The Simulation Is Only as Good as What You Put In
"We run MAGMAsoft simulation on every tool before we cut steel."
That sentence appears on many die casting services pages covering Aluminum Die Casting, Zinc Casting, and other casting services, and it is genuinely useful — when the simulation is set up correctly. The problem is that a simulation can produce a plausible-looking fill animation from almost any geometry and still tell you nothing reliable about what T1 will produce.
Fill simulation software — whether MAGMAsoft, FLOW-3D, ProCAST, or any equivalent — works by solving fluid dynamics and heat transfer equations in a numerical model of the die and the melt. The output is only as accurate as the inputs: material properties, boundary conditions, and the criteria by which results are judged. Three of those four input categories are provided by the supplier. The fourth — the definition of what a good result looks like for your part, in your application — can only come from the buyer.
When that fourth input is missing, the simulation runs, the report looks professional, and T1 may still fail. Not because the software was wrong, but because nobody told it what to be right about.
Input 1: Critical-to-function features must be flagged before the simulation runs
A fill simulation optimises for complete cavity fill — that is what it is designed to do. It does not have a list of which features matter most. In a casting with a thin structural rib 200 mm from the gate and a thick flange 30 mm from the gate, the simulation may predict a complete fill, because the flange filled and the runner and vent system looked balanced on paper.
The practical mechanism: most custom die casting suppliers run simulation with a gating arrangement that fills the part efficiently. Efficient general fill and reliable fill of a specific critical feature are often different gating strategies. Only the buyer knows which features are critical. Providing that list — ideally as a GD&T annotation on the drawing, or as a written list attached to the RFQ — is not engineering work for the buyer. It is a ten-minute exercise that directly determines whether the simulation produces a useful result.
Input 2: Die temperature must be specified, not assumed
Every simulation software has a default die temperature. For aluminium, that default is typically around 200 °C — a reasonable steady-state temperature for a well-warmed die running at production cycle time with balanced cooling.
It is not the temperature at T1. At T1, the die is cold at the start of sampling. If the supplier preheats with a torch or a short warm-up sequence, the die face may be at 160 °C for the first shots. If the cooling channels run too fast, the ejector side may be at 140 °C. After 50 shots without adjusted cooling, a hot spot can develop on the fixed side above 240 °C.
Input 3: Porosity acceptance zones must be defined before the simulation report is interpreted
Fill simulation predicts where porosity is likely to concentrate. It cannot decide whether that location is acceptable. A boss base with medium porosity risk may be structurally irrelevant for a bracket — the boss carries a low-torque fastener in a static assembly. The same porosity at a bore wall that carries fluid at 25 bar is a leak waiting to happen.
Input 4: Programme volume and target cycle time determine whether the simulation is optimised for the right machine
The same die can run at 30-second or 50-second cycle times depending on cooling channel efficiency, intensification pressure, and the target shot velocity profile. Each of those variables changes the thermal state of the die and the fill dynamics.
Providing annual volume, programme life, and target machine class before simulation is configured lets the supplier set the shot profile, intensification pressure, and cycle time that correspond to actual production. This matters equally for high-volume OEM casting programmes and for lower-volume custom die casting and custom casting programmes where the machine selection is still open.
The Simulation as a Contract
When a custom die casting supplier provides simulation output as part of DFM sign-off, that output is an engineering prediction. Like any prediction, it can be held accountable — but only if the conditions it was predicting are documented.
The four inputs above are not a burden on the buyer. They are the same information needed to write an RFQ, specify inspection, and design the production process. Providing them before simulation runs — not after T1 fails — is the only way to get a simulation result that means something.