4.3 Real Surfaces: Emissivity, Reflectivity, Transmissivity

Blackbody theory assumes ε = 1. Real surfaces emit less, reflect more, and sometimes transmit. Level II quantitative thermography stands or falls on correctly handling emissivity (ε), reflectivity (ρ), and transmissivity (τ). This section gives the conservation laws, Kirchhoff’s relation, the opaque simplification used daily in the field, the graybody model, and why bare metals break naive table lookups.

The Three-Way Split of Incident Radiation

When infrared radiation strikes a surface, each watt is accounted for in exactly one of three ways:

1. Absorbed — energy enters the material (absorptivity α)
2. Reflected — energy bounces away (reflectivity ρ)
3. Transmitted — energy passes through (transmissivity τ)

Energy conservation requires:

α + ρ + τ = 1

(All three are dimensionless fractions between 0 and 1, and may depend on wavelength, angle, temperature, and surface condition.)

Emissivity and Kirchhoff’s law

A surface not only receives radiation; it also emits thermal radiation according to its temperature and emissivity ε. Kirchhoff’s law of thermal radiation (in the form used for thermography) states that, for a body in thermal equilibrium, emissivity equals absorptivity at a given wavelength and direction:

ε(λ, θ) = α(λ, θ)

Intuition: a good absorber is a good emitter; a poor absorber (shiny metal) is a poor emitter and a strong reflector.

Combining Kirchhoff with conservation gives the working identity for Level II:

ε + ρ + τ = 1

(at the wavelength and geometry of interest, under the usual equilibrium assumptions taught in certification courses).

Opaque Surfaces: The Daily Field Case

Most targets in electrical, mechanical, and building thermography are opaque in the camera’s IR band: bus bars, painted enclosures, concrete, masonry, roofing membranes (as viewed for surface temperature), motors, and insulated pipes. For these:

τ ≈ 0

Therefore:

ε + ρ ≈ 1 → ρ ≈ 1 − ε

This single relation drives countless exam questions and field decisions:

If ε is low, the camera “sees” the surroundings almost as much as the target. A polished stainless panel may report something close to the reflected wall or sky temperature rather than the metal’s true temperature unless you correct aggressively or modify the surface (tape, paint, known emitter).

Opaque does not mean “no reflection”

Opaque only means transmission is negligible. A mirror is opaque and still highly reflective. In IR, polished aluminum behaves like a thermal mirror: ε very low, ρ very high.

Semi-Transparent Materials: When τ Matters

Some materials transmit in the IR band even if they look solid to the eye:

• Certain thin plastic films and bags
• Some glasses (transmission depends strongly on wavelength; common window glass is largely opaque in LWIR but can transmit in other bands)
• IR inspection windows (polymer or crystal viewports) — Level II work applies a transmittance parameter so the camera accounts for loss through the window
• Special filters and gas cells in advanced systems

For a non-opaque path: ε + ρ + τ = 1 still holds for the material’s interaction, but the radiometric problem becomes a layered problem (scene → window → atmosphere → optics). Later camera-parameter chapters treat window transmittance as an explicit input; here, remember that τ is not always zero.

Graybody Approximation

A graybody is an idealized surface whose emissivity is less than 1 but constant with wavelength (and often assumed independent of temperature for a given surface state). Then:

E = ε σ T⁴

with a single ε multiplying the blackbody curve without reshaping it. Most camera emissivity settings implement a graybody correction: one number scales the solution.

When graybody is acceptable

• Painted metals (paint dominates optical properties)
• Many nonmetals (wood, rubber, water, brick, most organics) over the LWIR band
• Surfaces where you measured effective ε with a contact thermometer or reference emitter in the same band and geometry

When graybody is weak

• Bare metals — ε varies strongly with wavelength; LWIR and MWIR values can differ
• Surfaces with selective coatings or spectral paints
• Large angle changes (emissivity often drops at grazing angles)
• Surfaces whose oxidation or roughness changes between inspections

Level II reports should state the emissivity used and the basis (table, measurement, tape method). Blind application of a single metal “handbook” number without surface condition is a classic failure mode in QA of Level I work.

Metals: Wavelength Dependence and Surface State

Metals deserve special emphasis because they cause disproportionate measurement errors.

Why metals are difficult

1. Low emissivity → weak emission signal relative to reflection
2. High reflectivity → strong contamination from ambient sources (heaters, sun, open panels, the thermographer’s own heat)
3. Spectral variation — ε(λ) for metals often increases toward longer wavelengths in some cases and is highly alloy- and finish-dependent; published curves are not universal
4. Temperature dependence — some metals show ε rising with temperature as oxidation develops
5. Geometry — cavities and threads can raise effective emissivity (multiple reflections) even when bulk metal ε is low

Wavelength dependence in one sentence

For many bare metals, you cannot move a visible-band “shiny means cold” intuition into LWIR without checking ε(λ); spectral emissivity tables and in-band effective values matter, and a number valid at 2 µm may be wrong at 10 µm.

Practical Measurement Strategies Rooted in ε–ρ–τ

Level II methods (expanded in later chapters) all flow from this physics:

1. Raise ε / lower ρ — apply high-emissivity tape or paint on a defined spot; measure there.
2. Measure reflected apparent temperature (RAT) — quantify the radiation field being reflected so the camera can subtract or compensate reflection.
3. Use contact cross-check — measure true temperature with a trusted contact sensor on a suitable surface, then solve for effective ε.
4. Account for windows — enter transmittance when viewing through IR ports so τ is not silently assumed to be 1.
5. Control geometry — avoid steep angles that change effective ε and increase specular reflection of unintended sources.

Mental checklist before trusting a metal temperature

• Is the surface opaque in-band? (usually yes)
• What is ε, and how was it determined?
• What is ρ ≈ 1 − ε saying about reflection risk?
• What warm or cold objects can the surface “see” in reflection?
• Is there a window (τ < 1) in the path?
• Is graybody acceptable, or is the metal spectral enough to demand a measured effective ε?

Connecting Back to Stefan–Boltzmann and Planck

• Stefan–Boltzmann / graybody: emitted exitance ≈ εσT⁴ (broadband idealization).
• Planck: emission spectrum shape for the blackbody portion; real spectral radiance ≈ ε(λ) × blackbody spectral radiance.
• Camera: measures band-limited mix of emitted + reflected (+ transmitted path effects).
• Your job: unmix that signal with correct parameters so reported temperature approaches true surface temperature.

If ε is set too high on a low-ε metal, the camera attributes too much of the collected radiation to emission and reports a temperature that is too low (classic systematic error). If ε is set too low, reported temperature is typically too high. Reflection errors can swing either way depending on whether the reflected scene is hotter or colder than the target.

Exam-Focused Summary Table

Summary for Recall

Real surfaces split incident IR into absorption, reflection, and transmission with α + ρ + τ = 1. Kirchhoff equates ε and α, producing ε + ρ + τ = 1. For the opaque equipment you inspect daily, τ ≈ 0 and ε + ρ ≈ 1, so shiny low-ε metals are thermal mirrors and dull high-ε coatings are trustworthy emitters. The graybody model (constant ε) is a workhorse for nonmetals and coatings but is often inadequate for bare metals, which show strong wavelength and surface-state dependence. Level II quantitative competence is the ability to choose ε, manage ρ, and include τ whenever the path is not a simple opaque free view of a well-characterized emitter.