Non Contact Core Temperature Estimation across Variable Emissivity Recycled Polymer Blends during Rapid Furnace Cycles

Flux

Incoming polymer slabs enter the furnace conveyor at twenty-two degrees Celsius with uncalibrated surface qualities. Quartz heater banks run at eight hundred degrees Celsius, driving rapid energy transfer toward the resin faces over a thirty-second transit. Single-wavelength pyrometers set to an assumed emittance of 0.95 report a surface value of one hundred sixty-five degrees Celsius.

A contact thermocouple plunged into the mid-plane reads one hundred thirty-eight degrees. That twenty-seven-degree spread splits the process window. The outer boundary reaches melt while the interior remains rigid.

The mould closes, tensile stress tears the cold interior, and the press operator discards the shot.

Every non-contact temperature measurement inside an industrial heating oven relies on the Stefan-Boltzmann relation and Planck spectral distribution law. The detector collects photonic power emitted across a narrow optical band, converting microvolts to a temperature output through an assumed emittance setting. Pure virgin polyolefins display predictable spectral absorption characteristics in infrared regions.

Polyethylene exhibits strong carbon-hydrogen absorption bands at 3.43 micrometers and 6.8 micrometers, alongside a distinct optical window between 8 and 14 micrometers where thin sections behave as semitransparent media. Adding post-consumer regrind introduces flake-to-flake variations in pigmentation, mineral filler content, moisture, and surface roughness. These additions shift surface emittance between 0.68 and 0.94 without operator notification.

Plunging a calibrated mineral-insulated needle probe into a 4.0-millimeter sheet after thirty seconds of furnace dwell reveals a twenty-seven-degree lag below the pyrometer reading.

Incident Thermal Power across Heating Zones

Heater arrays project energy through quartz tubes or ceramic elements spaced between one hundred and two hundred millimeters above the conveyor line. The flux field consists of direct emission from the heating filament, secondary emission from furnace wall refractory insulation, and convection currents from air movement. Energy absorption within the polymer obeys the Beer-Lambert law: internal absorption equals incident flux multiplied by the exponential decay of the absorption coefficient through thickness.

Virgin high-density polyethylene permits infrared penetration several hundred micrometers beneath the face before photon absorption occurs. Post-consumer lots loaded with furnace carbon black or titanium dioxide alter this penetration depth instantly.

Carbon particles drop the optical penetration depth to less than twenty-five micrometers. Energy that once distributed through the upper half-millimeter of resin now concentrates entirely at the boundary interface. The surface heats three times faster.

Internal conduction remains governed solely by the bulk thermal diffusivity of the polymer matrix, which sits at roughly 0.18 square millimeters per second for high-density polyethylene. The furnace speed remains constant while the thermal gradient across the thickness steepens. Optical pyrometers interpret the high surface temperature as proof of thorough heating, triggering premature transfer to the forming tool.

Furnace Dwell inside Fast Production Cycles

Cycle times in high-speed sheet thermoforming and composite blank preheating rarely exceed forty-five seconds. Machine-hour rates on large eight-axis forming machines reach two hundred forty dollars. Plant managers resist lengthening furnace dwell times to establish equilibrium through the cross-section.

Conduction through polymers is inherently slow. Fourier numbers below 0.3 indicate severe internal gradients where boundary temperature offers no correlation to mid-plane enthalpy without rigorous mathematical reconstruction.

DIN EN 60584-1 specifies tolerance bands for calibration thermocouples, requiring lot verification clauses to stipulate contact probe insertion depths on test blanks to penalize pyrometer drift exceeding three degrees Celsius.

Skin

Melt-reprocessed flake streams arrive with unpredictable compounding histories. Flakes derived from blow-moulded milk jugs carry residual surfactants, degraded antioxidants, and high-density polyethylene fractions with narrow molecular weight distributions. Automotive shredder fractions add talc, calcium carbonate, carbon black, and random inclusions of crosslinked ethylene-vinyl acetate.

Each contaminant shifts the optical boundary behavior. A pyrometer pointed at a passing sheet does not measure temperature alone. It reads an integrated radiance signal composed of emitted thermal photons, reflected background photons from the overhead heater elements, and transmitted photons from underlying conveyor links.

Mineral fillers alter infrared response across intermediate wavelengths. Calcium carbonate exhibits sharp vibrational absorption bands near 7.0 micrometers and 11.4 micrometers. When a pyrometer operates across broad spectral bands, variations in calcium carbonate concentration change the apparent emittance by up to twelve percent.

The instrument interprets a decrease in emittance as a drop in surface temperature, prompting the furnace automation to boost heater element power. The sheet degrades, smoke fills the canopy, and volatile degradation products condense on the sensor optics.

Under ASTM D1505 density checks and differential scanning calorimetry, raw flake variability shifts spectral absorption profiles between production lots.

