Die Casting yield improvement identifying gates and runners that minimize scrap generation

Content Menu

● Introduction: The Real Cost of Bad Gating

● Flow Behavior Inside the Cavity – What Actually Happens

● Gate Types That Consistently Deliver High Yield

● Runner Layout Rules That Pay Off Immediately

● Gate Thickness and Intensification Feeding

● Simulation Is No Longer Optional

● Three Production Examples Worth Copying

● Shop-Floor Checklist Before You Cut Steel

● Conclusion: Stop Accepting Mediocrity in Gating

● Q&A: Questions I Get Asked Every Week

 

Introduction: The Real Cost of Bad Gating

Yield numbers do not lie. When a 3000-ton machine spits out a 12 kg EV cross-car beam and only 72 % of the shots pass X-ray and pressure test, the scrap pile gets expensive fast. Most of that rejected metal is not random—it comes from the same repeatable defects: linear porosity bands that leak at 5 bar, shrinkage cavities under thick sections that machining opens up, cold shuts on long flow paths, and blisters that pop up after heat treatment.

After auditing more than 40 high-pressure die casting lines across North America, Europe, and China in the last eight years, I can tell you the single biggest offender is almost always the gating and runner system. The plunger does its job, the metal is clean and at temperature, the die is sprayed correctly, yet the part still leaks because the liquid front broke up the moment it left the gate. Fix the way metal enters the cavity and suddenly the same machine, same alloy, and same cycle time deliver 15–25 % more good parts.

The good news is that the knowledge to do this has been around for years. The even better news is that simulation tools finally make it practical for any shop to apply it without endless trial-and-error on the production floor.

Flow Behavior Inside the Cavity – What Actually Happens

Metal leaves the gate at 35–60 m/s depending on gate area and machine setting. That is fast enough for the stream to atomize if it is not attached to a wall. An unattached jet behaves like a fire hose in air: it breaks into droplets, oxidizes instantly, and traps microscopic air pockets. When intensification hits 1000 bar a few milliseconds later, those pockets compress into round gas pores that line up in bands—exactly the bands that leak during helium testing.

Attach the same stream to the cavity wall with a properly designed tangential gate and the picture changes completely. The metal spreads as a widening sheet, stays coherent, and pushes the air ahead of it toward vents and overflows. Wave formation is minimal, oxide generation drops dramatically, and the last areas to fill are fed with clean metal instead of contaminated spray.

Gate Types That Consistently Deliver High Yield

Tangential Gates on Real Parts

A North American Tier-1 running a 4-cavity telecom heat sink (wall thickness 1.2 mm, projected area 480 cm² per cavity) started with conventional fan gates. Leak rate was 21 %. They switched to dual tangential gates entering from the long sides. Gate velocity dropped from 58 m/s to 36 m/s, overflow mass was cut 42 %, and first-pass yield went from 79 % to 96 %. The only die change was re-machining the insert faces—no new steel, no vacuum hardware.

Tapered Fan Gates Done Right

Fan gates are cheap to machine and still have their place on thicker parts. The trick is aggressive tapering of the runner so the metal accelerates smoothly right up to the gate land. A European producer of gearbox housings (average wall 4.5 mm) had constant shrinkage porosity opposite the gate. Runner cross-section was constant, so gate velocity hit 72 m/s with massive jetting. They tapered the runner from biscuit to gate in three steps (area ratio 1.35 → 1.15 → 1.00). Velocity fell to 41 m/s, porosity disappeared, and yield rose from 81 % to 94 %.

Multiple Small Gates for Large Structural Castings

One-piece battery trays and shock towers are now common. A single large gate simply cannot fill a 1200 × 800 mm part calmly. Ten to fourteen small gates (0.7–0.9 mm thick, 8–15 mm wide each) placed around the perimeter give gate velocities around 30 m/s and radial fill patterns. A German OEM reports consistent 91–93 % yield on 18 kg trays using this approach, compared to 64–72 % when they tried a single center gate or two large side gates.

Vortex and Swirl Runners

Adding a simple cylindrical chamber or helical insert before the gate converts linear momentum into rotation. The effective forward velocity drops while fill time stays almost unchanged. Trials on a magnesium dashboard beam showed a swirl runner reduced measured air entrapment by 68 % and eliminated visible flow lines entirely.

Runner Layout Rules That Pay Off Immediately

Runner area must decrease toward the gate. Full stop. Constant-area runners create back-pressure waves that push contaminated metal backward into following cavities. A 15–20 % area reduction from biscuit to gate is a safe starting point for most aluminum parts.

