Plated metal brackets on a workbench
Knowledge Intermediate

Plating Thickness Measurement Methods: Which Test to Use

November 7, 2026 24 min read Lab Wizard Development Team
Choose a thickness test for plated and anodized parts: magnetic, eddy current, XRF, coulometric or cross section, plus the setup errors that skew readings.

The right plating thickness test depends on four things: what the coating and substrate are made of, how thick the coating is, whether you can accept a bare spot or a destroyed part, and the shape and size of the area you need to measure. If the plating specification or purchase order names a method, that method governs. Otherwise, start here:

  • Nonmagnetic coatings on steel, such as zinc, cadmium, tin or copper (not nickel): a magnetic gauge.
  • Nonconductive coatings on nonmagnetic metals, such as anodize on aluminum: a standard (amplitude-sensitive) eddy current gauge.
  • Thin metallic coatings on many substrates, including precious metal layers and some multilayer stacks: X-ray fluorescence (XRF), up to a maximum thickness that depends on the coating.
  • Layers a gauge cannot separate, locations it cannot reach, or a disputed reading: coulometric stripping or a microscopical cross section, both destructive. In a dispute, use the referee method the specification or purchase order names.

Each method family reports a different quantity and has its own calibration rules. A correctly chosen method still reads wrong when its setup does not match the part. For magnetic and eddy current gauges, that means base metal thickness, curvature, edges, roughness and extra layers above the coating; for XRF, it means spot size, focus distance, substrate and standards.

If you need to predict thickness from rectifier charge, start with Electroplating Thickness Calculations (Faraday’s Law Made Simple). If your calculation and your gauge disagree, Why Calculated and Measured Plating Thickness Differ explains how to compare them. This article is about choosing and setting up the measurement itself.

🧲 How to Choose a Method by Coating and Substrate

Most method choices can be made from two questions. Is the substrate magnetic? Is the coating magnetic, electrically conductive, or neither?

Thickness Method Selection by Coating and Substrate

Starting points based on the scope of each ASTM or ISO method. ND = nondestructive, D = destructive. Where an XRF thickness range is stated, it is a typical figure from an instrument maker or platers' experience and varies with the instrument, substrate, stack and calibration. Multilayer XRF applies only within instrument- and stack-specific validated limits. The method named in your plating specification or purchase order takes precedence.

Coating on substrateFirst choiceOther optionsWatch for
Zinc, cadmium, tin or copper on steelMagnetic, ASTM B499 / ISO 2178 (ND)XRF (ND); phase-sensitive eddy current, ISO 21968 (ND)Passivate, sealer and topcoat are read as coating
Hard chromium plated directly on steelMagnetic, ASTM B499 / ISO 2178 (ND)XRF within its range (ND); cross section (D)Thick deposits can exceed the XRF range
Electroplated nickel on steelXRF (ND), typically up to about 30 µm of nickelPhase-sensitive eddy current, or the ASTM B530 magnetic method for nickel, each calibrated on nickel-plated parts from the same process (ND); coulometric, ASTM B504 (D); cross section (D)Never B499; ordinary magnetic gauges cannot separate nickel from a steel base
Electroplated nickel on copper, brass or aluminumMagnetic method for nickel, ASTM B530 (ND)XRF (ND); coulometric (D); cross section (D)Calibration standards plated in the same nickel process
Electroless nickel on steel, as platedXRF within its validated measurement range (ND)Cross section (D); magnetic, ASTM B499 (ND), for sufficiently nonmagnetic as-plated deposits when the governing specification permits it and the method has been validated against an independent measurementMagnetic response depends on phosphorus content; see the nickel section for historical thresholds and limits
Electroless nickel below about 8% phosphorus, or any heat treated electroless nickel, on steelXRF calibrated for the deposit (ND), within the instrument’s validated rangeCross section (D)The deposit responds magnetically; do not treat B499 as the first choice
Electroless nickel on aluminumXRF with electroless nickel on aluminum standards (ND), typically up to about 25 to 30 µmMagnetic nickel gauge for deposits below about 7 to 8% phosphorus by weight, with standards plated in the same bath (ND); cross section (D); eddy current only if calibration on coated samples proves it (ND)Phosphorus content changes both XRF response and density
Anodic coating on aluminumEddy current, ASTM B244 / ISO 2360 (ND)Cross section (D)Oxide thickness is not the same as dimensional buildup
Gold, silver, palladium or tin on copper alloysXRF, ASTM B568 / ISO 3497 (ND)Beta backscatter (ND); coulometric (D); cross section (D)Spot size versus contact or feature size
Decorative chromium over nickel over copper, on steel or plasticXRF with a multilayer calibration for the stack’s layer order and elements, within instrument- and stack-specific validated limits (ND)Coulometric layer by layer, ASTM B504 / ISO 2177 (D); cross section (D)B499 does not apply because of the nickel; XRF capability depends on each layer, the stack sequence, the substrate, the instrument configuration and the calibration; stacks outside the validated model go to coulometric testing or a cross section, and the semibright and bright nickel split needs the STEP test or an etched cross section
Multilayer nickel (semibright plus bright)STEP test, ASTM B764 (D)Cross section (D)The only listed method that also reports the potential differences between layers
Chromate or other chemical conversion coatingsNo routine thickness gaugeCoating weight or the test named in the coating’s own specificationThese films are too thin for an eddy current gauge

