If you have ever stood in front of a failed part and asked whether the problem started at the surface or in the bulk, you already know why the words coating and surface treatment matter. A coating is a distinct layer added on top of a substrate. A surface treatment changes the near-surface region of the substrate itself, usually without creating a sharp, separable interface. For engineers and quality professionals who read fracture surfaces, measure microhardness, or interpret adhesion test data, the difference is not semantic. It changes what you look for under a microscope, what you measure with a profilometer, and what you write in a corrective action report.
We will walk through the definitions, the process signatures, the failure modes, and the measurement evidence that separate these two categories. You will see why a 25 µm electroless nickel deposit behaves differently from a 0.2 mm induction-hardened case, and why a phosphate conversion coating is not the same as a nitrided diffusion zone. By the end, you should be able to look at a cross-section or a test report and say which one you are dealing with.
What a Coating Is
A coating is a layer of material that is applied to a substrate and remains largely distinct from it. The interface may be mechanical, chemical, or metallurgical, but the coating has its own composition, its own thickness, and often its own grain structure. Common examples include electroplated zinc, hot-dip galvanizing, thermal-sprayed tungsten carbide, paint, and physical vapor deposition (PVD) chromium nitride.
When we measure a coating, we usually report a thickness in micrometers or mils. A zinc electroplate might be specified at 8–12 µm. A thermal spray coating might be 150–300 µm. The substrate underneath is unchanged in composition, except for a thin interdiffusion zone in some high-temperature processes.
From a failure analysis perspective, coatings fail in predictable ways. You see blistering, flaking, spalling, or interfacial corrosion. Adhesion is a key property. A pull-off test might show failure at 5 MPa, or a bend test might reveal cracking at a 10 mm mandrel diameter. The coating is a separate entity, and the evidence usually shows that separation.
What a Surface Treatment Is
A surface treatment modifies the existing surface or near-surface region of the part. There is no new layer with a sharp boundary. Instead, the composition, hardness, residual stress, or topography of the original material changes. Examples include carburizing, nitriding, induction hardening, shot peening, laser peening, anodizing, and phosphating.
Some treatments, such as anodizing, do build a layer, but that layer grows from the substrate by oxidation. The interface is gradual, and the anodic film is integral to the aluminum part. A hard anodize layer of 50 µm is not a coating in the same sense as a 50 µm electroplated nickel layer. The anodize layer has a cellular structure that connects to the base metal, and its hardness is a property of the oxide, not of a foreign material.
For measurement, we often report case depth rather than coating thickness. A carburized gear might have an effective case depth of 0.8–1.2 mm at 550 HV. A nitrided shaft might have a white layer of 5–15 µm and a diffusion zone of 0.3 mm. The transition from surface to core is graded, not sharp.
Why the Distinction Matters in Failure Analysis
When a coated part fails, the first question is usually about adhesion. Did the coating separate from the substrate? Was there corrosion under the coating? Did the coating crack and allow moisture to reach the base metal? The evidence is often visual: a flake of chrome plating on a hydraulic rod, a rust blister under paint on a steel panel, or a spalled thermal barrier coating on a turbine blade.
When a surface-treated part fails, the question is different. Did the case depth meet specification? Was the hardness profile correct? Did the treatment introduce harmful residual stresses or microcracks? A nitrided part that fails by spalling may show a white layer that is too thick and brittle. A carburized gear that pits may have a case that is too shallow or a core hardness that is too low.
We can see the difference in cross-section. A coated sample shows a clear boundary. A surface-treated sample shows a gradient. If you etch a carburized steel cross-section with 2% nital, you see a dark case that fades into a lighter core. If you etch a zinc-plated steel cross-section, you see a bright zinc layer sitting on a darker steel substrate. The metallographic evidence is not ambiguous.
Process Signatures You Can Measure
Every process leaves a signature. Coatings leave a thickness signature. Surface treatments leave a depth profile. Here are the measurements we use most often.
Coating Thickness and Adhesion
For coatings, we measure thickness with a magnetic gauge, eddy current gauge, X-ray fluorescence, or cross-section microscopy. A magnetic gauge on a steel panel might read 12 µm of zinc. A cross-section at 500x might show a 10 µm nickel layer with a sharp interface. Adhesion is tested with a pull-off tester, a bend test, or a tape test. The numbers are direct: 8 MPa pull-off strength, no flaking after a 180° bend over a 13 mm mandrel, or a 5B tape test rating.
Case Depth and Hardness Profiles
For surface treatments, we measure case depth with a microhardness traverse. A carburized part might be specified as 0.8 mm effective case depth at 550 HV. A nitrided part might be specified as 0.3 mm total case depth. The hardness profile is the evidence. A plot of Vickers hardness versus depth shows a plateau near the surface, then a gradual drop to core hardness. A coating, by contrast, shows a step change at the interface.
Residual Stress and Topography
Shot peening and laser peening are surface treatments that introduce compressive residual stress. We measure that stress with X-ray diffraction. A shot-peened spring might show -600 MPa at the surface, fading to near zero at 0.2 mm depth. A coated part might have residual stress in the coating itself, but the substrate stress is usually unchanged. Surface roughness is another signature. A phosphate coating might have a roughness of 1.5 µm Ra, while a ground and nitrided surface might be 0.2 µm Ra. The topography tells you what process was used.
Common Misclassifications
Some processes sit in a gray zone. Anodizing is a surface treatment because the oxide grows from the aluminum. But a sealed anodize layer can look like a coating in cross-section. The difference is the interface. An anodize layer has a gradual transition with a barrier layer at the base. A paint layer has a sharp line.
