Sandwich panel thickness inconsistency usually comes from press-gap errors, unstable core feeding, temperature drift, or a mismatch between chemical output and line speed. Start by mapping where the variation occurs. A panel that’s thick along one edge needs a different correction from one that slowly gains 2 mm over a 30-minute production run.
Find the Thickness Pattern
To troubleshoot sandwich panel thickness inconsistency, map the defect before adjusting the line. Measure cooled panels at fixed cross-width points, compare the pattern with press gap and process logs, correct one variable, then run a short confirmation batch. Edge bias usually indicates geometry; gradual drift points to temperature, flow, or speed.
- Stop random adjustments. Save the active recipe and record the current press gap, line speed, temperatures, pressure readings, and material batch numbers.
- Select 10 consecutive panels from the affected run. Mark each panel with its production time and position.
- Measure at five fixed points: 50 mm from each edge, 25% and 75% across the width, and the centerline.
- Repeat the measurements near the panel’s leading end, middle, and trailing end. Use the same datum face each time.
- Plot the results against production time. Look for edge bias, center crown, periodic waves, or gradual drift.
- Change one machine setting, produce a short trial, and compare the new measurement map with the original.
Don’t rely on a single caliper reading at the panel edge. Edge roll-forming, protective film, and local skin distortion can shift that reading without representing the full panel.
For rigid PU or PIR cores, ISO 1923:1981 provides a recognized method for determining linear dimensions. Your contractual tolerance should still come from the panel drawing, purchase specification, or applicable product standard.
How do you measure panel thickness?
Measure a cooled panel on a flat support at repeatable points across its width and length. Use a depth gauge, large-jaw caliper, or fixed inspection fixture with suitable resolution. Avoid clamping hard enough to compress the core, and record skin profile effects separately from the structural panel thickness.
Match Defects to Causes
The defect shape carries useful information. A 100 mm panel measuring 99.7 mm on the left, 101.5 mm at the center, and 100.1 mm on the right suggests press crown or uneven internal foam pressure. If the same panel stays 1.5 mm low only along its left edge, press parallelism becomes the first check.
| Thickness pattern | Likely causes | First plant-floor check |
|---|---|---|
| One edge consistently thick | Press gap out of parallel, worn guide, uneven side containment | Compare actual left and right gaps with calibrated blocks |
| Center thicker than both edges | Belt crown, platen distortion, excess foam pressure | Measure the hot operating gap across the full width |
| Gradual increase during a shift | Rising component temperature, viscosity change, output drift | Plot material temperature and kg/min against production time |
| Repeating high-low wave | Chain pitch error, roller runout, sprocket wear, belt contamination | Measure the distance between repeated defects |
| Thickness changes after speed adjustment | Chemical or core feed didn’t follow line speed | Recalculate mass flow for the new speed |
| Local bumps or depressions | Rockwool joint, adhesive lump, foam void, debris on belt | Cut and inspect the defect location |
| Leading end too thick | Foam reaches the press before speed stabilizes | Review startup timing and reject length |
| Trailing end too thin | Material feed stops before the panel clears the press | Check shutdown sequence and delay settings |
A repeating defect is especially revealing. If the distance between peaks matches one conveyor slat pitch or one sprocket revolution, mechanical inspection works better than changing the foaming recipe.
Why is one panel edge thicker?
A consistently thicker edge usually means the upper and lower conveyor faces aren’t parallel, or side containment is allowing the core to expand differently. Measure both edge gaps at operating temperature. If the gap is correct, inspect guide rails, edge blocks, facing tension, and core distribution.
Check Double-Belt Press Geometry
The HMI can display 100.0 mm while the actual exit gap is 101.3 mm. Position sensors report actuator movement; they don’t prove that the conveyor faces are flat, parallel, or free from thermal distortion.
Isolate the machine before entering guarded areas or placing gauges near chains and platens. The OSHA lockout/tagout rule, 29 CFR 1910.147, defines energy-control requirements for servicing equipment with hazardous stored or moving energy. Your plant’s local rules may require further controls.
Once the line is safe, check the gap at the entry, two intermediate positions, and exit. Across a 1,200 mm panel width, useful reference points are approximately 50, 400, 800, and 1,150 mm from one edge. Calibrated blocks and feeler gauges expose errors that an HMI value can hide.
Cold measurements alone aren’t enough. Run the press to its normal production temperature, stop it under the approved procedure, and repeat the survey. A frame or platen may move as it heats. Also inspect:
- Upper and lower chain tension
- Conveyor slat joints and worn bearings
- Foam or adhesive deposits on belt faces
- Left-right actuator synchronization
- Encoder counts for the upper and lower belts
- Calibration of gap sensors after maintenance
If you’re checking a recently installed PU sandwich panel production line, compare these readings with the commissioning record. That baseline tells you whether the line has shifted mechanically or whether the fault started after a recipe, raw-material, or speed change.
Stabilize Core Feed and Foaming
For PU and PIR panels, thickness is set by the double-belt gap, but stable filling depends on chemical output, reaction timing, and line speed. Too little material can leave low-density areas or voids. Excess output raises internal pressure and may lift the upper facing before the foam stabilizes.