Nonmagnetic metals on steel, such as zinc, cadmium, copper and tin, do not respond to a magnetic field, so a magnetic gauge reads them as a gap between probe and steel; ASTM B499 and ISO 2178 cover this case. Chromium reads the same way only when it is plated directly on steel, as hard chromium usually is; decorative chromium sits on nickel, which takes it out of B499’s scope. Phase-sensitive eddy current (ISO 21968) is the alternative for these coatings on steel. Instrument makers describe it as much less affected by curvature and roughness than magnetic or amplitude-sensitive eddy current gauges, which suits small plated parts. Its measuring range depends on the probe and on the coating and base, and the base must be thick enough for the probe. Because the reading depends on the electrical and magnetic properties of both, it needs calibration on the same coating and base combination as the parts.

Nonmagnetic metal coatings on nonmagnetic metals, such as tin on copper or the gold in gold over nickel on brass, cannot be read with a magnetic gauge, and amplitude-sensitive eddy current is meant for nonconductive coatings. The nickel layer under that gold is the exception: the ASTM B530 magnetic method for nickel can read it. Use XRF for these coatings: it is noncontact and nondestructive, and it measures many coating and substrate combinations that other techniques do not readily measure, within the thickness limits covered below. Phase-sensitive eddy current works for a metal on a metal only when the coating and base differ enough in electrical conductivity or magnetic permeability. Electroless nickel on aluminum can meet that condition, but rely on the method there only after calibration on coated samples of the same deposit confirms it.

Anodize on aluminum, a nonconductive oxide on a conductive, nonmagnetic metal, is the main use of ASTM B244. Aluminum alloys and tempers differ in conductivity, so the calibration standards need a base metal with the same electrical properties as the part. The gauge reports the total oxide thickness, but anodizing converts part of the aluminum surface into oxide, so the coating grows both into and out of the original surface. For hard anodize, NASA’s process specification PRC-5006 says that typically half of the coating is penetration and half is growth; the share varies with the process and alloy, so check it on your own parts when a dimension is tight. A micrometer measures the finished dimension, so the change it shows is the growth less any aluminum that etching or other pretreatment removed, and less any lapping or honing afterward; across a diameter it covers both sides. Compare the micrometer with a dimension “after anodize”, or halve the change in diameter before comparing it with a buildup per side, and compare the eddy current reading with a coating thickness requirement.

🔬 How the Methods Compare: What Each One Measures

Some methods infer a distance from a magnetic or electrical field, some respond to the mass of coating per unit area (XRF, beta backscatter, and coulometric stripping, which dissolves the coating), and a cross section cuts the part so the layer can be measured directly. That difference decides how a reading should be interpreted and what can bias it.

Common Coating Thickness Methods and What They Report

Summary of the main method families, with the ASTM and ISO test methods that describe them.