Surface Carbon Black Fractions in Regrind

Pigment distribution in recycled pellets never achieves virgin compounding uniformity. Agglomerates measuring between two and twenty micrometers float within the shear field during single-screw extrusion sheet line runs. These clusters aggregate near the extruded sheet surface due to melt rheology effects during die land exit.

An optical sensor focused on an agglomerated patch reads high local radiance. Two centimeters away, a resin-rich pool emits thirty percent fewer photons at identical thermodynamic temperature.

Extrusion lines running regrind compounds introduce variable gloss levels. Chill roll temperature fluctuations alter the crystalline skin morphology of polypropylenes. High chill roll temperatures yield slow cooling, producing large spherulites that scatter infrared wavelengths differently than rapidly quenched amorphous surfaces.

The surface emittance at 3.9 micrometers swings from 0.81 to 0.92 depending purely on the roll stack water temperature during sheet manufacture.

The compounding vendor always states that carbon black content remains stabilized at one percent across every delivered railcar.

Sensor

Single-band optical instruments fail when applied to recycled polyolefin heating processes. Operating an infrared sensor in the standard long-wave band between 8 and 14 micrometers exposes the detector to ambient reflection errors from ceramic emitters. Commercial furnaces generate ambient radiative temperatures exceeding six hundred degrees Celsius.

A polymer surface with an emittance of 0.85 reflects fifteen percent of that ambient power straight into the lens. The pyrometer calculates a temperature higher than reality.

1. Wavelength Selection Criteria require optical filtering at 3.43 micrometers for aliphatic C-H bonds, preventing heater element reflections from polluting the detector voltage.
2. Bandpass Optical Filters centered at 7.9 micrometers target ester and carbon-carbon linkages, isolating emission strictly to the first fifty micrometers of polymer skin.
3. Thermopile Array Integration provides spatial resolution across the entire transverse conveyor width, catching localized hot spots driven by pigment segregation.
4. Water Cooled Jacket Purges prevent ambient furnace draft temperatures from drifting the internal detector reference junction above forty degrees Celsius.

Calibrating pyrometers without accounting for background reflections introduces measurement errors exceeding twenty degrees Celsius.

Will Dual Wavelength Pyrometry Isolate Surface Emittance Drift?

Ratio pyrometers measure emitted energy across two adjacent spectral bands, dividing the signals to cancel the absolute emittance value. This technique functions reliably on greybody metals where emittance remains identical at both wavelengths. Polymers are non-greybody emitters.

Spectral emittance changes sharply with wavelength depending on polymer chemistry, pigment loading, and filler type. The ratio of emittance between 3.43 micrometers and 4.2 micrometers does not stay constant as carbon black loading changes. Two-color pyrometers introduce mathematical errors when applied to filled polyolefins unless continuous spectral correction factors update in real time.

Narrow-band pyrometry targeting strong absorption peaks provides superior stability. At 3.43 micrometers, high-density polyethylene absorption is sufficiently strong that emittance approaches 0.96 even on thin films. The resin behaves as an opaque body.

Transmitted energy drops to zero, removing reflections from the steel conveyor belt beneath the sheet. Background reflections from overhead quartz filaments still enter the optical path, necessitating physical sighting tubes or active background radiation compensation circuits.

Absorption Bands in Hydrocarbon Backbones

Targeting specific infrared bands changes what depth the sensor actually interrogates. At 3.43 micrometers, the optical penetration depth measures less than thirty micrometers; the sensor observes only the extreme outer skin. At 10 micrometers, unpigmented polyolefins become semitransparent, permitting the detector to capture photon emissions originating from deeper layers within the sheet.

The instrument reads an integrated volume average rather than a pure surface boundary. When regrind pellets enter the feed throat, the absorption coefficient at 10 micrometers changes instantly. The volume average collapses to a surface measurement, throwing uncorrected process controllers into unstable heating cycles.

Disregarding optical transmission changes causes cold sheet cores to crack during draw phases, damaging expensive machined aluminium tooling and forcing unscheduled toolroom repairs.

Filter

Direct non-contact measurement of internal slab temperatures during thirty-second furnace cycles is physically impossible with surface optical devices. Radiation originating from the interior cannot escape without undergoing absorption and re-emission through outer layers. The solution pairs non-contact surface pyrometry with real-time numerical state estimators.

An observer model runs parallel to the physical machine, executing a one-dimensional transient heat conduction equation across fifteen discrete spatial nodes through the sheet thickness.

Consider a worked calculation for a heavy-gauge thermoforming cell. Assume a 4.0-millimeter high-density polyethylene sheet heated from twenty degrees Celsius to forming state. Thermal conductivity equals 0.23 Watts per meter-Kelvin, density equals 950 kilograms per cubic meter, and specific heat capacity equals 2,100 Joules per kilogram-Kelvin.

Thermal diffusivity measures 1.15e-7 square meters per second. The Biot number exceeds 1.2 under rapid heating, proving that steep internal thermal gradients dominate. Surface pyrometry reads 160 degrees Celsius at twenty-four seconds.

