Piezoelectric Cavity Transducer Drift in High Temperature Injection Moulding

Crystal

Direct melt contact in high-performance tooling subjects cavity pressure sensors to localized thermal shocks above 300 degrees Celsius against sustained mould base temperatures past 180 degrees Celsius. When processing resins such as polyether ether ketone or polyphenylene sulfide, these conditions test the physical limits of the electromechanical transduction element. Standard Alpha-quartz (silicon dioxide) elements have a transverse piezoelectric coefficient (d11) of roughly -2.31 pico-Coulombs per Newton at room temperature.

As operating temperatures exceed 200 degrees Celsius, quartz suffers progressive thermal lattice degradation long before reaching its alpha-to-beta phase transition at 573 degrees Celsius. Well prior to that structural inversion, internal electrical resistivity drops off exponentially, compromising charge retention across the pressure holding phase.

Alternative single-crystal materials avoid this lattice instability at high temperatures. Gallium phosphate (GaPO4) undergoes no phase transition below 970 degrees Celsius and maintains a piezoelectric sensitivity of approximately 4.5 pico-Coulombs per Newton ~ roughly double that of natural quartz. That higher sensitivity delivers more charge output per unit of mechanical stress, which improves the signal-to-noise ratio when tracking low packing pressures in thin-wall parts.

Gallium phosphate crystals maintain stable piezoelectric coefficients up to 400 degrees Celsius while standard alpha-quartz exhibits signal attenuation beyond 200 degrees Celsius.

The electromechanical response under dynamic mechanical load follows a linear relationship within the elastic limit of the crystal assembly:

Q = d11 F

The generated electrical charge Q, measured in pico-Coulombs, scales directly with the applied normal force F on the sensor diaphragm. At high temperatures, drift is driven primarily by internal bulk resistivity loss. Once bulk resistance drops below 10 to the 12th power Ohms, charge drains across the crystal body instead of traveling through the high-impedance cable to the charge amplifier.

On the acquisition monitor, this appears as an artificial pressure decay during extended injection pack and hold phases.

Thermal Lattice Stability across Piezoelectric Materials

Selecting crystal chemistry comes down to balancing operating thermal limits against mechanical shear thresholds and transducer cost. Housings machined from precipitation-hardened stainless steels must also match the thermal expansion characteristics of the crystal stack so internal preload does not relax at process setpoints.

Ferroelectric ceramics like lead zirconate titanate provide high raw charge sensitivity, but they cannot handle direct-contact cavity monitoring at elevated temperatures. Their Curie points sit below 350 degrees Celsius, which causes complete depolarization under sustained tool heat. They are also intensely pyroelectric, producing false charge outputs from raw temperature fluctuations as the melt front passes.

Because of non-polar crystal symmetry, alpha-quartz and gallium phosphate generate no intrinsic pyroelectric charge along their primary sensing axes. Apparent thermal drift in these sensors stems instead from differential expansion between the quartz disc, preload ring, and outer sleeve.

Steep thermal gradients across the transducer face impose uneven mechanical stress through the crystal stack. When molten polymer at 380 degrees Celsius strikes a cooler sensor diaphragm, the initial contact creates a sharp compressive stress spike before heat distributes across the transducer housing. That localized thermal expansion shows up on the acquisition system as an instantaneous, non-mechanical pressure offset.

Tooling engineers must determine whether the sensor housing requires a floating diaphragm or a rigid front face. Rigid faces channel axial casing expansion straight into the crystal, multiplying measurement errors whenever mould base temperatures swing during a shift.

Leakage

Charge generated by dynamic cavity pressure must reach the operational amplifier’s input stage without transmission losses. Because the piezoelectric crystal acts as a charge source with very high internal resistance, cables and connectors require insulation resistance values above 10 to the 13th power Ohms. At elevated mould temperatures, insulation volume resistivity drops rapidly, opening parasitic discharge paths that distort cavity pressure curves.

Polymer breakdown, absorbed moisture, and ionic residues at cable junctions speed up this discharge. High tool temperatures degrade the dielectric strength along the transmission line, turning what should be a purely capacitive circuit into an RC discharge network.

