Eddy current inspection to verify heat treatment, confirm the temper and master the six factors that alter every reading — within the framework of AMS2658D.

Electrical conductivity is the ease with which electrons move through a material. The more freely they flow, the higher its conductivity.
In the aerospace industry this measurement is a safety tool: it lets you verify the heat treatment of aluminum structures, confirm the corrosion-resistant temper in 7xxx alloys and detect heat damage — all non-destructively, without compromising the part. It is the basis of specifications such as AMS2658D.
All matter is made of atoms with a nucleus of protons and neutrons surrounded by orbiting electrons. The flow of those electrons through the material is an electric current. The relative ability of a metal's atoms to allow that flow is precisely its conductivity: a physical property that depends on the composition, the crystal structure and the treatment condition of the material.
Conductivity is a highway for electrons. In a pure metal the highway is clear and current flows fast (high conductivity). Each alloying atom is like a car parked in a lane: it gets in the way. That is why a little alloying element slows traffic a lot — and why conductivity gives away the composition and condition of the material.
In non-destructive testing, a coil excited with alternating current induces eddy currents in the part. The material's conductivity changes the complex impedance of the coil; the instrument converts that impedance change into a conductivity value. That is why it is an indirect measurement, sensitive to anything that alters the path of the induced currents.
Every metal is assigned a conductivity value on a scale relative to pure annealed copper: the International Annealed Copper Standard.
The IACS standard defines a copper bar 1 meter long with a 1 mm² cross-section, held at 20 °C, with a resistance of 0.017241 Ω, which equals 100 %IACS. The SI unit is the siemens per meter (S/m); by convention, 100 %IACS = 58 MS/m.
The international primary standards are kept at NIST (USA) and NPL (UK), with an accuracy close to 0.2 %IACS, and are cross-calibrated against each other to maintain traceability. Secondary standards — the ones that normally come with a field instrument such as the SIGMATEST — reach a typical accuracy of 0.35 %IACS.
Measured conductivity does not depend on the material alone: geometry, temperature and test frequency change the result. Knowing them is the difference between a valid reading and an unjustified rejection.
Small amounts of alloying elements reduce conductivity drastically.
Temper and cold work change the structure — and conductivity with it.
The higher the temperature, the more the atoms vibrate and the lower the conductivity.
Below 3× the standard depth of penetration the reading is no longer reliable.
The cladding layer (Alclad) shifts the reading depending on the frequency used.
Convex and concave surfaces require correction factors.
Conductivity usually drops dramatically with very small amounts of an alloying element. The purity of the base metal matters less than what is added to it.
The correlation between hardness and conductivity is the basis of heat-treatment verification by eddy current. But neither one is enough on its own.
Heat treatment and cold work rearrange the crystal structure of the alloy. The annealed (O) form of any alloy has the most regular structure — and therefore the highest conductivity. When the alloy is aged or hardened, conductivity changes measurably: an annealed 2014 is around 50 %IACS, while in the 2014-T4 condition it drops to ~30 %IACS.
In aerospace aluminum alloys, this correlation is the practical tool for confirming that a part received the correct heat treatment without destroying it: measure conductivity in the field, compare it against the specification's acceptance bands, and confirm with hardness where needed.
As temperature rises, atoms vibrate with greater amplitude and interfere with the movement of electrons. The result: conductivity decreases.
The relationship is inverse and far from negligible: the same commercially pure aluminum goes from 69 %IACS at −10 °C to 62 %IACS at 20 °C. The acceptance bands in the standards are referenced to 20 °C, so every reading must be normalized to that temperature before comparing it.
Eddy currents attenuate exponentially with depth. For a valid reading, the material must be at least three times the standard depth of penetration (δ).
Eddy currents are like the echo of your voice in a cave: strong at the surface and weaker and weaker toward the bottom. The depth of penetration (δ) marks where they almost fade out. If the part is thinner than 3δ, the “echo” bounces off the far side and contaminates the reading — like trying to measure the depth of a pool that is really a puddle.
| Material | δ @ 60 kHz | Min. (3δ) | δ @ 500 kHz | Min. (3δ) |
|---|---|---|---|---|
| Al alloys (35 %IACS) | 0.46 mm | 1.4 mm | 0.16 mm | 0.47 mm |
| 304 stainless steel (2.5 %IACS) | 1.71 mm | 5.1 mm | 0.59 mm | 1.8 mm |
| Ti-6Al-4V (1 %IACS) | 2.70 mm | 8.1 mm | 0.94 mm | 2.8 mm |
Aluminum alloys are often produced in Alclad form: an alloy sheet (e.g. 2014) with a thin cladding of pure aluminum — typically up to 0.005″ — to resist corrosion.