Branching has to respect flow split. If you split a runner into two equal branches, each branch gets half the flow only if the areas are equal and the path lengths are similar. For asymmetrical parts, make the longer branch larger in cross-section to balance arrival time. Simulation makes this trivial now, but even a spreadsheet calculation gets you most of the benefit.

Keep the runner in the same half of the die as the gate whenever possible. Cross-ejector runners cool too fast and cause premature freezing.

Gate Thickness and Intensification Feeding

Gate thickness controls whether intensification pressure actually reaches the casting. Gates thicker than ≈1.1 mm often stay liquid until the end of intensification, so pressure bleeds back into the runner instead of feeding the part. A large North American structural caster reduced gate thickness from 1.4 mm to 0.85 mm on a node casting. External runner shrinkage vanished overnight and internal soundness improved so much that X-ray reject rate dropped from 11 % to 1.2 %.

Simulation Is No Longer Optional

Ten years ago simulation was a luxury. Today a decent seat of MAGMA, FLOW-3D CFD, or Anycasting costs less than one week of scrap on a 3500-ton cell. Run 20–30 gating variants in a weekend, pick the one that shows the calmest fill and best feeding, then cut steel. First-time yield above 90 % on new programs is now routine when the gating is designed with simulation instead of copied from the last job.

Three Production Examples Worth Copying

Example 1 – Transmission Case (AlSi9Cu3)

Original symmetrical H-runner, four cavities, center fan gates. Scrap 23 % from porosity opposite the gates. Redesigned with branched tapered runners and tangential gates on the two long sides. Gate velocity 38 m/s, balanced fill time ±4 ms across cavities. Scrap fell to 5 %, overflow mass reduced 38 %.

Example 2 – EV Inverter Housing (thin-wall, vacuum assisted)

Single large side gate created a massive wave that folded over the top wall. Blisters after T6 heat treatment were 18 %. Switched to six micro-gates with vortex elements. Fill became perfectly radial, vacuum level dropped from 80 mbar to 40 mbar (less air to pull), blister reject rate zero, yield 97 %.

Example 3 – Steering Knuckle (high integrity required)

Thick gate (1.6 mm) prevented proper intensification feeding. Shrinkage under bearing bosses. Gate thinned to 0.9 mm, runner tapered aggressively. Intensification now fully effective—mechanical properties in critical areas rose 12–18 MPa, yield from 76 % to 95 %.

Shop-Floor Checklist Before You Cut Steel

• Gate area calculated from PQ² diagram, target velocity 30–45 m/s for aluminum, 25–40 m/s for magnesium.
• Runner area decreases 12–20 % from biscuit to gate.
• Gate thickness 0.7–1.0 mm for structural aluminum, 0.5–0.8 mm for magnesium.
• Simulation shows no detached jets and air pushed only to planned vents/overflows.
• Overflow volume ≤12–15 % of casting volume (8–10 % is achievable with good gating).
• Gate land length 0.8–1.5 mm to avoid erosion but still freeze quickly.

Conclusion: Stop Accepting Mediocrity in Gating

The difference between a 75 % yield die and a 95 % yield die is rarely the alloy, the machine, or the vacuum system. Nine times out of ten it is the runner and gate design. Calm, attached flow with progressive acceleration and properly freezing thin gates eliminate the majority of porosity, shrinkage, and cold-shut defects that fill scrap bins today.

Every new program is a chance to get this right from day one. Every existing high-scrap die is a chance to reclaim 10–20 % more capacity without buying another machine. The tools, the research, and the proven layouts are all out there. The only thing missing in most shops is the decision to stop copying yesterday’s mediocre gating and start designing systems that actually work.

Q&A: Questions I Get Asked Every Week

Q1: Can I just add vacuum and keep my old bad gating?
A1: Vacuum helps, but it cannot fix a jetting stream. You still get oxide bifilms and compressed gas pores. Good gating + modest vacuum beats bad gating + perfect vacuum every time.

Q2: What is the maximum practical gate velocity before problems explode?
A2: Above 50 m/s for aluminum you are rolling the dice. Above 60 m/s you are almost guaranteed porosity bands.

Q3: Are tangential gates harder to machine?
A3: Slightly, but the extra electrode time is paid back in the first few thousand good shots.

Q4: How much yield gain is typical when we finally fix an old bad die?
A4: 12–22 % is normal. I have records of 28 % improvement on a mature transmission program just by re-cutting inserts.

Q5: Do I still need overflows with perfect tangential gating?
A5: Yes, but much smaller ones—usually 40–60 % less mass than with bad gating.