How to Measure Case Depth: Effective Depth, Total Depth and the Hardness Limit That Decides It

10 September 2026 | Metal Hardness Testing Equipment - Bench Top and Portable Solutions, Metal Testing Equipment, Vickers and Knoop Hardness Testers

Request Quote

A drawing asks for case depth in a single number, and that number carries a tolerance, an acceptance decision and sometimes a rejected heat lot. What the drawing rarely says is which of two different depths it means, or which hardness value marks the boundary. Two laboratories can follow the same procedure on the same part, measure carefully, and report depths that differ by a noticeable margin, because they drew the line in different places. The measurement itself is not the hard part. Agreeing on what is being measured is.

This article is about the traverse: how case depth is defined, what fixes the boundary, and what the work demands from the equipment that produces it. If you are still deciding whether microhardness is the right method at all, that question is answered separately in our guide to choosing a hardness testing method for reliable QC.

Two Depths That Sound Alike and Are Not

Case hardening leaves a gradient, not a layer with an edge. Hardness falls from the surface toward the core over a distance, and where you declare the case to end is a convention rather than a discovery. Two conventions are in common use, and they answer different questions.

Effective case depth, written as case hardening depth or CHD in ISO practice, is the perpendicular distance from the surface to the point where hardness first falls to a stated limiting value. It is a functional depth: it says how far into the part the material still meets a hardness the designer asked for. Total case depth is the metallurgical one, the distance to where the case and the core are no longer distinguishable. The two numbers are not interchangeable, they do not sit in a fixed ratio, and a part can pass on one and fail on the other.

Most acceptance work runs on effective depth, because that is the number tied to a service requirement such as wear resistance or contact fatigue. Total depth tends to appear in process development, failure investigation and disputes about how a furnace cycle actually ran. When a drawing says only “case depth” with a tolerance, the first job is to establish which of the two the customer means, before a single indentation is made.

Where the depth is measured from

One Traverse, Two Depths

The same set of indentations supports both answers. The effective depth ends where the curve crosses the hardness you agreed to treat as the boundary, so moving that line moves the result. The total depth ends where the material itself stops changing, which is why the two distances are not in a fixed ratio and cannot be converted into one another.

Scroll the diagram sideways to see all of it.

The Hardness Limit Is a Decision, Not a Constant

Effective depth only exists once somebody names the hardness that defines it. For carburized and carbonitrided steels the common limit is 550 HV, or the equivalent value of 550 HK where Knoop is used instead. Read literally, case hardening depth is the perpendicular distance from the surface to the layer that has a hardness of 550 HV.

That limit is not universal, and the reason it moves is worth understanding. If the core itself is already hard, a 550 HV line can fall in material that was never part of the case, and the reported depth stops meaning anything. Where the steel shows a hardness above 450 HV at a distance of three times the case hardening depth from the surface, the limiting value may be raised, in steps of 25 units, until the boundary again separates case from core. A harder core therefore pushes the limit upward by agreement rather than invalidating the measurement.

The practical consequence is administrative rather than technical. A case depth result is only comparable with another when three things travel with it: which depth was measured, what limiting hardness was used, and at what load. A number reported without them cannot be checked by anyone, including the laboratory that produced it.

The test that decides

Does the 550 HV Line Still Separate Case from Core?

Both panels carry the same curve shape and differ only in how hard the core is. Read the hardness at three times the case hardening depth and compare it with 450 HV. On the left it falls below, the 550 HV boundary sits on the falling part of the curve, and the limit stands. On the right the core itself is above 450 HV at that distance, the crossing has slid into ground where the curve is already flattening, and the limiting value is raised in steps of 25 units until the boundary separates case from core again.

Scroll the diagram sideways to see all of it.

The Standard on Your Drawing May No Longer Exist

Case depth callouts age badly. ISO 2639 is still printed on drawings and quality plans across the industry, and it is the document many engineers first learned the subject from. It is also withdrawn.

Its content did not disappear. ISO 18203 cancels and replaces ISO 2639:2002 together with ISO 3754:1976 and ISO 4970:1979, folding three documents that measured surface hardened layers in almost the same way into a single standard. For a laboratory, that means an incoming drawing citing ISO 2639 is a legacy callout, not an error, and the correct response is to confirm with the customer which document governs before quoting a result against it. Purchase specifications frequently lag the standards they reference by years, and a report that silently substitutes one standard for another creates a dispute rather than avoiding one.

Regional adoptions add their own layer, and a customer in one market may recognise a national version of the same method while rejecting the ISO number. This is why edition control belongs in the quality plan rather than in an operator’s memory.

One Standard, Three Depths, Because There Are Three Treatments

Merging three documents into one was not housekeeping. The three older standards existed because a hardened surface can be produced in fundamentally different ways, and each way leaves a gradient that has to be read against a different reference. The consolidated standard keeps all three, which is why the term “case depth” on its own is ambiguous until the treatment is named.

Carburized and carbonitrided parts are measured as case hardening depth against the 550 HV line described above. Parts hardened by heat alone, whether by flame, induction, electron beam or laser, are measured as surface hardening depth, and there the limiting hardness depends on what the component has to do in service, so it is agreed with the customer rather than assumed from a table. Nitrided parts are different again: nitriding hardness depth is referenced to the core, conventionally the core hardness plus 50 HV, which means the same nitriding cycle on two different core conditions legitimately produces two different depths.

That last case is the one that generates arguments. A supplier and a customer can both measure correctly, both report in good faith, and disagree, simply because one anchored the limit to the core it found and the other used a fixed number carried over from a carburizing specification. The treatment determines the convention, and the convention determines the number.