The forward finite-difference node equation calculates the mid-plane node at 128 degrees Celsius. The estimator fuses the model projection with the physical surface sensor reading, adjusting the internal profile through a Kalman gain matrix.

The Kalman filter models system uncertainty and measurement noise dynamically. Surface pyrometer noise stems from emittance fluctuations, draft turbulence, and electrical interference from thyristor power packs. Model uncertainty originates from unknown flake blends, changing latent heat during polymer melting, and unmeasured sheet thickness tolerances.

DIN 16742 allows extruded sheet thickness to vary by plus or minus six percent. A six percent thickness increase extends required conduction time to the mid-plane by twelve percent, because thermal diffusion time scales with the square of thickness.

Should State Estimators Track Internal Thermal Gradients?

Standard proportional-integral-derivative loops fail because they react only to the surface pyrometer. As the surface approaches the setpoint, the controller throttles heater power while the mid-plane remains too cold for tool engagement. A state observer tracks the reconstructed mid-plane temperature node directly.

The controller maintains high surface flux until the estimated interior approaches forming temperature, then drops emitter output to zero to allow thermal equalization throughout the thickness.

Nonlinear enthalpy methods accommodate the latent heat of fusion during crystal melting. High-density polyethylene absorbs roughly two hundred kilojoules per kilogram between 125 and 135 degrees Celsius. During this transition, temperature rise stalls while crystals absorb energy.

Linear thermal models fail here. The observer incorporates temperature-dependent specific heat curves obtained through baseline differential scanning calorimetry testing. When recycled lots contain mixed polypropylene fractions, a second melting peak appears near 165 degrees Celsius, shifting energy demand.

Transient Conduction and Midplane Thermal Lag

Tracking the thermal state requires continuous updating of boundary conditions. Pyrometers mounted at the entrance, middle, and exit of the furnace supply sequential data to the observer algorithm. If surface emittance shifts suddenly due to a dark regrind patch, a single-pyrometer filter produces corrupted internal estimates.

Dual-sensor systems viewing both top and bottom sheet faces cancel uneven heater zone imbalances.

Process engineering teams still debate whether low-cost microcontroller hardware running reduced-order state space models can accurately capture non-Newtonian viscous dissipation inside rapid forming stages without introducing numerical divergence.

Sheet

Tooling commitments require reproducible mechanical behavior during forming cycles. When the sheet index mechanism moves the heated polymer from the oven into the clamp frame, gravitational sag begins immediately. A sheet with an overheated surface and a cold interior experiences unpredictable stress distribution.

The hot outer layers stretch under their own weight, thinning prematurely before vacuum application. The cold internal core resists deep drawing over tool corners, causing localized wall thinning, poor impact resistance, and high scrap rates.

Sag measurement sensors verify thermal readiness mechanically. Ultrasonic or laser distance transducers positioned below the sheet track the catenary sag profile during the final five seconds of heating. High-density polyethylene exhibits distinct viscoelastic relaxation once the interior surpasses the crystalline melting point.

Linking non-contact optical temperature estimation with physical sag displacement data provides a dual-validation mechanism. If the state estimator signals that the interior is at forming temperature but the sag sensor shows zero deflection, the material retains unmelted crystalline structures, indicating a formulation shift in the regrind lot.

Tooling amortisation schedules collapse when unverified recycled feedstocks cause twenty percent scrap rates on multi-cavity thermoforming moulds.

Sag Kinetics at the Forming Station

Temperature gradients across sheet thickness govern catenary sag velocity. A homogeneous thermal profile yields uniform biaxial stretching during tool entry. A steep gradient results in surface tearing, surface haze, and micro-cracking across deep-draw male features.

Aluminum production tools running ten cavities cannot compensate for incoming temperature variations between parts. Cavity three draws properly while cavity eight blows out because the regrind emittance shifted along the outer conveyor edge.

Tool designers cut draft angles and corner radii based on predictable shrinkage figures. Virgin high-density polyethylene contracts between 1.5 and 2.5 percent during cooling. Post-consumer lots vary between 1.8 and 3.2 percent depending on cooling rates and polymer orientation.

Running parts with unresolved through-thickness thermal gradients locks severe residual stresses into the moulded product, driving warpage after part ejection and dimensional failure during downstream assembly.

Clamp Frame Transfer and Part Ejection

Transfer speeds from the furnace to the press platen cannot be delayed without heavy thermal penalties. In ambient shop air, a 4.0-millimeter sheet loses up to twelve degrees Celsius of surface temperature in three seconds through natural convection and radiation. The interior remains hot.

The thermal gradient reverses: the skin becomes cool and tough while the mid-plane remains liquid. Clamping this sheet produces poor surface replication of mold textures and high rejection rates during first-article inspection under DIN EN ISO 9001 audits.

Cold sheets break tools while overheated sheets melt into the vacuum holes.