Measuring DFT in a lab on a flat test panel is straightforward. Measuring it reliably on a real structure — with rough surface profiles, complex geometry, magnetic variation across the substrate, and time pressure from the application crew — takes a bit more care.
This is a practical walkthrough of how DFT measurement actually works in the field, what the numbers mean, and what to do when readings don’t look right.
The Equipment: Magnetic Induction Gauge
For coatings on steel — which covers the vast majority of industrial coating inspection — the standard instrument is a magnetic induction gauge. It works by generating a magnetic field from the probe tip and measuring how that field is affected by the steel substrate beneath the coating. The greater the distance between probe and steel (i.e. the thicker the coating), the weaker the interaction — and the instrument converts that into a DFT reading.
The instruments are handheld, usually battery-powered, and read in micrometres (µm) or mils. Most modern ones are digital with a display. They’re not expensive — a basic calibrated gauge suitable for site use costs a few hundred dollars; a top-end model with data logging and Bluetooth output costs more.
Two important limitations: magnetic induction only works on ferromagnetic substrates (steel, cast iron). On aluminium, copper, or concrete, you need a different measurement principle (eddy current or ultrasonic). And the readings are affected by the substrate’s magnetic properties — which can vary across a steel structure, particularly near welds.
For a broader explanation of what DFT means and why both minimum and maximum values are specified, see what is DFT in coating.
Calibration — the Step Most People Underdo
A DFT gauge needs to be calibrated against the actual substrate you’re measuring on, not just zeroed in a factory. The reason: the magnetic properties of steel vary by grade, thickness, and heat treatment. A gauge calibrated on mild steel plate may give inaccurate readings on a high-yield steel beam.
Calibration procedure:
- Clean an area of uncoated substrate — or measure on an uncoated area nearby that’s the same steel
- Zero the gauge on the bare steel surface (most instruments have a ‘zero’ function)
- Apply calibration foils of known thickness to the zeroed substrate and verify the gauge reads correctly
- If readings are off by more than the manufacturer’s stated accuracy (typically ±2–3%), adjust or service the instrument
Re-calibrate when you move to a significantly different area of the structure — particularly near welds (heat affected zone changes magnetic properties), and on sections with very different geometry from your calibration point. A gauge zeroed on a flat plate can give readings 10–15% off on a curved pipe section. This is not a theoretical concern — it happens in practice.
The SSPC-PA 2 Measurement Protocol
Most industrial coating specifications in North America and internationally reference SSPC-PA 2 for DFT inspection. It’s the standard that defines how many readings to take, how to interpret the results, and what to do about low readings.
The core approach: the structure is divided into measurement areas (typically defined by area in square metres or square feet). Within each area, a defined number of ‘spot measurements’ are taken, each spot being an average of three individual gauge readings within a small radius.
| Reading Type | Definition | Acceptance (typical) |
| Individual reading | Single gauge measurement | No individual reading below 80% of specified minimum DFT |
| Spot measurement | Average of 3 individual readings within ~40mm radius | Each spot ≥ minimum specified DFT |
| Area average | Average of all spots in the measurement area | Should meet or exceed specified DFT |
When readings are below minimum: SSPC-PA 2 distinguishes between areas with a few low readings (requiring additional application and re-measurement) and widespread low DFT (which implies a systematic problem with the application). The protocol tells you when spot repair is appropriate and when a broader investigation is needed.
Where to Measure — and Where Not to Forget
New or inexperienced inspectors tend to measure on flat, accessible surfaces and miss the areas where DFT is most likely to be out of specification. In practice, pay extra attention to:
- Edges and corners: coating naturally pulls away from sharp edges during application and curing; DFT at edges is typically lower than on flat faces
- Welds and weld toes: irregular surface profile makes uniform coating application difficult; these are often both thin and rough
- Bolt holes, nuts, and washers: recesses tend to pool coating (thick) or have very limited access (thin) depending on application method
- Overhead surfaces: coating sags on downward-facing surfaces if applied too thickly; DFT is often inconsistent
- Transitions between sections: where web meets flange, where members connect — geometry changes create application challenges
The stripe coat on edges and welds exists precisely because uniform DFT at these locations is difficult with spray application alone. Even with stripe coating, these areas warrant close inspection.
