Applications / non-contact thickness measurement of enamel coatings at every state of the process

01 — Introduction

Enamel coating: vitreous glass fused to metal and glass

Vitreous enamel — also called porcelain enamel — is a thin layer of glass fused onto a substrate at high temperature. A finely milled glass frit is applied as a wet slurry (slip), an electrostatic dry powder, or a screen-printed paste, dried, and then fired at roughly 800–900 °C, where it melts and bonds chemically to the part. The result is a hard, glossy, chemically inert surface that resists corrosion, abrasion, high temperature, and most acids and alkalis.

Because of this combination of properties, enamel is the finish of choice across a wide range of industries: cast-iron and steel cookware and bakeware, domestic-appliance oven muffles and trays, glass-lined hot-water storage tanks and boilers, chemical and pharmaceutical reactors, heat exchangers, architectural panels and signage, sanitary ware, and decorative prints on vitroceramic cooktops.

Cookware & bakewareAppliances & ovensWater-storage tanksChemical & process vesselsVitroceramic cooktopsArchitectural panels

An enamelled part usually carries one or more functional layers: a ground coat applied first for adhesion and corrosion protection, followed by one or more cover coats that provide the final colour, gloss, and chemical resistance. Total fired thickness commonly ranges from around 10 µm for thin screen-printed decorative layers up to about 1 mm — and, more rarely, beyond — for heavy glass linings, ground coats, and water-tank linings.


02 — Quality Requirements

Why controlling enamel thickness is essential

Coating thickness is one of the most critical quality parameters of an enamelled part, and the firing step makes it unforgiving: once the enamel is fired, it is chemically bonded to the substrate and cannot be reworked. A non-conform part is scrapped, not reprocessed.

Too thin: protection failures and pinholes

A layer that is too thin leaves the substrate exposed through pinholes and bare spots, eliminating the corrosion and chemical protection that the enamel is there to provide. On steel and cast iron this leads to rust-through and early failure; on water tanks and chemical vessels, even a localised thin zone can compromise the entire part.

Too thick: spalling, warping and waste

An over-applied layer is just as damaging. Excess enamel is prone to spalling and chipping, poor adhesion, and panel warping during firing, as well as defects such as fish-scaling. It also wastes expensive frit and increases firing energy.

Consistency: the key to process feedback

Beyond pass/fail limits, mapping the layer uniformity across the part exposes application problems that are invisible to the eye — spray-pattern drift, electrostatic Faraday effects on powder, uneven slip drainage, or dipping non-uniformities — long before they reach the furnace.

10 µm – 1 mm+
Enamel thickness range (up to ~1 mm, rarely beyond)
Up to 15%
Scrap driven by thickness alone
~800–900 °C
Firing — the point of no return
0.5–3%
Typical RMS measurement precision

03 — Measurement Challenges

Why non-contact, non-destructive measurement is essential

The constraints of enamelling lines make conventional thickness measurement inadequate. Reliable, representative, and actionable data requires overcoming several fundamental challenges.

The problem with contact techniques

The most common thickness gauges — eddy current and magnetic induction probes — are contact-based and require a consolidated coating over a metallic substrate. They cannot be used on the fragile, unconsolidated slurry, powder, or biscuit states, which the probe tip would simply destroy, and they are useless on glass and vitroceramic substrates. In practice they also require a trained operator and only a small, statistically unrepresentative sample of parts is measured.

The case for measuring before firing

Because contact gauges only work after firing, any non-conformity is detected at the one stage where nothing can be done about it. Measuring the wet, powder, or biscuit state — before the part enters the furnace — transforms quality control: an out-of-tolerance part can be reworked or stripped and re-applied before the energy-intensive firing cycle is spent.

Figure — The enamelling sequence: measurable at every step

Application slurry · powder · paste measurable & recoverable Drying biscuit state measurable & recoverable Firing ~800–900 °C point of no return Solid layer bonded — scrap if wrong measurable (final QC) ◆ Measurable by LPTR — before firing, the part can still be reworked Up to ~15% of scrap is driven by thickness alone — detected today only after firing, when rework is impossible.

Highlighted states can be measured by laser photothermal radiometry — long before the part is fired.

The need for in-line automation

Modern enamelling lines — appliance panels on conveyors, tanks, or batches of cookware — run at high throughput. A truly effective thickness control must measure automatically, track parts as they move, and feed data in real time to the line supervision system. This level of integration is impossible with contact technology.