MethodMain standardsWhat it reportsEffect on the part
Magnetic induction or magnetic pull-offASTM B499, ISO 2178Distance from probe to magnetic base metal, read as coating thickness through calibrationNondestructive
Magnetic method for nickel coatingsASTM B530, ISO 2361Nickel coating thickness from the magnetic response of the coating and substrate togetherNondestructive
Amplitude-sensitive eddy currentASTM B244, ISO 2360Distance from probe to a conductive, nonmagnetic base metal, read as thickness of a nonconductive coatingNondestructive
Phase-sensitive eddy currentISO 21968Thickness of a nonmagnetic metallic coating, from its electrical properties relative to the baseNondestructive
X-ray fluorescence (XRF)ASTM B568, ISO 3497Coating mass per unit area, estimated from X-ray intensities through a calibration or model and converted to thickness with a density valueNondestructive, no contact
Beta backscatterASTM B567, ISO 3543Coating mass per unit area, converted to thickness with a density valueNondestructive
Coulometric strippingASTM B504, ISO 2177Charge needed to dissolve the coating from a small defined area, converted to thickness through the metal’s electrochemical equivalent and densityDestroys the coating at the test spot
Optical cross sectionASTM B487, ISO 1463Local thickness measured directly on a polished sectionDestroys the part or a test coupon
SEM cross sectionASTM B748Local thickness of thin coatings and individual layers, measured with a scanning electron microscopeDestroys the part or a test coupon

XRF, beta backscatter and coulometric results depend on density. Beta backscatter measures mass of coating per unit area, and XRF estimates it from X-ray intensities through a calibration or model that also assumes the coating’s composition. Coulometric stripping measures the mass of metal dissolved. Each becomes thickness only through a density. A density set too high for the deposit makes the coating read thin, and one set too low makes it read thick, by about the same proportion, even when the instrument works perfectly. Alloy coatings are where this matters: nickel-phosphorus is less dense than pure nickel, and its density changes with phosphorus content.

Magnetic and amplitude-sensitive eddy current gauges measure everything between the probe and the metal they sense. They sense the base metal, and the coating is the gap between the probe and that base metal. So anything that changes the base metal’s response, such as its alloy, thickness, curvature or a nearby edge, changes the reading even when the coating is identical. Every nonmetallic layer in the gap is also counted: a chromate or trivalent passivate, a sealer, a topcoat, oil or dirt. Metal layers depend on the gauge. A magnetic gauge on steel also counts a nonmagnetic metal layer, such as a copper strike under the final plate, but reads a magnetic layer, such as a nickel strike, as part of the steel. An amplitude-sensitive eddy current gauge reads a metal layer as part of the base metal, completely when the layer is thick and only partly when it is thin. A thin metal layer under the coating therefore makes its reading unreliable, and a thick one needs a calibration made on that metal.

A zinc reading taken after sealer and topcoat reports the whole stack. Measure before those layers go on, or record which layers the reading includes. When only the zinc matters, use XRF calibrated with standards of the same coating combination, or a phase-sensitive eddy current gauge with a probe made to read zinc under paint, calibrated on an uncoated part and a zinc-plated part like the production parts. Instrument makers note that this works only when there is no zinc-iron diffusion layer, and electroplated zinc usually has none.

✅ A Practical Selection Sequence

For a new part or coating, work in this order:

  1. Read the specification first. If it names a method, a referee method, measurement locations or a sampling plan, those govern.
  2. Classify the coating and substrate. Decide whether each is magnetic, conductive, or neither, and whether the coating is a single metal, an alloy, or a stack of layers that a gauge will read together.
  3. Check the thickness range. Compare the nominal thickness with each candidate method’s range: at the heavy end, XRF’s maximum for that coating; at the thin end, the fixed part of a gauge’s error and the accuracy limit of optical cross sections, both explained under thickness limits below.
  4. Decide whether the part can be marked or destroyed. If not, choose among magnetic, eddy current and XRF. If a sample part or coupon can be sacrificed, coulometric testing and cross sections become available.
  5. Check the geometry against the method. Compare the smallest area you must measure with the probe or XRF spot, and note curvature, base metal thickness, and recesses the gauge cannot reach.
  6. Plan the calibration. Identify uncoated base metal that matches production for magnetic and amplitude-sensitive eddy current gauges, coated standards from the same process for nickel gauges and phase-sensitive eddy current, and certified standards of the same coating and substrate for XRF.
  7. Check capability against the tolerance. A gauge whose measurement variation is a large share of the thickness tolerance cannot reliably sort good parts from bad. Some specifications set the accuracy directly: the canceled 1990 military specification MIL-C-26074E, for example, required nondestructive readings within 10 percent of the true thickness, or an instrument type approved by the purchaser. Run a gauge study before production depends on it: MSA in Plating Labs: Gage R&R for Titration & pH shows the study design on titration and pH, and for a thickness gauge the result is judged against the thickness tolerance.
  8. Verify against a second method. For a new part, compare the production gauge with a cross section or coulometric result on a sample. Either catches gross calibration errors. When the production gauge is XRF or beta backscatter, a wrong density shows up only in a check that measures thickness as a length: a cross section, or a magnetic gauge where one applies. A coulometric result also converts mass to thickness with a density, so it can share the gauge’s error. Repeat the comparison when the substrate, the process or the standards change.