What the Traverse Asks of the Instrument

A case depth profile is an unforgiving use of a microhardness tester, and it exposes weaknesses that a single spot check never will. The measurement is optical: a diamond is pressed into a prepared cross-section under a light load, and the hardness follows from the diagonals of the indentation it leaves. Everything that blurs, shifts or drifts between one indentation and the next shows up directly in the depth number.

Load range is the first constraint. Our NG-1000 micro Vickers and Knoop tester works between 10 g and 1 kg, and that span matters because the indentation has to stay small enough to resolve a steep gradient while remaining large enough to measure reliably. A load chosen for convenience near the surface can put neighbouring indentations close enough to interfere with each other, and the profile flattens in exactly the region where it should be steepest.

Why a crowded profile reads flat

Spacing Decides What the Traverse Sees

The lower strip shows the two spacings: zones of deformed material that stay apart, and zones that overlap. The plot above shows what that costs. A crowded indentation lands in metal its neighbour has already work hardened, and the reading is lifted most where the gradient is steepest, because that is where two neighbouring points differ most. The recorded curve is therefore flatter than the real one and meets the limiting hardness further from the surface, so the case is reported deeper and gentler than it is. The curves show the mechanism, not a measured profile.

Scroll the diagram sideways to see all of it.

The choice between the two indenters belongs to the traverse rather than to the laboratory’s habits. A Vickers indentation is symmetrical and is the default for hardness profiles, measured under ISO 6507. A Knoop indentation is elongated, so its long diagonal can be read at a useful size while the short one stays narrow, and indentations can sit closer together along the direction the gradient runs. On a steep case, or on a thin layer where a Vickers impression would run out of room, that geometry is what keeps the profile from being averaged into a straight line. The method for it is ISO 4545, and an instrument that carries both indenters lets the operator pick per job instead of per machine.

Optics and stage behaviour decide whether the profile repeats. Reading a diagonal is a human judgement unless the system measures it for you, and across a traverse of many indentations that judgement drifts. Positioning has to be equally disciplined, because a depth result is a hardness plotted against a distance, and an uncertain distance corrupts a perfectly good hardness. Repetitive work of this kind is where motorised stages, automated focus and image based measurement earn their place, and where manual instruments quietly lose repeatability between operators.

Optical analysis software on the NG-1000 showing a Vickers indentation with the measured diagonals and the resulting hardness value

The diagonals are the measurement. Image based measurement puts the edges of the indentation under the same rule on every reading, which is what keeps a traverse of many indentations comparable from the first to the last.

Preparation sits underneath all of it. The cross-section has to be cut without heating the surface, mounted with the edge properly supported, and polished so the first indentation can sit close to the surface without falling into a rounded edge. A specimen that loses its edge during preparation shortens every depth measured on it, and no amount of instrument quality recovers that. Our ASTM E384 page covers the microindentation method these traverses run under.

Hot mounting press closing on a specimen holder, forming the mount that supports the edge of a cross-section

Mounting decides the edge. The mount has to hold the surface of the section flat and supported right up to its edge, because that edge is where the first and most important indentation of the traverse has to sit.

Specimens loaded into the holder of an automatic grinder and polisher before a preparation cycle

Polishing decides whether it survives. A rounded edge shortens every depth measured on it, and no amount of instrument quality recovers that, so the preparation route is part of the measurement rather than a step before it.

Microhardness for case depth

What the Traverse Runs On

A case depth profile is a long series of small indentations read optically. The tester decides whether the readings repeat, and the section it reads decides whether the first indentation can sit close enough to the surface to be worth anything.

Micro Vickers and Knoop

NG-1000

Optical microhardness testing between 10 g and 1 kg, with both indenters available, so the traverse can be run in Vickers or switched to Knoop where the gradient is steep or the layer is thin.

Case depth profiles, thin coatings, individual phasesLoads from 10 g to 1 kg

View product

NG-1000 micro Vickers and Knoop hardness tester

Low-speed precision cutting

GenCut GL100E

A low speed diamond wheel cutter for small and delicate samples, with micrometer positioning of the specimen, so the section is opened where the traverse has to start.

Sectioning parts for a cross-sectionMicrometer travel 0 to 25 mm, position accuracy 0.01 mm

View product

GenCut GL100E low speed precision metallographic cutter

Reporting a Number That Survives an Audit

The last step is the one that turns a measurement into evidence. A case depth entry that an auditor can follow names the depth type, the limiting hardness, the test load, the standard and edition it was measured to, the instrument, and the verification status of that instrument on the day of the test. Traverses are usually run more than once on a part, because a single line of indentations through a gradient is a small sample of a surface treated component.

What has to travel with the number

Four Entries That Make a Case Depth Checkable

A depth on its own cannot be verified by anyone, including the laboratory that produced it. These four turn it into a result someone else can reproduce.

Which depth

Effective, written as case hardening depth, or total. They are not interchangeable and they do not sit in a fixed ratio, so the type is not optional.

Which limiting hardness

550 HV for carburized and carbonitrided work, or the value it was raised to in steps of 25 where a hard core forced the boundary to move.

At what load

The indentation size follows the load, and the load decides how close to the surface the first reading could legitimately sit.

Standard and edition

Plus the instrument and its verification status on the day. An incoming drawing citing a withdrawn standard is a legacy callout, not an error.

None of this is difficult once it is written into a procedure. All of it is expensive to reconstruct afterwards, and the most common reason a case depth result cannot be defended is not a bad instrument but a report that recorded a depth and nothing else.

Talk to us about case depth

Tell Us the Depth Range and the Limiting Hardness

Which depths your drawings call for, what limiting hardness your customers specify, and whether the parts arrive as cut sections or as whole components. That is enough for us to say what load range, what optics and what preparation the traverse needs, and what the verification schedule should look like.

Request a quotation