04 — Technology Comparison

Why laser photothermal radiometry is the ideal solution

Laser photothermal radiometry (LPTR) — the patented technology developed by Enovasense — resolves all of these limitations in a single, compact, integrable sensor. A power-modulated laser beam generates a controlled thermal wave at the coating surface; the wave diffuses into the layer and back-diffuses at the layer–substrate interface, and an infrared detector captures the returning thermal flux. The phase shift between the emitted laser and the thermal response is correlated to the coating thickness, while the signal amplitude reveals defects such as porosity or lack of adhesion.

Because the measurement relies on heat diffusion rather than contact or magnetic coupling, it works on cured and uncured enamel alike, is independent of colour and pigment, and — crucially for enamel — works on glass and vitroceramic substrates where eddy current cannot.

Figure — LPTR principle: phaseshift and amplitude

Modulated laser excitation Thermal response — phase-shifted & attenuated phaseshift φ → thickness A → defects

The phaseshift correlates to layer thickness; the amplitude correlates to defects in the layer.

Criteria Enovasense laser photothermal (LPTR) Eddy current / magnetic induction Cross-section microscopy (destructive)
Non-contact measurement Yes No — probe contact required No — part is cut
Measurement before firing (slurry · powder · paste · biscuit) Yes — every process state No — contact destroys the unfired layer No
Works on a glass / vitroceramic substrate Yes — absorbs into heat No — needs a metallic substrate Partial — fragile sample prep
Non-destructive Yes Yes No — part is destroyed
Defect / porosity / adhesion detection Yes — amplitude signal No Partial — visual, sampling only
Full-field / 100% mapping Yes — with Field Sensor Sampling only Sampling only
Tolerant to rough, porous biscuit surfaces Yes Partial — flat zones only N/A
In-line automation capability Yes — designed for automated lines Difficult — manual operation No
Measurement speed <1 s per point ~1 s per point — manual repositioning Hours per sample

A further practical advantage at the fired state is that a single calibration covers all enamel colours on a given substrate, so many production references can be measured without recalibrating between colour changes.


05 — Metrological Performance

How Enovasense laser photothermal performs on enamel

The performance of LPTR on enamel has been validated across four representative industrial applications spanning every process state — slurry, powder, paste, biscuit, and fired — and several substrates, from cast iron and steel to vitroceramic glass.

Application 1 — Dutch ovens (biscuit & fired, on cast iron)

A first anticorrosion enamel layer on cast-iron Dutch ovens, applied by wet slurry and measured both at the dried biscuit state and after firing.

Ten samples were prepared at various enamel thickness levels. At the dried biscuit state, ten photothermal measurements were taken at different locations on each sample to extract the phaseshift; the samples were then fired and measured by cross-section microscopy as the destructive reference. Averaging per sample, the biscuit-state phaseshift correlates to the final fired thickness at 98.61% — so, once calibrated, the sensor predicts the fired thickness directly from a biscuit-state reading, with an average absolute difference of 1.4 µm to the cross-section. Repeating the test at the fired state gave a comparable 98.42% correlation (1.2 µm average difference). Similar results were obtained on the second, colour cover-coat layer.

Sensor parameters

Sensor Enovasense Point
Wavelength 980 nm
Power 10 W max
Measurement time 700 ms
Distance 35 mm · spot Ø3.3 mm

Layer structure — Application 1

Enamel Cast iron 20–80 µm

First (anticorrosion) layer on cast iron · 20–80 µm

98.6%
Correlation — biscuit state vs fired reference
1.4 µm
Average absolute difference to cross-section
98.4%
Correlation — fired state vs reference
1.2 µm
Average absolute difference to cross-section

Figure — Biscuit state: Enovasense vs reference thickness

0 0 20 20 40 40 60 60 80 80 1:1 line R² = 0.9861 Reference thickness — cross-section (µm) Enovasense thickness — biscuit state (µm)

Enovasense thickness (phaseshift converted via the biscuit-state calibration) vs the cross-section reference thickness at the fired state.

Figure — Fired state: Enovasense vs reference thickness

0 0 20 20 40 40 60 60 80 80 1:1 line R² = 0.9842 Reference thickness — cross-section (µm) Enovasense thickness — fired state (µm)

Enovasense thickness (phaseshift converted via the fired-state calibration) vs the cross-section reference thickness at the fired state.