Example: A Zinc Plated, Sealed Steel Bracket

A thin sheet steel bracket is zinc plated, passivated and sealed, and the drawing sets a minimum zinc thickness on its flat face and inside a bend.

  • Method (steps 1 and 2). The order names no method, and zinc on steel points to a B499 magnetic gauge. Take the reading before sealing, with the passivate noted on the record, because the gauge counts every layer above the steel; after sealing, XRF reports the zinc alone.
  • Setup (steps 5 and 6). Calibrate on uncoated steel of the same sheet thickness, since a gauge set on thick steel reads high on thin parts. Inside the bend a flat calibration reads low, so match the curvature there and keep the probe away from the edges, where its readings tend to come out high.
  • Location and check (step 8). The inside of the bend plates thin, so measure it as well as the flat and record which reading came from where. Section one bracket and compare it with the gauge before production relies on it.

📐 Thickness Limits: Which Methods Work on Very Thin or Very Thick Coatings?

Coating thickness can rule a method out even when the materials fit. XRF has an upper limit for each coating: once the coating is thick enough, its X-ray signal stops changing with further thickness, so the instrument can no longer tell thicknesses apart. NIST’s guide to the thickness test methods gives a range of roughly 0.01 µm to about 75 µm at most, depending on the coating and substrate. Thick hard chromium, heavy nickel built up for wear or salvage, and thick zinc can exceed it. Magnetic gauges usually handle thick hard chromium and thick zinc on steel, and an optical cross section is accurate at those thicknesses. Heavy nickel on steel is harder; the nickel section below gives each method’s practical limit.

At the thin end the ranking reverses: flash coatings of a micrometer or less are where XRF is strongest and where an optical cross section struggles. According to the same NIST guide, which summarizes the optical cross-section standards, an optical cross section gives thickness to within 1 µm or 10%, whichever is greater, and can reach an absolute accuracy of about 0.8 µm under good conditions. That is a large fraction of a thin coating, so thin layers are measured on sections in a scanning electron microscope, the subject of ASTM B748. Magnetic and eddy current gauges also lose relative accuracy on thin coatings, because part of their error is fixed. DeFelsko, for example, specifies ±(1 µm + 1%) for readings up to 50 µm on its standard probes, and on a coating only a few micrometers thick that tolerance is a large share of the reading.

⚠️ How Do You Measure Nickel and Electroless Nickel Thickness?

Nickel is ferromagnetic, so a B499 gauge on nickel-plated steel cannot tell where the nickel ends and the steel begins. That puts electrodeposited nickel on steel outside ASTM B499, which covers nonmagnetic coatings on magnetic base metals; ASTM B530 is the magnetic method written for electrodeposited nickel coatings.

B530 covers electrodeposited nickel on both magnetic and nonmagnetic substrates and handles it by calibrating against nickel itself. It works only when the reference standards match the magnetic properties of the coating and substrate. Plated nickel’s magnetic properties change with the plating conditions, so that means standards plated in the same process, on the same substrate, as production. Instrument makers are more cautious about nickel on steel than the standard’s scope suggests. Fischer, for example, gives roughly 150 µm as a practical magnetic range for nickel on nonmagnetic bases and says magnetic methods cannot measure nickel on steel. On steel it measures nickel with phase-sensitive eddy current (to roughly 100 µm) or coulometric stripping (to roughly 50 µm), and it notes that XRF becomes difficult above roughly 30 µm of nickel.

Electroless nickel adds a second variable, phosphorus content, and published sources draw the magnetic line in different places. ASTM B499 Note 1 and the canceled 1990 military specification MIL-C-26074E both treat electroless nickel above 8% phosphorus by weight, as plated on a magnetic base metal, as nonmagnetic enough for the B499 magnetic method, and they do not recommend that method below 8%. Published research puts the point where the deposit is reliably nonmagnetic higher, at about 10% phosphorus, the start of the high phosphorus grades, and plating suppliers often quote about 11%. Deposits between 8% and about 11% phosphorus fall between those lines; that band covers the upper end of mid phosphorus grades and the lower end of high phosphorus grades. How Phosphorus Content Shapes the Hardness of Electroless Nickel gives the low, medium and high phosphorus ranges.

