Mold structure selection drives everything downstream, so start there before touching a single process parameter. In practice, four categories cover most work: two-plate molds for straightforward single-face parts, three-plate molds when the gate must be centrally located or automatically degated, hot runner systems for engineering resins where sprue waste and cycle time matter, and stack molds when output per cycle needs to double within a fixed tonnage. Each choice locks in gate type, ejection layout, and cooling circuit routing — decisions that are far cheaper to get right at the design review than on the bench. A two-plate tool with a side gate behaves nothing like a hot runner tool with a valve gate, and the parameter window shifts accordingly.
Once the tool is mounted, tune parameters in a fixed sequence rather than chasing symptoms randomly. The ten-step order that holds up on the floor: barrel temperature, mold temperature, injection speed, injection pressure, hold pressure, hold time, cooling time, screw speed and back pressure, shot cushion, and finally clamp force. Set temperatures first and let them stabilize — a 10°C shift in melt temperature can move cavity pressure enough to flash a part. Establish fill speed before pressure, since pressure only trims what speed has already delivered. Hold time should be set from gate freeze analysis, not guesswork; a short cushion of 3–5 mm protects against shot-size drift. Clamp force comes last, tuned to just above the point where flash disappears.
Document every stable process as a setup sheet with the actual values, not ranges — the next shift needs a baseline it can return to after a mold pull. Watch cavity pressure if you have sensors; it catches material lot variation before parts do. For mold sourcing and structure comparisons across these four categories, visit MoldSo at www.moldso.com.