In glass heating, the most expensive defects are rarely caused by “not enough heat.” They are caused by temperature non-uniformity at the critical point (right before quench or forming). Practical tempering guidance highlights that capturing a thermal image as each lite enters the quench enables adjustments to the heating profile to maintain a uniform glass temperature—a direct pointer to where zoning and stability work should focus.
This article gives a practical method to deliver uniform heating across width and over time using IR zoning, measurement discipline, and a commissioning workflow that production teams can repeat.
What “uniform heating” means in production
Uniform heating is two requirements at once:
- Across the lite (spatial uniformity):edge vs center, and lane-to-lane across the width.
- Across time (stability): the same recipe produces the same temperature profile throughout the shift, not only after “everything is warm.”
If you only solve one of these, defects will return as soon as speed, glass thickness, or ambient conditions change.
Why IR zoning is the right lever
IR is fast and local. That is the advantage—and also the risk.
- Advantage: You can add “trim heat” exactly where the glass is consistently cold (edges, specific lanes, coated regions).
- Risk: If energy is not distributed evenly, IR can create hot lanes quickly and repeatedly.
To manage this, you need zoning that respects the physics of radiation: radiative transfer depends on emissivity, geometry, and “how much of the heater the glass sees” (view factor), not only nameplate kW.
Zoning architecture that works on glass lines
Most successful glass IR systems use two dimensions of control:
Cross-width zoning (left → right correction)
Use it to correct persistent patterns such as:
Machine-direction staging (entry → exit stability)
Use it to shape the heating profile so the glass arrives at the critical point with a tight temperature band:
- controlled ramp behavior
- equalization behavior before quench/press
- reduced peak gradients that imprint distortion
Measurement strategy: you cannot zone what you cannot measure
If your temperature signal is unreliable, zoning becomes guesswork.
Two commonly used approaches in the glass industry are:
- Thermal imaging / line scanners at a critical location to visualize the temperature field across the lite.
- Thermal profiling systems designed for tempering/bending to capture “entry-to-exit” profiles and support recipe optimization, including for coated low-e glass.
For bending specifically, industry application guidance notes that quality issues can occur when blanks are non-uniformly heated prior to bending, reinforcing the need to verify uniformity immediately before the forming step.
A repeatable commissioning workflow for zoning and stability
The goal is a controlled loop: Map → Trim → Verify → Lock.
- Pick a representative lite and a worst-case lite (largest size, most sensitive thickness, or highest throughput setting).
- Choose a critical point to map (pre-quench entry or pre-press/bend). Tempering guidance explicitly points to quench entry thermal imaging for profile adjustment.
- Run a baseline pass with zoning trims set to neutral, and record the temperature field (image/profile).
- Classify the pattern as one of the following: edge-cold, edge-hot, lane hot spot, one-side bias, mixed.
- Apply small cross-width trims to correct the dominant pattern (do not change speed and trims simultaneously).
- Re-map and compare until the temperature band tightens at the critical point, then lock the trims into the recipe.
- Prove stability by repeating the same test later in the shift, and once at a second operating point (higher speed or different thickness).
Practical zoning rules that reduce trial-and-error
Use simple rules that prevent over-correction:
- Trim first, then stage. Fix cross-width bias before you reshape the full heating curve.
- Small changes win. A few percent trim can remove most lane bias; large trims often create new distortions.
- Equalization is not wasted time. A short equalization behavior can reduce gradients that become “locked in” during quench/press.
Zoning design checklist (operations-friendly)
Item | What “good” looks like | What fails |
Zone width | Matches recurring bias patterns (edges vs center, known lanes) | Zones too wide to correct, or too narrow to control stably |
Trim limits | Defined ± range to prevent over-correction | No limits; operators chase defects with large swings |
Measurement point | At the true critical location (pre-quench/pre-press) | Measuring too early where gradients haven’t developed |
Recipe ownership | One “gold recipe” + controlled variants | Many ad-hoc recipes per shift |
Troubleshooting: pattern → cause → first move
Temperature pattern you see | Likely cause | First move |
Edges consistently colder | edge loss + geometry | add edge trims; confirm distance uniformity at edges |
One lane always hotter | reflector geometry / alignment | reduce that lane trim; inspect reflector and mounting geometry |
Bias changes through shift | measurement drift or heat-up drift | verify sensor cleanliness/aim; compare early vs late shift maps |
Good map, bad bending | non-uniformity right before bend | move measurement closer to bend/press; verify uniformity pre-bend |
Case example (representative): hot lane eliminated with trims + measurement discipline
Starting issue: recurring optical distortion aligned with one cross-width lane.
What the map showed: a stable hot lane at the critical point, repeatable across batches.
Fix: reduced the single lane’s trim within a capped range, then re-mapped to confirm the temperature band tightened at the critical point. This approach matches tempering guidance that uses thermal imaging at quench entry to adjust the heating profile for uniform temperature.
FAQ
Where should I measure to tune zoning: inside the heater or right before quench/bend?
Tune zoning where defects are “locked in.” Tempering guidance emphasizes thermal imaging at quench entry to adjust the heating profile for uniform temperature.
Do I need cross-width zoning if I already have a stable furnace profile?
If you see repeatable edge/center or lane bias, yes. Zoning is the most direct control lever for spatial uniformity.
Why does zoning sometimes “fix one defect but create another”?
Over-correction. Radiation is geometry-driven; aggressive trims can create new gradients. Keep trims small and verify with maps.
How do I handle coated/low-e glass?
Use a measurement approach proven for coated glass and treat it as a separate recipe family. Profiling systems are explicitly marketed for accurate profiles across glass types including coated low-e.
Call to action
Share your glass thickness, lite size, line speed, belt/roller width, and whether the issue is edge bias or lane hot spots. YFR can propose an IR zoning layout and a mapping-based commissioning recipe to tighten temperature uniformity before quench or bending.
Data sources
Last modified: 2026-01-22