Reading Interpretation: What the Numbers Are Telling You
A few scenarios and what they mean in practice:
Consistently low across the whole area: the spray settings, application speed, or material mix ratio are probably off. This needs to be investigated and corrected before applying the next coat — applying more over a consistently thin coat without finding the cause just masks the problem.
Isolated low readings in an otherwise acceptable area: likely thin spots from spray technique — missed areas, insufficient overlap at pass edges. Typically addressed by stripe-coat touch-up and re-measurement.
Consistently high in corners and low on flat faces: indicates pooling in corners from gravity or over-spray. For some coating types (intumescent, glass flake) this is a problem in both directions — too thick in corners can cause cracking, too thin on faces leaves gaps.
Very high variability (some readings far above, some far below): usually indicates inconsistent application technique or equipment problems. Worth stopping application, investigating, and correcting before continuing.
Practical Tips That Don’t Always Make It Into the Standards
Keep the probe perpendicular to the surface. Tilting it even slightly introduces error — and on curved surfaces it’s easy to do without noticing.
Let the gauge stabilise for a few seconds after placing the probe, especially in cold weather when the instrument may be coming from a warm vehicle.
On rough blast-cleaned profiles, the gauge reads from the average surface level — not from the peaks or valleys of the profile. If the blast profile is 80 µm Rz and you’re measuring a coating specified at 250 µm, your readings will be somewhat affected by the profile roughness. This is normal and is why specifications are written with this in mind.
If the gauge gives wildly inconsistent readings in a small area — jumping from 150 µm to 400 µm over a span of 10cm on what looks like a uniform surface — check for buried steel features (bolts, reinforcement, plate edges) that might be distorting the magnetic field.
💡 Document everything. DFT readings are an inspection record — date, location reference, individual readings, gauge serial number, calibration date. On a project where coating performance is later questioned, this documentation is the evidence. A verbal ‘it looked fine’ is not.
Quick Questions
Can I measure DFT through multiple coats at once?
Yes — the gauge reads total film thickness above the steel regardless of how many coats are present. To determine individual coat DFTs, measure after each coat is applied. Once a subsequent coat is over the top, you can only back-calculate the individual coat thickness if you recorded the previous total. This is why hold-point inspections are done after each coat, not just at the end.
My gauge gives different readings on different parts of the same steel plate. Is it broken?
Probably not. Variation in magnetic permeability across a steel plate — from rolling direction, heat treatment variation, or proximity to welds — causes genuine variation in gauge readings. This is why calibration is done as close as possible to the actual measurement location, not just once at the start of the day. Also check that there are no steel features on the back face of the plate that might be influencing the readings from the front.
Is there a minimum DFT that’s too thin to measure reliably?
Most magnetic induction gauges lose accuracy at very low DFT — below about 10–15 µm, depending on the instrument. For most industrial coating applications this isn’t a practical constraint since even primer coats are specified at 50 µm or above. For very thin coatings (conversion coatings, pretreatment films), specialist instruments or destructive cross-section examination is used instead.
Related Reading
- Related Reading
- What is DFT in coating — dry film thickness explained, including minimum and maximum values and what happens when DFT is wrong.
- Steel structure coating inspection checklist — surface prep, DFT hold points, recoat intervals, and touch-up procedures.
- How to calculate intumescent coating thickness — section factor, fire curve, and DFT table lookup for fireproof coating specification.
- Surface preparation for industrial coatings — engineering standards and practical selection guide for blast cleaning and surface profile.
Send your project coating specification, steel section details, and inspection requirements via the project inquiry form and our technical team will advise on DFT specification and system selection.
