Which Layer Did the Gauge Measure? Publishing Useful Coating Thickness Evidence
By Daoyu Guan
A coating-thickness result identifies a measurement of a particular coating system on a particular substrate, using a suitable method at defined locations. It is not automatically the thickness of the top layer, a minimum over the whole component or evidence of corrosion life. Before comparing suppliers, identify the layer stack, gauge principle, substrate, preparation and acceptance rule. Keep those fields beside the readings. Otherwise, two genuine measurements can be presented as equivalent even though the instruments responded to different parts of the coating system.
A digital display does not identify the layer
Suppose a buyer requests the thickness of an organic finish on a plated steel component. The supplier sends a photograph of a gauge reading. The number looks precise, but the photograph does not establish whether the response represents the organic layer, the combined nonmagnetic layers above the steel or a different measurement arrangement.
The buyer needs the stack before interpreting the reading: substrate, intermediate metallic coating, primer, topcoat and any other relevant layer. A result described simply as “coating thickness” can conceal a different boundary from the one in the purchase requirement.
This is not solved by renaming the column “paint thickness.” The method and configuration must support the intended measurement. If a coating or substrate has changed, the earlier setup may need re-evaluation even when the required nominal thickness remains the same.
Begin with the physical measurement principle
Magnetic methods are used for suitable nonmagnetic coatings over ferromagnetic substrates. Eddy-current methods can be used for suitable nonconductive coatings over conductive, nonferrous substrates. Other systems require other methods or an appropriately demonstrated configuration. “Metal part” is not a sufficient method-selection description.
DeFelsko's coating-thickness measurement guide explains these principles and identifies effects such as curvature, roughness and substrate properties. A manufacturer's technical guide helps explain the measurement; it does not independently validate a supplier's own results.
For a multilayer finish, a single displayed number may not separate every layer. The PosiTector 6000 manual discusses particular configurations for plated substrates, illustrating why the instrument, probe and mode need identification. Do not generalize one model's capability to every gauge using a similar principle.
| Question | Why it changes the interpretation |
|---|---|
| What is the substrate? | The method's response depends on its relevant physical properties |
| Which layers sit above it? | A response may include more than the layer named in the order |
| Which probe and mode were used? | The selected configuration determines the intended measurement |
| Where was the probe placed? | Edges, curvature and access can affect applicability |
| What check or adjustment was performed? | The setup needs evidence appropriate to the actual part |
| What rule decides acceptance? | A mean, a local reading and a minimum are different quantities |
A simple stack shows the ambiguity
Consider a fictional idealized stack: 8 µm of a nonmagnetic metallic layer beneath 60 µm of an organic coating on steel. The arithmetic sum above the steel is 68 µm. That arithmetic alone does not predict a real gauge's exact displayed value, because the instrument and part effects still matter.
It does show why a 68 µm result cannot automatically be relabelled “68 µm of organic topcoat.” A magnetic method responding to the combined separation from the steel is answering a different question from one that isolates the organic layer.
Nor should the buyer simply subtract a nominal 8 µm from every result. The underlayer's actual thickness, measurement location, variation and uncertainty would need an appropriate basis. Subtracting two measurements introduces its own uncertainty; subtracting an unverified nominal value may be worse.
A useful report states which boundary the measurement represents. If layer separation requires another technique or a qualified configuration, agree it before production. An English product page should not promise layer-specific evidence that the supplier's routine measurement does not produce.
Calibration, verification and adjustment are different activities
A calibration record describes an instrument's response under specified conditions. A verification checks its performance against a suitable reference. An adjustment changes the instrument's response for the intended use. The exact terminology and required practice should follow the applicable method and instrument documentation.
DeFelsko's calibration-terms guide distinguishes these activities. Its reference-standard descriptions also show why a certified reference and an unverified practice shim are not interchangeable evidence.
For a buyer, the practical question is whether the setup is appropriate to the actual component. A calibration certificate for a gauge does not, alone, establish that a probe can reach a narrow recess or that a flat reference represents a sharply curved surface.
Ask the supplier to identify the reference used, the relevant check results and any part-specific adjustment. Preserve the record before and after the production examination where the agreed procedure requires it. Do not rewrite a failed check as a pass by showing only a later adjusted reading.
Decide whether the requirement concerns local thickness or an average
Three readings can support several different summaries. The correct one depends on the agreed specification, not on which makes the shipment look acceptable.