Every decade drop in line insulation resistance shortens the usable quasi-static measurement duration by a factor of ten.

The measurement circuit’s time constant tau governs signal retention during the injection cycle:

tau = R_ins (C_sensor + C_cable + C_range)

Here, R_ins represents the combined parallel insulation resistance of the sensor, cabling, and amplifier input, while the capacitances belong to the sensor crystal, line, and range capacitor. If continuous 200 degrees Celsius operation drags R_ins down from 10 to the 14th power Ohms to 10 to the 10th power Ohms, the time constant plummets from thousands of seconds to a fraction of a second. As a result, during an 8-to-10-second hold, the indicated cavity pressure falls toward zero while actual mechanical pressure in the cavity remains completely steady.

Parasitic Resistive Networks and Signal Decay

Electrical leakage leaves a negative baseline offset when the cavity depressurizes during part ejection. Because the charge amplifier bleeds charge through compromised insulation while under load, removing that mechanical force pulls the output signal below zero volts. This negative drift corrupts automated switchover routines that rely on integrated cavity pressure thresholds.

• Mineral Insulated Cabling prevents high-temperature insulation breakdown by utilizing compacted magnesium oxide dielectrics rated up to 450 degrees Celsius, avoiding the thermal degradation typical of standard PTFE lines.
• Hermetic Glass-to-Metal Seals block moisture and volatile polymer outgassing from entering the internal transducer cavity, protecting internal insulation resistance across extended production runs.
• Dynamic Drift Compensation algorithms calculate real-time charge decay rates during idle open periods, mathematically restoring peak hold values before the acquisition system triggers the machine screw position switchover.

Reset timing on the charge amplifier sets the reference baseline for each cycle. Tripping the reset relay while residual thermal charge lingers on the cable introduces an immediate offset into the upcoming measurement.

Engaging the input switch too early traps transient thermal recovery charges directly on the integration capacitor. Delaying the reset sequence until immediately before mould close allows the connector pins to reach thermal equilibrium, preventing baseline drift from compounding over continuous multi-day runs.

Tooling

Pocket geometry and machining tolerances establish the mechanical boundary conditions around the transducer body. Inaccurate pocket dimensions cause binding, localized thermal pinching, and uneven preload across the sensor face. At elevated temperatures, differential thermal expansion between the tool steel (typically 1.2343 or H13) and the transducer casing (often 1.4542 or 17-4 PH stainless) generates mechanical stress unrelated to actual melt pressure.

Bore diameters require machining within H7 tolerances, and pocket floor flatness must stay within 0.005 millimeters. Any deviation creates point loads against the transducer base, triggering mechanical hysteresis and zero-point drift as the tool cycles thermally.

Thermal Gradients and Mechanical Binding Constraints

Expanding tool steel exerts heavy radial clamping force on the sensor body whenever bore clearances fail to account for running temperatures. This radial binding restricts axial diaphragm travel, shifting calibrated sensitivity and pulling the baseline out of true over production shifts.

Differential expansion between the casing and the cavity plate significantly alters radial clearances at heat. Sizing pockets strictly for 20 degrees Celsius without expansion allowances invites mechanical interference once the tool reaches operating equilibrium.

Spacer rings behind indirect sensors deform under compression if thermal growth has nowhere to go. Bronze spacers expand faster than adjacent tool steels, increasing mechanical preload on indirect load cells and skewing the calibrated zero.

Indirect sensor pins in ejector assemblies require constant lubrication and tight sliding clearances to stop polymer flash from creeping into the bore. Ingress creates sliding friction that dulls force transfer to the sensor underneath, reading as an artificial drop in filling pressure and holding false pressure when the mould cracks open.

Drift

Isolating drift requires separating mechanical transmission losses from electrical leakage and thermal distortion. In the press, drift rate appears as changing output voltage over time under a steady mechanical load, recorded in pico-Coulombs per second or Bar per second. Evaluating these errors requires distinguishing quasi-static holding phases from dynamic injection peaks.

