining after the fact cannot recover.
Why does a fillet need coring out behind it?
Because a fillet adds metal at the junction of two walls, which is already the last region to solidify. On a thin-walled part an oversized radius creates a hot spot that freezes after its feed path has closed, producing microshrinkage in a highly stressed corner. Keeping the radius and removing the mass behind it gives you the stress relief without the heavy section.
Our machinist wants 0.3 mm of stock. Our caster wants 1.0 mm. Who decides?
The casting's measured variation on that specific feature decides. If the feature holds ±0.25 mm as cast, 0.3 mm of stock guarantees that some parts clean up and others do not. Set the allowance above the observed spread, then reduce cost by machining fewer faces rather than by cutting the depth on each one.
Can I specify a closed internal channel in a die casting?
Not in high pressure die casting. Any feature that would trap die steel inside the solidified part cannot be withdrawn, and no slide arrangement changes that. The normal solutions are to split the part into two castings joined at a machined face, or to move it to sand or gravity casting where expendable cores can be used — or in some geometries to aluminum fabrication from extrusion and plate. A metal casting service that quotes an enclosed passage without raising this has not opened the model.
How do I know if my fin design will wreck the tool?
Look at the gaps, not the fins. Each gap becomes a blade of die steel, and blades much below about 3 mm heat-check early and wear fast. Ask the supplier to mark every thin steel condition on your model and state the maintenance interval each one drives. Fewer, thicker, taller fins with more draft usually give similar thermal performance with a fraction of the tool wear.
Is this hierarchy different for zinc?
The order is the same, but the pressure on each tier changes. Zinc casting in a hot chamber machine runs cooler and is far gentler on tool steel, so tier four relaxes considerably and thinner walls become achievable. Ejection and filling still outrank everything else.
 
How We Apply This
We publish these numbers with the conditions attached, because a figure without its qualifier is where these conflicts start.
As-cast tolerance:
●±0.10 mm per 25 mm within one die half · ±0.25 mm across a parting line, datum to be stated · ±0.02 mm only where a machining operation is quoted separately
Minimum wall:
●1.2 mm aluminium within 80 mm flow length of a gate · 0.8 mm zinc within 60 mm
Practical wall range:
●2.0 – 5.0 mm; sections above about 8 mm should be cored and ribbed
Draft:
●1° – 2° typical, more on deep cores and textured faces
Tool life:
●100,000 – 150,000 shots in aluminium with polishing every 20,000 – 30,000 shots
Alloys:
●A380 · A383 / ADC12 · A360 · A413 · ZAMAK 3/5. A360 or A413 where bright anodise is required, because A380 at 3 – 4% copper will not anodise to a clean bright finish
Certification:
●ISO 9001 · ISO 14001 · CE · RoHS · SGS, issued in our own legal entity name
In-house: tooling design and manufacture, cold chamber aluminum die casting, hot chamber zinc, 3/4/5-axis CNC machining, blasting, anodising, powder coating, plating, insert installation, sub-assembly and leak testing — one aluminum alloy processing manufacturing chain with no undeclared handovers. Our metal fabrication and custom metal fabrication lines — laser cutting, press brake forming, TIG and MIG welding — often let a die cast metal body carry fabricated brackets, which removes side actions from the die entirely; for low volumes structural fabrication needs no tooling at all.
Typical aluminium casting products: LED heat sinks, telecom enclosures, motor housings, gearbox covers, valve bodies and mounting brackets. We hold no stock. Every custom casting and custom fabrication job is built to drawing or sample, with permanent tooling ownership and 30+ years of factory processing experience behind it.
Send a STEP file with your functional dimensions tagged, and we will mark every rule collision we can see in it — including the ones that argue against our own tooling revenue. When comparing aluminum die casting suppliers and their casting services, that list is worth more than a price.