Here is an invented example with three designated locations. Assume, for this example only, that the order requires each designated location to meet a 25 µm lower limit under an agreed decision rule. No actual product or order is represented.
| Designated location | Illustrative reading, µm |
|---|---|
| A | 23 |
| B | 28 |
| C | 31 |
| Arithmetic mean | 27.33 |
The mean exceeds 25 µm, but location A does not meet the example's stated point requirement on the displayed value. Reporting only the mean would not answer that requirement. In a real order, measurement uncertainty, allowable local exceptions and any defined spot or area averaging must follow the actual method and contract.
Reverse the situation and the question changes again. A specification based on an agreed area average should not be silently replaced with a requirement that every individual reading exceed the average limit. Define the measured and accepted quantities before collecting data.
The report also needs a map. “Point A” is not useful if a later inspector cannot locate it on the correct product variant. An approved sketch or region description can preserve the relationship without exposing a confidential customer drawing publicly.
Separate the witness coupon from the shipped component
A flat witness panel can provide useful process evidence. It does not automatically establish thickness inside every recess or on every edge of the delivered part. The relationship between the panel and the part needs an agreed technical basis.
Identify whether a published result comes from a production component, witness panel, development sample or destructive section. Keep the batch, process state and preparation attached to it. A supplier capability page may describe available measurement methods; a lot-release report must identify what was actually examined.
This is especially important when marketing reuses an old laboratory image. The image can illustrate an instrument or technique. Unless the records establish more, it is not evidence that the current shipment received that examination.
Build a coating evidence request the supplier can answer
- Name the stack. Identify the substrate and layers relevant to the requirement, including whether the finish is measured in its specified final condition.
- Define the quantity. State whether the buyer needs a particular layer, total thickness, local results or a specified aggregate.
- Choose an appropriate method. Have the responsible technical parties confirm its suitability for the materials, geometry and required range.
- Define locations and coverage. Map accessible regions, exclusions and any supplementary method needed for inaccessible areas.
- Specify the setup evidence. Identify the instrument, probe, relevant mode, reference and checks required by the agreed procedure.
- Agree acceptance and disposition. Record how results, uncertainty and any exceptions will be evaluated.
- Preserve the source report. Keep the original readings and location map alongside the summary used in the English page or buyer response.
This sequence is a communication framework, not a universal coating standard. A sector-specific specification may impose different or additional requirements.
Thickness does not establish every property of the finish
A coating can meet a thickness requirement while still requiring separate evidence for other characteristics. Thickness alone does not establish adhesion, composition, cure, coverage of unmeasured regions or service performance. Which other properties matter is an engineering and purchasing decision for the actual application.
Likewise, a laboratory corrosion exposure is a separate result. The salt-spray claim guide explains why an exposure duration should not become an unsupported number of years outdoors. Combining a thickness figure with that duration does not automatically create a service-life prediction.
For Xindar's manufacturing content work, the useful output is a clear comparison record: what was measured, how, where and for which claim. The manufacturer and authorized test provider approve the technical facts. An AI-readable table should preserve those facts, rather than compress them into “premium coating, fully verified.”
Buyer questions
Does one thickness reading identify the topcoat thickness?
Not necessarily. Confirm the stack, instrument principle and configuration. A method may respond to combined layers.
Can I subtract a nominal underlayer thickness?
Do not treat that subtraction as a verified layer measurement without a suitable basis. Actual underlayer variation and measurement uncertainty matter.
Does a calibrated gauge suit every part geometry?
No. Confirm the applicable range, probe access, substrate and geometry conditions for the actual measurement.
Does a passing average prove every location passed?
No. Apply the specification's defined quantity and decision rule. The fictional three-point example demonstrates why those rules cannot be exchanged silently.
What belongs next to a coating-thickness claim online?
Identify the coating system, relevant substrate, method, scope and status of the evidence. Keep an individual test result separate from a general capability or order guarantee.
Sources and scope
- DeFelsko: coating-thickness measurement — method principles and part effects.
- DeFelsko: calibration terms and reference standards — instrument checks and references.
- PosiTector 6000 manual — specific instrument behaviour, not a claim about all gauges.
The stack and readings are fictional. No coating system, supplier inspection or service-life claim has been qualified by this article.