For production measurement, prefer XRF within its validated range for the deposit and substrate, or a cross section. Magnetic measurement under ASTM B499 may still be appropriate for a sufficiently nonmagnetic as-plated deposit when the governing specification permits it and the method has been validated against an independent measurement such as XRF or a cross section on parts from the same bath. Heat treatment makes the deposit more magnetic, so an as plated calibration is no longer valid after the bake; where heat treated deposits are measured magnetically, MIL-C-26074E advised calibrating against heat treated standards. Electroless Nickel Heat Treatment: Hardness, Temperature, and Time covers why the structure changes.

For low phosphorus or heat treated deposits, and for electroless nickel on aluminum, the magnetic response does not matter to XRF or a cross section. XRF suits these deposits when the instrument supports that coating and substrate and the thickness is within its calibrated range. When XRF is used, calibrate with electroless nickel standards of similar phosphorus content on the same substrate, because phosphorus changes both the X-ray response and the density.

✂️ When a Destructive Method Is the Better Answer

Nondestructive gauges are fast, but some questions need a method that removes or cuts the coating.

Coulometric Stripping

The coulometric method dissolves the coating anodically from a small area sealed by a gasket, using an electrolyte chosen for that coating and substrate. When the coating is gone, the cell voltage jumps as a different metal is exposed, and the charge passed up to that point gives the thickness through the metal’s electrochemical equivalent and density. ASTM B504 notes that it destroys the coating over about 0.1 cm², so it suits parts where a small bare spot is acceptable, or test pieces that can be scrapped.

Its strength is layered coatings. A coulometric instrument can strip chromium, then nickel, then copper and report each layer, when exposing each new metal produces a clear voltage jump. It needs an electrolyte suited to each layer and the metal beneath it, changed between layers. The STEP test (ASTM B764) is a modification for multilayer nickel that reports each nickel layer’s thickness and the potential differences between layers that decide how a duplex or triplex nickel system protects the substrate. The conversion uses the electrochemical equivalent and density of the deposit being stripped and assumes a current efficiency close to 100%, so calibrate or check the instrument on thickness standards of the same coating and substrate. An alloy deposit needs an electrolyte and calibration suited to that alloy. After each test, check that the coating dissolved cleanly to the substrate. A leak at the gasket enlarges the stripped area and makes the result read high, and an unsuitable electrolyte, or a diffusion layer between coating and substrate, as under tin on copper, can stop the test early or late.

Microscopical Cross Section

A cross section measures local thickness directly. The part is cut, mounted, ground and polished, and the coating is measured under a microscope. ASTM B487 covers optical sections and ASTM B748 sections measured in a scanning electron microscope. Both depend on preparation, and NIST’s summary of ISO 1463 notes that careful preparation and suitable instruments can bring the optical uncertainty down from about 0.8 µm to about 0.4 µm under reproducible conditions.

A cross section shows things no gauge can show: every layer in a stack, thickness inside a narrow recess or bore, the coating at an edge, and interface defects such as porosity, voids, contamination or separation that point to poor adhesion. Adhesion itself is judged by the test the specification names, such as a bend, file or heat-quench test. Preparation decides whether the number is right. In standard metallographic practice, the coated surface is overplated or otherwise supported before mounting so the coating edge does not round off during polishing, and a light etch makes the boundaries between layers visible. A cut that is not perpendicular to the coating makes the layer look thicker than it is, while a rounded or smeared edge can make it look thinner. Either error can hide or invent a thin spot.

Cross sections are also how many shops verify their gauges and settle disputes. No single method is the universal referee: the plating specification or purchase order names it, and MIL-C-26074E, for example, used periodic microscopical cross sections as the referee for its nondestructive readings. Check which method yours names before the first part ships.

📏 Setup Errors That Make a Correct Method Read Wrong

Many thickness disputes involve a method that was right for the coating and wrong for the way it was calibrated or applied.

Magnetic and Eddy Current Gauges

Calibration has to reproduce the part. Instrument makers give the same rule for both technologies: calibrate on an uncoated area of the measuring surface, or on uncoated base metal of the same alloy, thickness and shape, then check on a coated part. Nickel gauges and phase-sensitive eddy current probes are the exception: they need coated standards from the same plating process.