Use this calculation as a first check:
Required core output (kg/min) = panel width (m) × core thickness (m) × line speed (m/min) × target core density (kg/m³)
A 1.0 m wide, 100 mm thick panel running at 5 m/min with a target density of 40 kg/m³ needs about 20 kg/min of finished core mass before allowing for edge trim and normal process loss. If speed rises from 5 to 6 m/min, core demand rises by 20%. Leaving the metering output unchanged will change fill conditions immediately.
Compare actual component output with the approved formulation. Check both material temperatures against the chemical supplier’s processing window. For example, if the validated recipe calls for 22 ±2°C, one component at 19°C and the other at 27°C isn’t an acceptable average. The viscosity and pump delivery of each stream matter separately.
Verify the A-to-B ratio by mass-flow data or a supplier-approved calibration test. Then check mixing pressure, nozzle condition, pour pattern, and the delay between application and press entry. A partly blocked mixing head can distribute the correct total kg/min unevenly across the width.
Can line speed change thickness?
Yes. Line speed changes the amount of core material delivered per square metre and the time available for expansion inside the press. Any speed adjustment should trigger a matched output change and a check of cream, gel, and tack-free timing under the actual component temperatures.
Rockwool needs a different diagnosis. Check incoming slab thickness, lamella cutting accuracy, joint gaps, compression setting, and adhesive distribution. A thick adhesive bead can create a local high point; a poorly closed lamella joint produces a depression after pressing.
On a PU and rockwool panel production line, saved changeover recipes should separate PU/PIR metering settings from rockwool feed, glue, and compression parameters. Copying the same press setup between core types often creates avoidable variation.
Does core density change thickness?
Core density can affect thickness when material output and press conditions aren’t balanced. Low output may leave voids or weak contact with the facings, while excess output increases foam pressure. Density should be checked from cut samples; panel thickness alone can’t confirm that the core mass is correct.
Prove the Thickness Correction
A good-looking first panel proves very little. The line must hold thickness after temperatures, pump output, and conveyor load have stabilized.
Use a controlled confirmation run:
| Verification item | Practical acceptance check |
|---|---|
| Cross-width thickness | All five measurement points stay inside the internal action band |
| Production trend | Ten consecutive panels show no upward or downward drift |
| Core condition | Cut sample has no voids, crushed areas, open joints, or poor skin contact |
| Metering stability | Ratio and total kg/min remain within the validated recipe limits |
| Press condition | Actual hot gap agrees with the corrected mechanical setting |
| Surface quality | No belt marks, edge lift, facing waves, or adhesive bumps |
Set the internal action band tighter than the customer’s rejection limit. If the contract permits ±2 mm, waiting until readings reach 1.9 mm leaves no room for measurement error or process movement. An internal warning at ±1.0 mm is a practical starting point, though the right limit depends on gauge repeatability and the panel specification.
Measure at consistent cooling times. Record one set at the line exit for process control, then measure again after 30 to 60 minutes for release. During a new foam-system qualification, another check after 24 hours can expose post-expansion or shrinkage that an immediate reading misses.
Keep thickness acceptance separate from bond and density acceptance. A panel can measure exactly 100 mm while hiding a low-density pocket or weak facing adhesion.
Prevent Thickness Drift
Most repeat defects return because the corrected setting never becomes part of the production standard. Save the verified recipe under a controlled product code that includes core type, nominal thickness, width, facing profile, and approved speed range.
A simple inspection schedule works:
- At startup and each product change: complete a five-point thickness map.
- Every 30 minutes: record panel thickness, speed, component temperatures, and total output.
- Each week: inspect belt surfaces, chain condition, guides, and side containment.
- After mechanical work: verify actual press gap before releasing the line.
- At planned calibration intervals: compare gap sensors, flow readings, and temperature probes with reference instruments.
- For every raw-material lot change: retain the lot number and one qualified panel sample.
When we investigate a thickness complaint, time-aligned records are far more useful than a photo of one rejected panel. A Siemens S7-1200 or S7-1500 trend export, for example, can show whether the defect began with a temperature rise, speed command, pressure fluctuation, or operator setting change.
FAQ
What thickness variation is acceptable?
Use the tolerance stated on the approved panel drawing, contract, or applicable product standard. Set the factory action limit tighter than that tolerance so operators can correct drift before panels become rejectable.
Can uneven metal skins affect readings?
Yes. Roll-forming distortion, coil camber, profiling depth, and excessive facing tension can change local readings. Measure from a defined datum and inspect the bare facing profile before changing the press gap.
When should production be stopped?
Stop when thickness crosses the internal action limit, trends steadily toward the rejection limit, or appears with voids, weak adhesion, belt damage, or abnormal press pressure. Continuing production turns a diagnosable fault into a full batch of scrap.
For faster remote diagnosis, send YZ Welding a 10-panel measurement map, the HMI trend export, raw-material lot details, and photos of a cut core section. YZ Welding can then separate recipe faults from mechanical alignment problems and prepare the right commissioning or service checks before a technician arrives.