Phosphating is a surface treatment. It converts the steel surface to a crystalline phosphate layer. The layer is thin, usually 2–20 µm, and it is integral to the surface. It is not a coating in the same way that an electroplated zinc layer is. The phosphate crystals grow from the substrate, and the interface is interlocking.
Thermal spray is a coating. The molten particles splat onto the surface and solidify. There is a mechanical bond, sometimes with localized diffusion, but the coating is a separate material. A tungsten carbide thermal spray coating is not a surface treatment, even though it changes the surface properties.
What the Standards Say
Standards help keep the language straight. ISO 2064 defines coating thickness measurement methods. ASTM B487 is for measuring metal and oxide coating thickness by cross-section microscopy. For surface treatments, SAE J423 covers case depth measurement for carburized and carbonitrided parts. ISO 18203 covers nitriding depth measurement. When you write a report, citing the right standard tells the reader which category you are in.
For adhesion, ASTM D4541 is the pull-off test for coatings. ASTM B571 covers qualitative adhesion tests for metallic coatings. For surface treatments, there is no adhesion test because there is no separate layer to pull off. Instead, you measure case depth, hardness, and residual stress. The test method itself reveals the category.
Failure Modes: Coating vs. Surface Treatment
Let us look at two real-world examples. A zinc-plated bolt fails by red rust after 200 hours in salt spray. The coating is 8 µm thick, but the salt spray test shows white rust at 48 hours and red rust at 200 hours. The failure is in the coating. The fix might be a thicker coating, a different passivation, or a different coating system.
A carburized gear fails by pitting after 10,000 cycles. The case depth is 0.6 mm, but the specification called for 0.9 mm. The hardness at 0.6 mm depth is 500 HV, but the specification called for 550 HV minimum. The failure is in the surface treatment. The fix is a deeper case or a higher surface hardness.
In both cases, the evidence is in the cross-section and the hardness profile. The coating failure shows a sharp interface and a thin layer. The surface treatment failure shows a shallow case and a hardness drop. The corrective actions are different because the processes are different.
How to Choose Between a Coating and a Surface Treatment
The choice depends on the failure mode you are trying to prevent. If you need corrosion resistance, a coating is often the answer. Zinc, nickel, chrome, and paint all provide a barrier. If you need wear resistance or fatigue strength, a surface treatment is often better. Carburizing, nitriding, and shot peening change the material itself.
Sometimes you need both. A carburized gear might also be phosphate coated for break-in lubrication. A nitrided shaft might be coated with a thin PVD layer for low friction. The combination works because the surface treatment provides the load-bearing case, and the coating provides the surface chemistry. But you must measure both. The case depth and the coating thickness are separate specifications.
Measurement Pitfalls to Avoid
One common mistake is measuring a surface treatment with a coating thickness gauge. A magnetic gauge on a carburized part will give a number, but that number is meaningless. The gauge is calibrated for a non-magnetic coating on a magnetic substrate. A carburized case is still magnetic, so the gauge reads zero or a nonsense value. Use a microhardness traverse instead.
Another mistake is reporting a nitrided white layer as a coating thickness. The white layer is a compound layer of iron nitrides. It is part of the surface treatment, not a separate coating. If you report it as a coating, you will confuse the reader and possibly trigger the wrong corrective action.
When you prepare a cross-section, be careful with edge retention. A soft coating can smear during grinding and polishing. A hard case can chip. Use a hard mounting compound, gentle pressure, and a final polish with 1 µm diamond. The interface must be sharp and clean for accurate measurement.
What to Put in Your Report
When you write a failure analysis report, state the category clearly. Say “the part has a 12 µm electroplated zinc coating” or “the part has a 0.8 mm carburized case.” Do not say “the part has a surface treatment” when you mean a coating. The reader needs to know what to look for in the cross-section and what test methods apply.
Include the measurement data. For a coating, report thickness, adhesion, and composition. For a surface treatment, report case depth, hardness profile, and residual stress if measured. If you used a standard, cite it. If you used a cross-section, include the magnification and etch. The evidence should be reproducible.
FAQ
Is anodizing a coating or a surface treatment?
Anodizing is a surface treatment. The anodic oxide grows from the aluminum substrate by electrochemical oxidation. The interface is gradual, and the oxide is integral to the part. It is not a separate applied layer like paint or electroplated metal.
Can a part have both a coating and a surface treatment?
Yes. A common example is a carburized gear that is also phosphate coated. The carburizing provides a hard, wear-resistant case. The phosphate coating provides break-in lubrication and mild corrosion protection. The two are measured separately: case depth by microhardness, coating thickness by weight or microscopy.
Why does a coating fail by flaking while a surface treatment fails by spalling?
Flaking in a coating usually means poor adhesion or interfacial corrosion. The coating separates from the substrate at a sharp interface. Spalling in a surface treatment usually means the treated case is too brittle or the residual stress is too high. The failure occurs within the treated layer or at the case-core transition, not at a sharp interface.
How do I measure case depth on a nitrided part?
Use a microhardness traverse from the surface inward. The total case depth is the distance where the hardness drops to a specified value, often 50 HV above core hardness. The white layer, or compound layer, is measured separately by microscopy. Do not use a coating thickness gauge.
Next Step for This Blog
This article is the first in a series on surface engineering for failure analysts. The next article will cover how to prepare and etch cross-sections for coating and case depth measurement. We will look at specific etchants, mounting techniques, and common artifacts that can fool you under the microscope. If you have a cross-section that is giving you trouble, send a note through the contact page and we may use it as an example.