That cladding conducts differently from the base metal, so the reading is split between the two according to how much of the eddy current energy is dissipated in the cladding — and that depends on frequency:
Measuring conductivity on clad material is like taking someone's temperature through a coat: at low frequency you mostly “feel” the body (the base metal, which is what you care about); at high frequency you only feel the coat (the cladding). That is why procedures are written at 60 kHz.
Eddy current conductivity meters are calibrated for use on flat surfaces. Measuring on curved surfaces requires care to maintain accuracy.
| Situation | Main effect | Mitigation |
|---|---|---|
| Concave surface | Probe lift-off | The instrument compensates up to a limit; watch the lift-off at the center of the probe |
| Convex surface | Indicated conductivity drops as the radius falls below ~3″ | Apply curvature correction tables |
| General rule | — | Lower frequencies give better results (except on thin-wall tubing) |
Three factors — thickness, conductivity and cladding — come together in a single practical decision: which frequency to use.
Aluminum is grouped into series according to its main alloying element. The key distinction for the inspector is whether the alloy is heat-treatable: that decides what conductivity can verify.
This is the flagship application. High-strength 7xxx alloys are tempered to peak hardness (T6), but that condition is susceptible to stress corrosion cracking (SCC). Overaging the part to T73 or T76 gains SCC resistance at the cost of some mechanical strength — and conductivity increases measurably. That is why conductivity is the non-destructive method for confirming that a part received the corrosion-resistant temper.
Back to the highway: in the T6 temper the obstacles (fine precipitates) are spread across every lane — they slow the current (low conductivity) but anchor the structure (high strength). With overaging, those obstacles gather into a few large clumps: they leave lanes clear (conductivity rises) but hold less (strength drops). That is why, in a 7xxx, more conductivity = more overaged = more corrosion resistance.
| Temper | Conductivity | Brinell min. | Condition and use |
|---|---|---|---|
| T6 | 30.5–36.0 % | 135 | Peak-aged · maximum strength · susceptible to SCC |
| T73 | 38.0–43.0 % | 125 | Overaged · resistant to stress corrosion cracking |
| T76 | 38.0–42.0 % | 130 | Overaged · resistant to exfoliation |
Conductivity bands and minimum Brinell hardness per AMS2658D, Table 1 (bare alloys). Conductivity measurement per ASTM E1004. Table 2 governs clad material.
Two warnings straight from AMS2658D that change how you inspect:
Two readings on nominally identical material can differ. Before making a call, rule out these causes:
| Variation | Why it happens |
|---|---|
| Same metal, different temper | 2024-O gives 46–51 % but 2024-T3 drops to 27.5–32.5 %. The temper rules. |
| Overlapping bands | Neighboring tempers share a range — which is why hardness is also needed to tell them apart. |
| Cladding | Reading weighted toward the cladding; no acceptance band for clad material. |
| Temperature not normalized | Shifts the reading; normalize to 20 °C. |
| Thickness < 3δ | Below the minimum the reading is no longer reliable. |
| Curved, rough or corroded surface | Does not represent the base material; apply a correction factor or recondition the surface. |
Quick field reference. Each one teaches a different inspection lesson — acceptance values per AMS2658D (bare material, 60 kHz, normalized to 20 °C).
This is how the pieces come together. A conductivity reading without hardness — and without knowing the temper's band — can be misleading. Follow the reasoning.
Test what you have learned: 5 random questions from a question bank. You will get an explanation for each one and, at the end, your score — with the option to try again with different questions.
A portable eddy current instrument that measures the electrical conductivity of non-ferromagnetic metals from the complex impedance of the probe. The measuring range is set by calibration; on an unknown part, the instrument converts the impedance into a conductivity value shown on the display.
Its measurement quality meets Boeing (BAC 5651) and Airbus standards, as well as DIN 50994, ASTM E1004 and DIN EN 2004-1 — the same standards AMS2658D invokes for conductivity measurement.
Confirm that an aluminum alloy received the correct temper by comparing its conductivity against the AMS2658D bands.
Separate and sort alloys, identify mixed materials and control purity through their conductivity signature.
Detect heat damage in wheels, engine components and lightning-strike areas, where heat alters the local conductivity.