Cavity Pressure Sensors and Smart Mold Monitoring

The injection molding machine knows what it is doing at the nozzle. It knows nothing about what is happening inside the cavity. Screw position, injection pressure and hydraulic response are proxies — useful ones, but proxies. The mold itself is the only place where the actual filling, packing and cooling state of the part can be measured.

Figure 1. Cavity pressure curve anatomy with sensor placement and data uses.

That gap is what cavity instrumentation closes, and it is also the foundation of every credible "smart molding" or Industry 4.0 program in injection molding. You cannot control or trace what you do not measure.

Sensor types and what each one tells you

• Direct cavity pressure sensors — flush-mounted into the cavity surface, sensing the melt pressure right at the wall. The most accurate and the most informative signal; also the most intrusive and the most expensive to install.
• Indirect cavity pressure sensors — mounted behind an ejector pin, so the pin transmits the force. Cheaper and easier to retrofit, but the reading depends on pin fit and friction, and it only works where there is an ejector pin.
• Melt pressure sensors — in the nozzle, the sprue, the runner or the hot runner drop. Useful for repeatability and for detecting a blocked drop, but they do not see the cavity.
• Mold temperature sensors — in the cavity surface or in the cooling circuit. Surface temperature at the end of cooling is often the best predictor of dimensional variation; circuit temperature differentials reveal cooling problems.
• Mold deflection or strain sensors, tie-bar sensors — used for clamp monitoring and for cavity balance in large tools.
• Machine-side signals — screw position, injection pressure, cushion, cycle times; the baseline that everything else is compared against.

Where to place sensors

Location determines what the sensor can tell you. Useful placements:

• End of fill — the classic location. It shows when the cavity is full, how much pressure it takes to fill, and whether the cavity is consistent from shot to shot.
• Near the gate — the best place to see viscosity variation, since pressure drop from the gate reflects the melt's state.
• Mid-cavity — reveals the balance between filling and packing and helps identify hesitation.
• One sensor per cavity in a multi-cavity tool — the only reliable way to prove cavity-to-cavity balance, and the foundation of cavity-specific part sorting.

Practical installation rules: keep sensors away from weld lines, ejector strokes and areas that see wear; protect cables with proper routing and connectors designed for mold handling; and instrument the cavity that is hardest to fill, not the easiest one.

What the data is actually used for

### Switchover control

Traditional switchover is by screw position — but screw position does not account for check-valve leakage, viscosity changes or barrel wear. Cavity-pressure switchover triggers the transition from fill to pack when the cavity actually reaches a defined pressure. This makes the fill-to-pack transition independent of the machine's hydraulic condition and gives noticeably more repeatable part weight.

### Process window and part quality

A cavity pressure curve contains the fill peak, the packing pressure plateau and the gate-seal decay. Comparing curves shot to shot compresses a lot of information into one glance: a lower fill peak means a higher-viscosity melt, a shorter pack plateau means the gate sealed early, a slow decay means the gate is passing material longer than intended.

### 100% part sorting

With per-cavity instrumentation and a monitoring system, every shot can be compared against the accepted curve. Parts from a cavity that deviated are separated automatically — short shots, flashed parts and partially packed parts never reach the customer. This is often the single strongest economic argument for instrumenting a tool: replacing sampling inspection with automatic, per-part evidence.

### Traceability and integration

• Per-part records — shot number, curve, cavity, timestamp and machine parameters stored against the serial or lot number.
• MES and OPC-UA / MQTT integration — the sensor system as a data source for the plant system, so quality data joins production and order data.
• Alarms and escalation — deviation alarms to the operator, and automatic escalation if a cavity drifts persistently.
• Predictive maintenance — trends in fill pressure and cooling behavior point to a worn check valve, a partially blocked cooling line, a degrading heater, or an approaching gate-wear problem, before they produce scrap.
• Digital twin and simulation feedback — measured cavity pressure is the best calibration data available for a moldflow model, which turns simulation from a design tool into an ongoing process tool.

Implementing without over-building

• Start with one part-critical tool. Pick a tool with a chronic quality cost — tight tolerances, cosmetic sensitivity, or a customer quality requirement. Instrument it, learn on it, and quantify the benefit.
• Define KPIs before buying hardware. Scrap rate, part weight variation, Cpk improvement, inspection cost removed, and unscheduled downtime avoided. Sensors with no KPI attached become unused cables within a year.
• Instrument the critical cavity, then expand. Full multi-cavity instrumentation can follow once the value is proven.
• Design the mold for sensors. Retrofitting is possible, but a mold designed with sensor pockets, wiring channels and a junction box is far cleaner and far more reliable.
• Calibrate and check. Sensors drift and get damaged. A periodic check against a known reference, and a record of it, is what keeps the traceability claim credible.
• Train the people who will use it. A cavity pressure curve is only useful if the process technician can read it and act on it.

Limitations to respect

• Sensors add cost, and every sensor is a component that can fail, get damaged in handling, or leak if a seal fails.
• A sensor reading is local. One sensor does not describe a whole cavity, and a multi-gate part may need several.
• Indirect sensing through an ejector pin is only as good as the pin's fit and condition.
• More data is not more quality. Instrumentation is an enabler; the process window, the tool condition and the people still decide the outcome.

Figure 2. Quick reference: the key checks and reference values from this guide.

Checklist

• Critical tool selected on the basis of a quantified quality cost
• Measurement objective defined per sensor (switchover, balance, sorting, trend)
• Sensor locations chosen deliberately: end of fill, near gate, and per-cavity for balance
• Cavity pressure switchover implemented and compared against screw-position switchover
• Fill peak, pack plateau and gate-seal behavior used as the routine monitoring signature
• Per-cavity part sorting enabled where the customer or the cost model requires it
• Data recorded per part for traceability, and integrated with the plant system where applicable
• Predictive maintenance trends defined for fill pressure, cooling differential and machine response
• Mold designed with sensor pockets, cable routing and a protected junction box
• Calibration and sensor-condition checks scheduled and documented
• Operators and technicians trained on reading and responding to the curves

Cavity instrumentation turns the mold from a passive tool into a measuring instrument. Used with a clear objective — repeatable switchover, proven cavity balance, per-part sorting, or predictive maintenance — it converts process knowledge that used to live in one experienced technician's head into data the plant can act on. Start small, define the KPI, and expand from evidence.