Result  Predict the final fired thickness directly from a measurement at the biscuit state — before firing.

Application 2 — Oven trays (powder & fired, on steel)

A dry powder enamel coating on steel oven trays, mapped at the powder state and after firing. Eddy current cannot measure the powder state — contact would wipe the coating away — whereas photothermal maps the whole part and spots overspray at the bottom edge before any firing energy is spent.

Two trays were tested: Part 1 at the powder state, then fired and measured with an eddy-current reference, and Part 2 directly at the fired state.

Correlating the powder-state photothermal readings on Part 1 to the eddy-current values across thickness areas from 100 to 350 µm gave a 96% correlation; the fired-state comparison on Part 2 reached 97%. 

A full thickness map was also built at the powder state by stepping the Point sensor over a 5 mm grid covering the whole tray — the resulting colormap measured every position correctly and clearly revealed an overspray zone at the bottom of the part, all before any firing energy was spent.

Figure — POWDER STATE : ENOVASENSE THICKNESS COLORMAP


Colormap of coating thickness values measured at powder state on part 1

Figure — FIRED state: Enovasense THICKNESS COLORMAP


Colormap of coating thickness values measured at fired state on part 2

Sensor parameters

Sensor Enovasense Point
Wavelength 980 nm
Power 10 W max
Measurement time 1 s
Distance 35 mm · spot Ø3.3 mm

Layer structure — Application 2

Enamel Steel 100–300 µm

Dry powder enamel on steel panel · 100–300 µm

96%
Correlation at powder state (100–350 µm, vs eddy current, fired)
97%
Correlation at fired state (100–350 µm, vs eddy current)
Result  Map the complete tray at the powder state and catch defects such as overspray before any firing energy is spent.

Application 3 — Vitroceramic glass (paste state)

A thin screen-printed enamel on vitroceramic glass — around 10 µm — captured at the paste state with the Field Sensor in a single full-field snapshot. No glass reference level is needed and no sensor movement is required; the field map resolves the raised edges on each printed dot, a typical screen-printing signature.

The conventional alternative on glass is to measure the height of the printed markings against the bare-glass level with a profile sensor — but this is only accurate close to the marking edges where a glass reference is available, requires mechanical movement and tilt referencing, and a contact probe is not advisable on the fragile paste. The Field Sensor instead images the whole 21 × 15 mm area in a single one-second snapshot: each of its 384 × 288 pixels returns a local thickness over a 55 µm spot. On a vitroceramic glass printed with dots of various sizes, measured before curing, the 1-second colormap resolves the raised bumps at the edge of each dot — a characteristic screen-printing pattern — with no glass reference level and no moving sensor.

Figure — Picture of the measured part


Picture of the measured part

Figure — PASTE state: Enovasense FIELD SENSOR THICKNESS MAPPING


Colormap of the coating thickness values measured at paste state

Sensor parameters

Sensor Enovasense Field
Wavelength 980 nm
Power 10 W max
Measurement time 1 s (full field)
Distance 50 mm
Field of view 21 × 15 mm · 384 × 288 px

Layer structure — Application 3

Enamel print Vitroceramic glass ≈ 10 µm

Thin screen-printed enamel on vitroceramic glass · ≈ 10 µm

110 592 pts
Full-field acquisition in a single snapshot
55 µm
Spatial resolution per pixel
Result  High-resolution full-field mapping of a 10 µm print at the paste state — in a single one-second snapshot.

Application 4 — Water-storage tanks (wet slip & fired, on steel)

Glass-lined steel water-storage tanks carry a heavy enamel layer applied as a wet slip and then fired. Controlling this lining is critical: a localised thin zone exposes the steel to corrosion over the life of the tank. The enamel was measured at two states — the wet-applied slip and the fired lining — against two independent reference techniques.