For gauges that sense the base metal through magnetic induction (ASTM B499 / ISO 2178) or amplitude-sensitive eddy current (ASTM B244 / ISO 2360), instrument makers describe typical directions for the common geometry errors. These directions apply when the gauge is calibrated on a flat, thick, smooth piece of matching base metal and then used on a different geometry. Probe design, substrate, calibration and measurement principle can change the size of the error, and near a probe’s minimum base thickness even the direction can reverse, so treat the examples as starting guidance and confirm them for your probe.

  • Base metal thickness. If the base metal is thin, the probe’s field passes through it. For magnetic induction and amplitude-sensitive eddy current, a gauge calibrated on thin base metal typically reads low on thick parts, and one calibrated on thick base metal typically reads high on thin parts. Near the probe maker’s minimum base thickness, the size and even the direction of the error depend on the probe and the metal. Calibrate on base metal of the same thickness.
  • Curvature. On a convex surface such as a shaft or tube, more of the field of a base-metal-sensing probe travels through air than on a flat. A magnetic induction or amplitude-sensitive eddy current gauge calibrated on a flat typically reads high on outside diameters and low on inside diameters. Calibrate on the same curvature, or use a probe designed for it. A nickel gauge under ASTM B530 on a nonmagnetic base senses the coating instead. In 1937 the US National Bureau of Standards (NBS, now NIST) found that a magnetic attraction gauge gave a weaker signal on convex nickel surfaces, so it read low on outside diameters. Calibrate a modern nickel probe on nickel-plated parts of the same curvature.
  • Edges and small areas. Near an edge or on a small face, part of a base-metal-sensing probe’s field misses the base metal and runs in air, as it does over a convex surface, so the reading typically comes out high and can hide a thin coating. NBS’s 1937 attraction gauge read the other way on nickel over a nonmagnetic base: a smaller nickel area gave a weaker signal, so it read low on small areas and near edges. How close to an edge a probe can measure depends on its design: measure away from edges, and check the probe maker’s minimum measuring area before measuring small parts.
  • Roughness. A rough or blasted base surface makes individual readings scatter and, for magnetic and eddy current gauges zeroed on a smooth surface, commonly biases them high. Zero on uncoated base metal with the same finish, take several readings at each location, and report their average along with the location. Fischer considers readings on rough surfaces reliable only when the coating is at least twice as thick as the roughness peaks; that rule is instrument-maker guidance, not a universal standard limit.
  • Probe handling and soft coatings. On electronic magnetic and eddy current gauges with contact probes, a tilted probe lengthens the gap and typically reads high, a probe tip that presses into a soft coating such as tin shortens it and typically reads low, and uneven pressure adds scatter. Hold the probe square, let it settle, and keep the tip and the part clean. Constant-pressure probe sleeves reduce operator pressure effects when used as designed.

Calibration foils, thin plastic shims of known thickness laid on uncoated base metal, are convenient for setting electronic magnetic and eddy current gauges, though not magnetic pull-off gauges. A metal foil can stand in for the coating only on a magnetic gauge, and only if it is nonmagnetic: an amplitude-sensitive eddy current gauge reads a conductive foil as part of the base metal. A coated standard made in the same process is closer to the real part. Either way, check the gauge with a standard close to the thickness you are measuring, not only at zero.

XRF

  • The spot has to fit the feature. The measurement area depends on the X-ray optics and the distance to the part, not only the collimator size printed on the instrument. A spot that overlaps the edge of a contact pad averages coated and uncoated areas together and reads low, and a spot wider than a thread crest also takes in the flanks, which sit farther from the detector. A smaller collimator fits a small feature but collects fewer counts, so it needs a longer measuring time for the same repeatability.
  • Focus distance matters. When the instrument reads the coating’s own X-ray emission, as in the usual measurement, a part that sits farther from the tube and detector than during calibration makes the coating appear thinner, and a part that sits closer makes it appear thicker. Hitachi High-Tech notes that on multilayer coatings these errors add up. Use the instrument’s focusing aid every time, and check whether it compensates for distance.
  • The substrate must match the calibration. Elements in the substrate can interfere with the coating signal. A calibration made on one brass or steel can be wrong on another alloy. The error is largest, and the coating reads thick, when the substrate contains the same element as the coating, such as nickel plating on a nickel-bearing steel.
  • Stay inside the calibrated range. The relationship between X-ray intensity and thickness is not linear over large ranges. Readings outside the range of the calibration standards are unreliable.
  • Use certified standards and check them. Reference standards from an accredited supplier, periodic recertification, and routine checks against a known standard keep the calibration honest.

