Die spray is a process variable, not a cleanup step
Walk into a die casting cell and you can usually tell how the shift is going by looking at the spray head. If the nozzles are crusted white and the operator hits the same button twice "to be safe," the die is running blind. Porosity, soldering and cold shuts follow within a week.
The release agent does three jobs at once: it forms a sub-micron release film, it removes heat from the die face through evaporation, and it lubricates the gate and runner where metal velocity is highest. Treating it as a wiping step throws away the only cheap control you have over die thermal balance.
Alloy-specific die face temperature
Every alloy has a working window, and outside it the defect mode changes rather than the defect rate. These are the ranges most shops hold on the die face, measured right after ejection:
- Aluminum (ADC12, A380): 180–250 °C
- Zinc (Zamak 3, Zamak 5): 150–200 °C
- Magnesium (AZ91D): 200–260 °C
Below the window you get cold shuts, flow lines and short fill on 1 mm ribs. Above it you get shrinkage porosity in thick bosses, soldering at the gate, longer cycles and faster heat checking on the die steel. Both failure modes look like "bad casting" on the inspection bench, but they need opposite corrections.
Measure with a contact probe at six to eight fixed points, not with an infrared gun alone. Polished aluminum has an emissivity around 0.1–0.3, so an IR reading on a shiny die face can be 50 °C low. If you must use IR, tape a known-emissivity patch at each point and shoot the patch.
Where the hot spots actually sit
Average die temperature is a comfortable number that hides the problem. Gate, runner, overflow and any boss above 6 mm wall thickness typically run 40–60 °C hotter than the surrounding cavity. That local delta is what drives soldering and shrinkage porosity, and it is also where spray coverage is usually thinnest because the robot path was taught around the part shape.
Three fixes, in order of cost: extend spray dwell at the hot spot by 0.5–1 s, add a bubbler or baffle in the core, and only then consider a copper or tungsten-copper insert. The insert works, but it also changes heat flow to the rest of the cavity, so re-baseline the spray settings afterwards.
Spray settings that repeat shot to shot
Most of the variation between good and bad shifts comes down to four numbers that nobody writes down.
- Dilution. Follow the supplier ratio, typically 1:80 to 1:200 with DI water. Hard water leaves mineral scale that insulates the die and clogs nozzles. Mix in a graduated container, not by eye.
- Distance and angle. 150–250 mm, perpendicular to the surface. At 400 mm the pattern is wide and thin; the gate gets nothing.
- Sequence. Spray the gate and overflow first while they are hottest, then the general cavity. Reverse the order and the hot spots flash off the water before it can do any cooling.
- Blow-off. 1–2 s of dry air before the shot. This is not optional. Water trapped in a deep pocket turns to steam and shows up as gas porosity or blisters, and A380 already carries roughly 0.5–1.5 % dissolved gas from the shot itself.
Check nozzle patterns weekly against a card. One partially clogged nozzle is a localized hot spot that no amount of extra spray elsewhere will fix.
Thermal control units and what the gauges hide
Aluminum dies are usually run on oil at 180–250 °C; pressurized water units reach 140–160 °C and transfer heat faster per unit of flow. Whatever the medium, the useful number is the temperature drop across each circuit. A rise of 2–5 °C between inlet and outlet is normal. Ten degrees means restricted flow, and restricted flow means the controller is holding setpoint while the cavity runs cold in one zone.
Hoses delaminate from the inside, so a bubble test once a quarter is worth more than a flow meter. Scale buildup in channels is slower but more expensive: after 12–24 months, fouled channels can cut cooling capacity by 20–30 %, which shows up first as a cycle time that creeps and then as warpage that was never there before.
Mapping defects back to the spray
The point of all this instrumentation is a short diagnostic path. When a defect appears, the spray and thermal log should tell you which way to move:
- Soldering at the gate: local overheat, starved spray at the gate, or gate velocity above roughly 60 m/s for aluminum, which erodes the die and welds aluminum to steel.
- Cold shut or flow line: die face too cold, fill too slow, or spray so wet that the cavity is chilled before metal arrives.
- Shrinkage porosity in a boss: that zone is too hot and the intensification pressure is not reaching it before the gate freezes.
- Blisters after ejection: trapped water, insufficient blow-off, or a die face cold enough to freeze a skin before gas can escape.
Note that only two of those four are spray problems. That is the discipline worth building: the log tells you whether to change the spray, the thermal unit, or the shot profile — rather than changing all three and learning nothing.
What to ask a casting supplier for
For any die casting program that matters, ask for a die temperature log at fixed points, the release agent dilution and spray time per shot, blow-off duration, thermal controller setpoints, and thirty days of X-ray reject rate sorted by defect type. A supplier who cannot produce a die temperature log is guessing, and guessing is expensive when the tool is already cut.
It also helps to agree on the machining allowance up front. Castings that will be CNC machined afterwards need 0.3–0.7 mm of stock on functional faces, and those faces should sit in the same die half where possible. If the spray and thermal balance are unstable, the as-cast surface moves, and the machine shop ends up chasing a datum that shifts from lot to lot.
The prototype die is a different animal
A single-cavity prototype tool in H13 is often sprayed by hand, which means cycle-to-cycle variation is larger than on a production cell. Two habits pay for themselves: spray by count and timer rather than by feel, and log die temperature every 20 shots. Expect the first 20–50 shots to run cold while the die reaches thermal equilibrium — plan for scrap in that window instead of re-cutting the tool because the first ten parts showed flow lines.
Get the spray and thermal balance documented during the prototype run, and the production tool will inherit a process instead of a rumor.
