Guide
How do you call out effective case depth so a heat treater cannot misread it?
Effective case depth is the distance from the surface of a case-hardened steel part to the depth where hardness falls to a stated limit value, measured on a polished cross section with a microhardness traverse. It tells you how much hard, load-carrying case the part really has, which is why a callout that names the depth but not the limit hardness, the test method, the location or the stage (before or after grinding) can be read three different ways and still be “met”.
By Saola Manufacturing engineering team · Published · Updated
Carburized, carbonitrided, nitrided and induction-hardened parts fail in the field in a few recognizable ways: a gear tooth that pits or spalls, a pin that wears through to soft core, a cam follower whose case crushes, or a thin section that cracks because the case ran all the way through it. Very often the heat treater did exactly what the drawing said, and the drawing said too little. This guide shows how to recognize a case depth problem, what causes it, how case depth is measured under the published methods, how to fix a lot that is out, and a callout format that leaves nothing for the heat treater or the inspector to interpret.
On this page
How do you know a case-hardened part failed because of its case depth?
A shallow case shows up as wear through to the softer core, pitting that turns into deep spalls, or case crushing under heavy contact; an excessive case shows up as brittle cracking at tooth tips, edges, thin webs and holes. Surface hardness is often in range on both kinds of failed part, which is the clue that depth, not hardness, is the problem.
Look at where the damage starts. Wear that suddenly accelerates after a period of normal running means the hard layer has been worn through. Spalls with a flat floor below the surface, rather than shallow surface pits, point to subsurface fatigue: under rolling or sliding contact the highest shear stress sits some distance below the surface, and if the hardness at that depth is too low the case cracks from underneath and breaks out in flakes. Cracks that run across a tooth tip, a sharp edge or the wall of a cross hole point the other way, to a section that was hardened through and has no tough core left to stop a crack.
The quickest confirmation is a cross section through the failed feature with a microhardness traverse from the surface into the core. Compare the depth where hardness drops below the drawing limit with the depth of the spall floor or the crack origin. If they match, the case depth (or its definition) is the root cause, and the surface hardness certificate that came with the lot will not have caught it, because a surface reading says nothing about depth. The hardness scale guide explains why a standard Rockwell C reading on a thin case also reads low by pushing through into the core.
What causes case depth problems on case-hardened parts?
In rough order of how often they explain a dispute: an ambiguous callout (effective versus total depth, or no limit hardness), grinding after hardening that removes case the drawing never accounted for, the wrong depth for the contact load or section size, and process variation in the furnace or the induction setup.
1. The callout does not define the depth. “Case depth” alone can mean total case depth, which runs to where the case can no longer be told from the core, or effective case depth, which stops at a limit hardness. The two can differ widely, and even “effective” depends on which standard sets the limit. Per SAE J423, case depth can be reported as effective or total on hardened specimens, and the relationship between depths measured by different methods can vary widely with the steel, the case and the quench. Per ISO 2639, the case-hardened depth of a carburized case runs to a limiting hardness of 550 HV 1 unless another value is agreed, and that limit is not the same as the 50 HRC equivalent commonly used under SAE J423. A drawing written to SAE J423 and a heat treater working to ISO can both be right and still disagree.
2. Grinding after heat treatment. Carburized and induction-hardened parts are usually ground on the functional surfaces after hardening, and the stock comes out of the case. Heat treat distortion makes the removal uneven, so the high side of a journal or the loaded flank of a tooth can lose much more case than the nominal stock. If the drawing does not say whether the depth applies before or after grinding, the heat treater will measure it as heat treated.
3. The wrong depth for the job. Too shallow, and the contact stress peak sits in softer material below the case. Too deep on a small part, and thin sections harden through and turn brittle. Gear rating practice from AGMA treats minimum effective case depth as a design value tied to tooth size and contact load, not a habit carried from the last drawing.
4. Process variation. Carbon potential, time at temperature, quench severity and load position all move carburized case depth; coil coupling, frequency, power and scan speed move an induction case. A depth that is right on a test coupon can be wrong at a tooth root or inside a bore where the atmosphere or the field does not reach the same way.
| Process | SAE J423 practice (US drawings) | ISO 18203 depth and limit hardness |
|---|---|---|
| Carburized or carbonitrided | Effective case depth to 50 HRC, usually read as about 513 HV on a microhardness traverse | Case hardening depth (CHD) to 550 HV 1 (about 52 HRC) unless another limit is agreed |
| Induction or flame hardened | Effective case depth to a limit hardness that depends on the carbon content of the steel | Surface hardening depth (SHD) to 0.8 times the minimum specified surface hardness |
| Nitrided or nitrocarburized | No single carburizing-style limit to rely on; state the limit hardness on the drawing | Nitriding hardness depth (NHD) to core hardness plus 50 HV |
| Any process, total depth | Total case depth to where case and core can no longer be told apart (hardness, etch or chemistry) | Total thickness of the surface-hardened layer, method stated |
How is effective case depth measured and verified?