📍 Where to Measure: Significant Surfaces and Minimum Thickness

Most plating specifications state thickness requirements for significant surfaces, the surfaces where the coating must meet the specification, usually identified by the purchaser on the drawing. The requirement is usually a minimum local thickness, not an average over the part. Some coating specifications state a nominal thickness with a tolerance, or an average, instead: NASA’s PRC-5006 sets hard anodize at 0.0020 in ± 0.0004 in (about 51 ± 10 µm) unless the drawing says otherwise. Many specifications also limit where a minimum applies. For example, the federal nickel plating specification QQ-N-290A, now canceled and superseded by SAE AMS-QQ-N-290, applied its Class 1 thickness requirement to visible surfaces that a ball 19 mm (0.75 in) in diameter can touch and set a thinner requirement for threads, holes and deep recesses. Other specifications use similar ball rules. Check where your specification applies the minimum before measuring inside a recess.

ISO 2064 sets the definitions and conventions behind that requirement. A specified minimum thickness is a local value, not an average: readings taken within a small reference area give one local thickness, and every local thickness on the significant surface has to meet the minimum, as NIST’s guide summarizes it. Each reading itself covers only the instrument’s small measuring area, the probe footprint or the XRF spot. QQ-N-290A, for example, rejected a part if any of its measurements fell below the minimum. The governing specification sets how many readings make up a local thickness and how many locations to check. How many parts to check per lot is a separate decision: when the specification does not set its own plan, use ASTM B602 for attribute sampling or ASTM B762 for variables sampling of metallic and inorganic coatings.

Electroplated thickness is rarely uniform: edges and corners concentrate current and build heavy, while recesses, bores and areas shielded by the rack or by other parts plate thin, so the easiest flat face is rarely the thinnest place. Measure where thickness is expected to be lowest within the area the specification covers, as well as on a reference flat, and record which is which. Where a gauge cannot reach the low current density area, plan a cross section on a sample part. A coulometric test also works there if the cell can seal flat, because a leak near a bend or edge enlarges the stripped area. A Hull cell panel shows how a bath deposits across a range of current density before you measure a difficult part; Precision Electroplating: Hull Cell Analysis covers running one.

📋 What to Record With Every Thickness Result

A thickness number without its context cannot be checked later. When a reading is questioned, these records show whether the method and setup supported it.

Records That Make a Thickness Result Traceable

A checklist for production and lab thickness records. Keep what your specification and quality system require; this list covers the information most often missing when a result is disputed.

RecordWhy it matters
Method and standard (for example ASTM B568 or B499)Shows the method was within its scope for this coating and substrate
Instrument, probe or collimator, and serial numberLinks the result to a specific calibration and maintenance history
Calibration standards and the date of the last checkShows the standards matched the coating, substrate and thickness range
Base metal alloy, thickness and curvatureExplains geometry corrections and substrate effects
Layers present when the reading was takenSeparates the plated metal from passivate, sealer or topcoat in a magnetic or eddy current reading
Measurement locations on the partSeparates minimum local thickness on significant surfaces from a convenient flat
Number of readings and how they were combinedDistinguishes a single reading, an average, and a minimum
Density value used (XRF, beta backscatter, coulometric)Mass per area becomes thickness only through this value
Phosphorus content and heat treatment state (electroless nickel)Determine whether a magnetic reading was valid and which XRF standards and density apply

Gauge checks themselves are worth recording. Recording scheduled checks against known standards helps identify changes in instrument performance and supports investigation of measurements taken since the previous acceptable check. The same control chart logic used for bath analyses applies to those check readings; How to Set Control Limits for Plating & Metal Finishing shows how the limits are set.