Sensor parameters

Sensor Enovasense Point — T033 head
Wavelength 980 nm
Power 1 W
Measurement time 1 500 ms
Distance 40 mm · spot Ø2.6 mm

Layer structure — Application 4

Enamel (glass lining) Tank steel ≈ 280–710 µm

Enamel glass lining on tank steel · ≈ 280–710 µm

Wet state — compared against a wet-film disc contact gauge, the only conventional option on a fresh slip (and one that disturbs the film it measures):

Figure — Wet state: Enovasense vs wet-film disc gauge

0 0 200 200 400 400 600 600 800 800 1:1 line R² = 0.96 Wet-film disc gauge reference (µm) Enovasense thickness, wet state (µm)

Anonymised dataset · 8 thickness levels, 3 repeats each · R² = 0.96 · repeatability < 2% (≈ 9 µm) · mean deviation ≈ 18 µm

Fired state — compared against eddy current on the fired lining, the sensor tracks the reference closely up to about 1 mm, beyond which the signal saturates:

Figure — Fired state: Enovasense vs eddy current

0 0 250 250 500 500 750 750 1000 1000 1:1 line R² = 0.998 Eddy-current reference, fired state (µm) Enovasense thickness, fired state (µm)

Anonymised dataset · R² = 0.998 · measurable up to ≈ 1 mm, saturating beyond

Result  Control the glass-lining thickness on tank steel both at the wet-applied state and after firing — and catch under- or over-application of the slip before the energy-intensive firing cycle.

06 — Industrial Integration

How to integrate Enovasense sensors on an enamelling line

Depending on whether measurement is needed off-line at quality control, as a high-resolution full-field map, or fully in-line on a powder enamelling conveyor, Enovasense offers three complementary integration routes built on the same patented LPTR sensor.

HKL — off-line control station

The HKL is a compact off-line control station for laboratory and end-of-line quality control. An Enovasense Point sensor is positioned over a series of measurement points across a part — or a batch of parts — to record thickness in a fast, repeatable sequence. It is the natural fit for cookware, tank components, and trays inspected at the quality lab, and benefits from the technology's low sensitivity to part position, curvature, and shape.

Field Sensor — full-field imaging in one snapshot

The Field Sensor is an all-in-one imaging head that captures 384 × 288 — 110 592 measurement points — in a single, roughly one-second acquisition at a 55 µm pixel size, with no moving sensor required. It is ideal for high-resolution thickness mapping and defect detection: resolving screen-printed prints on vitroceramic glass, or mapping overspray across a complete powder-coated panel.

Sensor head dimensions 93 × 85 × 174 mm
Sensor head weight 400 g
Acquisition 384 × 288 — 110 592 points
Pixel size 55 µm
Maximum penetration depth Up to 2 mm
Measurement time ≈ 1 s (full field)

HSR — turnkey in-line station for powder enamel

For automated in-line control on an electrostatic powder enamelling line, the HSR platform is a fully self-contained, turnkey measurement station installed alongside the conveyor — no external controller, power station, or water cooling required. A compact sensor head on a 3-axis motorized gantry, guided by a structured-laser tracking module, automatically detects parts as they move and runs the measurement sequence with no operator intervention, adapting to variable part geometries within a predefined grid.

Sensor head dimensions 272 × 160 × 129 mm (axes excl.)
Sensor head weight 4.3 kg
Measurement range 1–1000 µm
Measurement precision <3% of the measured value
Following / height / distance stroke 2 500 / 2 000 / 300 mm
Maximum line speed Up to 10 m/min
Supply voltage 230 V / 50–60 Hz, 16 A single phase
Laser 980 nm · 1–10 W
Interface Ethernet RJ45 (TCP/IP)
Operating temperature +5 °C to +50 °C
Machine footprint 2 000 × 1 200 mm

Deployment steps (HSR)

1

Site preparation

Client provides a 230 V / 16 A power outlet and an Ethernet network point at the machine footprint location. No compressed air or water cooling required.

2

Design & engineering phase

Enovasense delivers mechanical implantation plans, electrical schematics, safety risk analysis (EN 13849-1), and functional description within 4 weeks of order.

3

Factory acceptance test

Reduced functional test performed at Enovasense premises before delivery to validate conformity of all HSR functions against specification.

4

Installation, commissioning & training

On-site installation and set-up of the full machine, followed by Expert, Operator, and Maintenance training sessions delivered on-site. Full documentation (CE certificate, user manuals, electrical and mechanical drawings) provided.

5

12-month warranty & ongoing support

Full machine warranty for 12 months after provisional acceptance, with 1-week intervention commitment. Optional annual maintenance contract covering hotline, preventive maintenance visit, and software updates.