A 4-cavity high-temperature tool moulding PEEK electrical connectors with 1.2 millimeter walls demonstrates how drift develops across a cycle:

• Target Injection Pressure ~ 1200 Bar melt pressure at the cavity sensor location.
• Pack and Hold Time ~ 8.5 seconds at 800 Bar target holding pressure.
• Mould Temperature ~ 195 degrees Celsius continuously regulated via pressurized water units.
• Transducer Sensitivity ~ -4.0 pico-Coulombs per Bar nominal calibration factor.
• Cable-Amplifier System Resistance ~ 5.0 10 to the 11th power Ohms at operating thermal steady-state.

Here, the peak charge at full injection pressure reaches 4800 pico-Coulombs. Across the 8.5-second hold, this charge decays exponentially through the degraded insulation network:

Q(t) = Q_0 exp(-t / (R_ins C_total))

With total line and amplifier range capacitance at 1000 pico-Farads, the calculated RC time constant is 500 seconds. Over the 8.5-second hold duration, the charge loss totals:

Delta_Q = 4800 (1 – exp(-8.5 / 500)) = 80.9 pico-Coulombs

An 80.9 pico-Coulomb drop translates to an apparent 20.2 Bar loss on the monitor, irrespective of actual polymer holding pressure. In automated sorting systems, that 20.2 Bar drift will prompt false scrap rejections whenever monitoring windows are tightened to tolerances like plus or minus 15 Bar.

Thermal Recovery Cycles and Long Term Zero Shifts

Thermal cycling during continuous runs leads to baseline walking, in which the zero reference drifts from shot to shot. This drift reflects uneven thermal recovery across the sensor assembly while the tool is open for part ejection.

1. Melt contact raises diaphragm face temperatures by up to 140 degrees Celsius within 0.15 seconds of cavity filling.
2. Heat conducts through the diaphragm into the front crystal stack during the pack-and-hold phase.
3. Part ejection exposes the sensor face to ambient air or localized mold cooling, initiating rapid surface cooling.
4. Transducer housing temperatures equalize over the subsequent 10 to 15 seconds, creating transient thermal expansion stresses.
5. Amplifier reset commands execute during transient stress states, locking in false voltage offsets for the next production cycle.

Failing to account for thermal recovery during reset sequences leads to compounding baseline errors over subsequent cycles, eventually faulting the controller on false minimum fill limits.

Process stabilization requires charge amplifier reset commands to fire within a 100-millisecond window immediately preceding screw movement.

Uncorrected drift destabilizes process control. When screw switchover relies on compromised cavity pressure values, transition timing wanders, causing alternating short shots and parting-line flash across production shifts.

Protocol

Managing piezoelectric drift requires rigorous setup, calibration, and maintenance routines matched to high-temperature tooling demands. Running tools above 150 degrees Celsius without disciplined baseline control undermines cavity-based switchover and scrap sorting.

Operating procedures must include insulation resistance checks alongside scheduled in-press sensitivity validations. Cleaning routines must also prevent solvents from creeping into cable junctions, where they permanently destroy dielectric performance under heat.

Operational Integration and Preventative Maintenance

Reliable high-temperature instrumentation demands systematic attention across tool design, machine setup, and electrical hygiene.

1. Inspect cavity pocket dimensions with precision depth micrometers prior to sensor installation, verifying base flatness within 0.005 millimeters and corner radii clearances according to transducer engineering drawings.
2. Torque mounting bolts using calibrated torque wrenches to specified values, preventing asymmetric mechanical preloads from inducing sensor signal distortion.
3. Verify sensor and cable insulation resistance using a dedicated high-voltage tera-ohmmeter operating at 50 to 100 Volts direct current, rejecting assemblies showing values below 10 to the 12th power Ohms at ambient temperatures.
4. Route high-impedance cabling through dedicated tool wiring channels equipped with thermal barrier plates, maintaining isolation from high-current heater cartridge lines to eliminate induced electromagnetic noise.
5. Configure machine controller charge amplifier reset triggers to clear residual line charges within a narrow 50 to 100 millisecond window directly before injection initiation, preventing baseline walking across shifts.

Neglecting these installation and operational protocols corrupts sensor data, triggers unnecessary scrap rejections, and damages transducers early, resulting in preventable tool downtime and high repair costs.