❌ Common Mistakes When Measuring Plating Thickness

  • Gauging zinc after sealer or topcoat and reporting it as zinc. The number is high by every layer on top.
  • Using a B499 magnetic gauge on electroplated nickel over steel, or on decorative chromium over nickel. Use XRF, coulometric stripping, a cross section, or a method calibrated on nickel-plated parts from the same process.
  • Treating magnetic measurement as the default for electroless nickel on steel. Prefer XRF or a cross section unless the deposit is sufficiently nonmagnetic as plated, the governing specification permits ASTM B499, and an independent check confirms the reading. After heat treatment, do not trust an as-plated magnetic calibration.
  • Comparing an anodize gauge reading with a micrometer buildup without converting. The two agree only after allowing for the oxide that grew into the aluminum and for any metal removed before anodizing.
  • Measuring heavy hard chromium or nickel by XRF. Check the maximum measurable thickness for that coating first.
  • Measuring only the easy flat. It overstates the minimum, because the lowest readings are usually in recesses and shielded areas; measure there unless the specification exempts them.
  • Letting reference standards go uncertified. A worn or expired standard can move every reading made with it.

🔍 Keeping Gauge Checks Next to the Process Record

Lab Wizard Cloud is where a plating shop can schedule routine gauge checks as recurring tasks, each one timestamped, logged and auditable. It also keeps a history of bath analyses, with concentrations calculated from titration results, along with tank readings, chemical additions, SPC charts and alerts tracked with an audit trail.

When a thickness reading from your gauge is questioned, engineers can use those dated records to review the bath conditions during the period when the parts were plated and whether the scheduled gauge checks were completed. To talk about how your lab schedules and reviews those checks, contact the Lab Wizard team.

Frequently Asked Questions

Which thickness test should I use for zinc plating on steel?
Zinc is nonmagnetic and steel is magnetic, so a magnetic gauge under ASTM B499 or ISO 2178 is the usual nondestructive choice. Calibrate it on uncoated steel of the same alloy, thickness and curvature as the parts. The gauge also counts any passivate, sealer or topcoat above the steel, so measure before those layers go on or record which layers the reading includes. XRF measures the zinc itself and is useful on small features.
Can a magnetic thickness gauge measure nickel plating on steel?
Not with an ordinary magnetic gauge under ASTM B499, which covers nonmagnetic coatings on magnetic base metals: electrodeposited nickel is itself magnetic, so the gauge cannot separate it from the steel. ASTM B530 (ISO 2361) describes a magnetic method for nickel, but only with standards that match the deposit and substrate, and instrument makers are cautious about it on steel. XRF, coulometric stripping and cross sections are the usual methods, and some instrument makers use phase-sensitive eddy current calibrated on nickel-plated parts from the same process.
How do you measure electroless nickel thickness?
Usually by XRF calibrated with electroless nickel standards of similar phosphorus content on the same substrate, if the deposit is within the instrument’s validated range, or by a cross section. On steel, a magnetic gauge under ASTM B499 may be appropriate for a sufficiently nonmagnetic as-plated deposit when the governing specification permits it and the method has been validated against an independent measurement on parts from the same bath. Magnetic response depends on phosphorus content and heat treatment, as the nickel section explains. After heat treatment, calibrate only against standards heat treated the same way.
Is an XRF thickness reading the same as a cross section measurement?
No. XRF estimates coating mass per unit area over its measurement spot from characteristic X-ray intensities, through a calibration or model that assumes the coating’s composition, and converts that to thickness with an assumed density. A cross section measures local thickness directly under a microscope at one cut. They agree when the composition and density assumptions, the XRF calibration, the section preparation and the measurement location are right, and they can differ when any of those is wrong or when the coating varies within the XRF spot.
What is the best way to measure anodize thickness on aluminum?
An eddy current gauge under ASTM B244 or ISO 2360 is the standard nondestructive method for anodic coatings on aluminum, calibrated on the same alloy and temper as the parts. It reports the total oxide thickness, but part of the oxide grows into the aluminum, so each surface builds up by less than that. NASA’s process specification PRC-5006 describes hard anodize as typically half growth, so a diameter grows by about one coating thickness, although the share varies with the process and alloy; check whether your specification calls for coating thickness or a finished dimension. A cross section under ASTM B487 is the destructive check.
Which coating thickness measurement method is the most accurate?
No method is most accurate for every part. NIST’s summary of the optical cross-section method (ASTM B487) gives thickness to within 1 µm or 10%, whichever is greater, and an absolute accuracy of about 0.8 µm under good conditions, which suits thick coatings; XRF and SEM cross sections (ASTM B748) suit thin ones. Accuracy depends on calibration, standards and part geometry as much as on the instrument, so check the gauge’s measurement capability against the thickness tolerance. If a reading is disputed, the governing specification or purchase order decides which method is the referee.