# Dry film thickness measurement Source: https://coatmentor.com/learn/dft-measurement/ · Updated 2026-10-04 How DFT gauges work, how they are verified, and how SSPC-PA 2 and ISO 19840 sample and accept the readings. *Figure: ISO 19840: mean ≥ nominal, every reading ≥ 80 %, fewer than 20 % of readings between 80 % and 100 %.* Dry film thickness (DFT) is the thickness of the cured coating. It is the number most specifications accept or reject a coat on. For the two sampling rules compared step by step, read the [SSPC-PA 2 vs ISO 19840 guide](https://coatmentor.com/blog/dft-measurement-sspc-pa2-vs-iso-19840). ## How the gauges work Every gauge senses the distance between the probe and the substrate. The coating fills that distance. - **Type 1** (magnetic pull-off): a magnet is drawn to the steel. The thicker the coating, the less force holds it. The gauge turns the pull-off force into thickness. - **Type 2** (electronic, magnetic): the probe senses magnetic flux. The flux depends on the distance to the steel. Use it for non-magnetic coatings on ferrous metal. - **Type 2** (electronic, eddy current): a coil makes an alternating field and induces eddy currents in a non-ferrous metal. Their opposing field depends on the distance to the metal. Use it for non-conductive coatings on non-ferrous metal ([ASTM D7091](https://coatmentor.com/standards/astm-d7091/)). - On concrete, wood or plastic, an ultrasonic gauge (ASTM D6132) needs a couplant under the probe. The substrate affects the reading. Magnetic gauges are sensitive to roughness, curvature and the thickness of the steel. Eddy-current gauges are sensitive to the distance from an edge. Follow the maker's instructions near edges and on small or curved parts. Blasted steel has a special problem. The gauge senses a plane between the peaks and the valleys of the profile. The coating is measured over the peaks, so an uncorrected reading is too high. The standards correct this in different ways, as the next sections show. ## In the field 1. Verify the gauge on reference standards before the work and at the end of each shift. This is a user check. Calibration is done by the maker or a laboratory. 2. Do not verify a Type 1 gauge with plastic shims. If a check fails, treat every reading since the last good check as suspect. 3. On blasted steel, adjust the gauge on a blasted test panel or the bare steel with a shim near the expected thickness. Or subtract a correction value from each reading, if the standard allows it. 4. Take the readings that the standard asks for. Move the probe to a new place for every reading. 5. Compare the results with the specified range and the restriction level or acceptance criteria. 6. If an area fails, find its extent. SSPC-PA 2 asks for spots at about 1.5 m (5 ft) intervals in eight directions until two conforming spots in a row. Mark the failed area and measure again after repair. A destructive method also exists. ASTM D4138 reads the DFT under a microscope in a small angled cut made in place, and the film then needs repair. ## Two acceptance systems **SSPC-PA 2** judges spots (the average of 3 or more readings in a 4 cm circle) and areas (5 spots). **ISO 19840** judges individual readings corrected for profile. The mean reaches the nominal DFT. Every reading is at least 80 % of nominal. Fewer than 20 % fall between 80 % and 100 %. The ISO correction value is 10 µm for a fine profile, 25 µm for a medium one and 40 µm for a coarse one. It is 25 µm when the profile is unknown. Subtract it once from each reading, whatever the number of coats. ISO 19840 applies when the nominal DFT is 40 µm or more. The [comparison guide](https://coatmentor.com/blog/dft-measurement-sspc-pa2-vs-iso-19840) sets the two systems side by side. See also [SSPC-PA 2](https://coatmentor.com/standards/sspc-pa-2/), [ISO 19840](https://coatmentor.com/standards/iso-19840/) and the [DFT acceptance calculator](https://coatmentor.com/calculators/dft-acceptance/). ## A worked example An area of 8 m² has a nominal DFT of 200 µm over a medium profile. ISO 19840 