Quick answer. Density gradient in a continuous PU panel line is caused by a mismatch between foam reactivity (cream, gel, and tack-free time) and line speed, laydown pattern, or top/bottom belt temperature differential. Fixing it requires cross-section density mapping per ASTM D1622 or ISO 845, then correcting either the polyol's catalyst package or the mixing head/belt settings depending on the gradient pattern found. Most gradients trace to insufficient overpack ratio or uneven mix laydown across the pour beam width, not a bad batch of raw material.
A buyer who calls a polyol supplier about "density gradient" is usually holding cut panel samples that failed a QC slice test. The core reads 34 kg/m³ at the edges and 41 kg/m³ at the center, or the top skin is visibly denser than the bottom third.
That pattern is diagnostic information, not noise. Before changing suppliers or reformulating from scratch, map the gradient shape first — it tells you whether the fix belongs in the polyol drum or on the machine.
What Actually Causes Density Gradient on a Continuous Line
A continuous double-belt line pours a liquid PU mix onto a moving facer, lets it rise and gel between the belts, and cuts the cured board to length. Any inconsistency in how fast the mix rises relative to belt travel shows up as a density gradient.
Reactivity vs. Line Speed Mismatch
If cream time is too long for the belt speed, the mix is still liquid when it enters the shaping zone and settles unevenly before gelling — producing a bottom-heavy core. If gel time is too short, the foam sets before it has finished expanding against the belts, leaving a low-density, coarse-cell top layer.
Mixing Head Laydown Pattern
Oscillating pour beams lay the mix down in overlapping fan or sweep passes across the panel width. A worn nozzle, an uneven traverse speed, or output drift at the mixing head (typically run at 150–200 bar impingement pressure) produces streaks of heavier and lighter foam running in the machine direction.
Belt Temperature Differential
Top and bottom belts are usually held a few degrees apart by design — commonly top belt 45–55°C and bottom belt 40–50°C. A differential beyond roughly 8°C accelerates skin formation on one face before the core has finished rising, locking in a top-to-bottom density split.
How to Diagnose the Gradient Before Changing Anything
Cut a full cross-section sample and slice it into 5–7 horizontal layers. Measure apparent density of each slice per ASTM D1622 or ISO 845, then plot density against slice position from bottom facer to top facer.
The shape of that plot points to a specific fix rather than a generic "reformulate" instruction.
| Gradient Pattern Observed | Likely Root Cause | Corrective Action |
|---|---|---|
| Top skin 8–15% denser than core | Gel time too short for belt gap / foam sets before full rise | Extend gel time 5–10 s via catalyst adjustment; confirm belt gap setting |
| Bottom layer denser than top | Cream time too long; mix pools before gelling | Shorten cream time 2–4 s; check bottom belt temperature is not running cold |
| Edge density 10%+ below center | Uneven laydown pattern from mixing head traverse or edge cooling | Recalibrate pour beam sweep width and overlap; check edge belt insulation |
| Streaking in machine direction | Mixing head output drift or nozzle wear | Service mixing head, verify output rate at 200–600 g/s depending on line width |
| Gradual drift over a production shift | Raw material temperature drift (polyol or MDI side tank) | Tighten tank temperature control, typically ±1°C around setpoint |
Formulation-Side Fixes: Matching Polyol Reactivity to Your Line
Once the gradient pattern points to a reactivity mismatch, the fix sits in the polyol's catalyst package and index, not in machine settings. Three reactivity bands cover most continuous lines in service.
| Line Speed Band | Target Cream Time | Target Gel Time | Target Tack-Free Time |
|---|---|---|---|
| Slow (≤6 m/min) | 10–14 s | 55–70 s | 85–100 s |
| Medium (6–9 m/min) | 8–11 s | 45–58 s | 70–85 s |
| Fast (9–12+ m/min) | 5–8 s | 35–48 s | 55–70 s |
A polyol blended for a slow 6 m/min line and run unmodified on a 10 m/min line will consistently show a bottom-heavy gradient — the reactivity was never wrong, the line speed simply outran it. This is the single most common mismatch we see when a buyer switches suppliers without disclosing actual line speed.
We custom-tune the amine/tin catalyst ratio to your stated line speed and belt gap before shipment rather than shipping a generic rigid foam polyol and leaving the customer to compensate with process changes. Standard MOQ for a reactivity-adjusted batch is 20 metric tons, with lead time of 15–20 days from formula confirmation to loading.
Process-Side Fixes: Overpack Ratio and Mixing Head Calibration
Overpack ratio — the ratio of in-panel core density to free-rise density — governs how much the foam is compressed against the belts as it cures. Running overpack below roughly 1.10 lets the foam under-fill at the edges before gelling, producing exactly the edge-low gradient pattern in the table above. Overpack above 1.25 wastes material and raises core density across the whole board, not just the gradient zone.
Target overpack for most rigid PU sandwich panel cores runs 1.10–1.20. If your gradient plot shows edge density consistently below center with formulation already matched to line speed, raise overpack incrementally by 0.02–0.03 per trial run and re-slice.
Mixing head laydown width should overlap by roughly 10–15% between adjacent passes; less overlap leaves under-mixed strips, more wastes reagent and can create localized high-density streaks from double coverage.
When It Is Not the Polyol
Before reformulating, rule out mechanical drift: MDI-side isocyanate handling and dosing accuracy matter as much as polyol reactivity, and isocyanate exposure controls at the mixing head are covered under OSHA isocyanate guidance — a poorly sealed or leaking dosing valve changes the actual index delivered to the mix even when the formulation on paper is correct.
A gradual gradient drift across a shift, rather than a consistent pattern from startup, almost always points to raw material tank temperature drift rather than formulation. Check this before requesting a reformulated batch — it saves a 15–20 day reorder cycle.
For panel lines running rigid foam cores, our rigid foam polyol systems are shipped with a line-speed-specific reactivity datasheet so the cream/gel/tack-free targets above are set at the blending stage, not adjusted after the first bad panel run.
FAQ
Q: What density variation across a panel cross-section is normal versus a real gradient problem?
A small variation between adjacent slices is expected from any liquid-to-solid rise process. A gradient becomes a quality problem when it exceeds roughly 5% deviation from your nominal design density between slices, or when it correlates with a specific belt or edge location run after run.
Q: Can density gradient be fixed with formulation changes alone?
Only if the diagnostic slice test shows a reactivity-vs-speed mismatch. Edge-density and streaking patterns are mechanical (laydown, belt temperature) and will not resolve through catalyst adjustment alone.
Q: How often should we re-run the slice test after making a change?
Re-test after every formulation, line speed, or belt temperature change, and again after any nozzle service. A single passing slice test after a change does not confirm stability — run it across a full shift to catch drift.
Q: Does switching polyol suppliers usually fix a density gradient?
Rarely, unless the new supplier tunes reactivity to your specific line speed and belt gap rather than shipping a standard catalog grade. A generic rigid foam polyol run on a mismatched line reproduces the same gradient regardless of brand.
Q: What overpack ratio should we target for a rigid PU sandwich panel core?
Most continuous lines run 1.10–1.20. Below that range, edges tend to under-fill before gelling; above it, you are compressing already-cured foam and raising average core density without improving uniformity.