asks for 15 readings in an area above 3 and up to 10 m². A raw reading of 238 µm becomes 238 − 25 = 213 µm. The 15 corrected readings have a mean of 212 µm and a lowest value of 171 µm. Two readings lie between 160 µm and 199 µm. The limit for the 80 % rule is 0.80 × 200 = 160 µm. Fewer than 20 % of 15 readings is fewer than 3, so two is acceptable. The area passes. Check: 20 % of 15 is 3. A third reading in the band would make the area fail. Now take SSPC-PA 2 at the default Level 3 with a specified range of 150 to 250 µm. A spot must lie between 0.80 × 150 = 120 µm and 1.20 × 250 = 300 µm. A spot with readings 132, 128 and 136 µm averages 132 µm and passes. If the five spots of an area average 140 µm, the area fails. The area average must stay in the 150 to 250 µm range. ## What the inspector records - The standard, the restriction level and the specified range. - The gauge type and number, the reference standards and each verification result. - The adjustment method: shim, base metal reading or correction value. - The spot or reading values, the area averages, the lowest and highest values. - Failed areas, repairs and the re-measurement. ## Exam traps - Trap: adjusting the gauge on a smooth surface and then measuring on blasted steel with no correction. - Trap: subtracting the correction value again after the gauge was adjusted on the blasted steel. - Trap: applying the 80 % spot rule at every restriction level. It applies at Levels 3 to 5. - Trap: restricting single gauge readings under SSPC-PA 2. Only spots and areas are judged. - Trap: counting a reading below 80 % in the ISO 20 % band. Any reading below 80 % of nominal fails the area, after the allowed repeats. - Trap: using a magnetic gauge on aluminium or concrete. ## How the exams treat it The CIP 1 guide names SSPC-PA 2 and magnetic DFT gauges. Its practical has a Type 1 gauge station (10 points) and a Type 2 probe station (20 points). The CIP 2 guide names eddy-current and ultrasonic DFT. Its practical has an eddy-current station (20 points), a Paint Inspection Gauge station (15) and a thickness-over-non-metallic-substrate station (10). The CIP 1 guide does not name ISO 19840. **Use it:** [DFT acceptance calculator](https://coatmentor.com/calculators/dft-acceptance/) · [SSPC-PA 2 vs ISO 19840](https://coatmentor.com/blog/dft-measurement-sspc-pa2-vs-iso-19840) **247** practice items in the members area mention this topic (CIP 1: 103 · CIP 2: 84 · FROSIO written: 60). ## Related - **Standards:** [ISO 19840](https://coatmentor.com/standards/iso-19840/) · [ISO 2808](https://coatmentor.com/standards/iso-2808/) · [ASTM D7091](https://coatmentor.com/standards/astm-d7091/) · [SSPC-PA 2](https://coatmentor.com/standards/sspc-pa-2/) · [ASTM D6132](https://coatmentor.com/standards/astm-d6132/) · [SSPC-PA 9](https://coatmentor.com/standards/sspc-pa-9/) · [ASTM D4138](https://coatmentor.com/standards/astm-d4138/) - **Defects:** [Pinpoint rusting](https://coatmentor.com/defects/pinpoint-rusting/) · [Cracking](https://coatmentor.com/defects/cracking/) · [Solvent entrapment and popping](https://coatmentor.com/defects/solvent-entrapment/) - **Inspection and testing:** [Holiday (discontinuity) testing](https://coatmentor.com/learn/holiday-testing/) · [Adhesion testing](https://coatmentor.com/learn/adhesion-testing/) ## About this page **Sources:** SSPC-PA 2 (2022) · ISO 19840:2012 · ASTM D7091-22 · ASTM D6132-13(2022) · ASTM D4138-07a(2022) · DeFelsko, Coating thickness measurement and Calibration terms for coating thickness gages. **Also published openly:** [DeFelsko: SSPC-PA 2 restriction levels](https://www.defelsko.com/pa2) · [DeFelsko: ISO 19840 sampling plan and acceptance](https://www.defelsko.com/coating-thickness/positector-6000-iso-19840). Written from the published standards and coating practice, in our own words, and checked on 2026-10-04. Found an error? It goes in the [corrections log](https://coatmentor.com/method/). CoatMentor is an independent study aid. It is not affiliated with ISO, AMPP, SSPC, NACE International, ASTM or FROSIO.