By sectioning a part or a representative coupon, polishing the section, and running a row of Knoop or Vickers microhardness indents from the surface inward until hardness passes the limit value; the effective case depth is where the hardness curve crosses that limit. A surface hardness test, a file check or a macro-etch cannot give you that number on its own.
The traverse is made under ASTM E384, which covers microindentation hardness with Knoop and Vickers indenters at test forces from 1 to 1000 gf; ISO practice uses a Vickers HV 1 traverse by default. The load matters because a heavier indent near the surface averages over a deeper band of material, so the report must state the indenter and the force. Convert between HRC and HV only with the steel table in ASTM E140, and treat the converted value as a reference, not the acceptance number.
Three choices decide whether the measurement means anything:
- Where. On a gear, the pitch line and the root fillet can carry very different case depths, and the root is where bending fatigue starts. On a shaft, measure on the journal that carries the load, not on an end face. Mark the location on the drawing.
- On what. A coupon of the same steel and section, run in the same load, is cheaper than scrapping a part, but it only represents the part if its section and position in the load are comparable. Say whether a coupon is acceptable and what it must match.
- When. Measure on the finished surface if the requirement is a minimum after grinding. A depth measured before grinding has to be reduced by the stock actually removed at that spot, which is not the nominal stock on a distorted part.
Read the rest of the section while it is polished. Retained austenite, intergranular oxidation at the surface of an atmosphere-carburized part, a decarburized skin or grinding burn can all make a part with a correct depth fail anyway. Carburizing specifications such as SAE AMS2759/7, which covers gas and vacuum carburizing of carburizing-grade steel parts, are where those limits are usually written, together with the general requirements of SAE AMS2759. For grinding damage on gears, a surface temper etch inspection such as ISO 14104 is the usual screen.
How do you fix or prevent a case depth problem?
A case that is too shallow can sometimes be re-carburized before final grinding, but a finished, ground part with too little case usually cannot be rescued; a case that is too deep cannot be reduced. Prevention is a defined callout, enough case to survive grinding, and a sectioned sample per heat treat lot.
Use the table to decide what to do with a suspect lot and what to change before the next one. The first question is always whether the lot actually failed the drawing, or failed an assumption that was never written on it.
| What the section shows | Likely cause | Action on this lot | Change for the next lot |
|---|---|---|---|
| Depth meets the supplier’s definition but not yours | Callout did not name effective vs total depth or the limit hardness | Material review; judge fitness for use by the load case | Rewrite the callout with limit hardness, method and location |
| Depth in range before grinding, short after | Grind stock and distortion not allowed for | Sort by sectioned samples; scrap parts short on the loaded surface | Specify minimum depth after grinding; control distortion and stock |
| Depth short everywhere, before grinding | Cycle too short or carbon potential too low | Re-carburize only if not yet finish ground and the specification allows rework | Ask for the cycle record and a coupon per load |
| Depth short only at roots, bores or masked edges | Atmosphere or field not reaching the feature | Section the feature itself before accepting | Name the feature as a measurement location |
| Thin sections hardened through and cracked | Depth too large for the section | Reject; cracked parts are not reworkable | Reduce or cap the depth; add a maximum to the range |
| Depth correct, part still spalled | Case too shallow for the contact load, or a metallurgical defect | Engineering review of the design depth; check microstructure | Recalculate the required depth from the contact stress |
Two habits prevent most of these disputes. Give the depth as a range with a minimum and a maximum, because a minimum alone invites a deep case on small parts and an open maximum makes cracking a supplier’s gamble. And allow for grinding in the routing: rough machine, case harden, then finish grind, with the case depth required after grinding and the heat treater told how much stock will come off. The stress relieving guide covers sequencing that keeps distortion, and therefore uneven grind stock, under control.
What should the effective case depth callout on the drawing say?
Seven items: the process, the depth as a range, the word effective or total, the limit hardness that ends the depth, the test method and load, where it is measured, and whether it applies before or after finish grinding. Add surface and core hardness so the profile is defined at both ends.
The examples below show the format; the depths and hardness values in them are placeholders showing where each item goes, not design values for your part. Put the callout on the feature it applies to, and add the same items to the RFQ so the quote prices the coupon, the sectioning and the grind stock. The 1018 vs 1045 guide shows the material line that goes above it.
| Item | Weak wording | Wording that cannot be misread |
|---|---|---|
| Process | “Case harden” | “Gas carburize, quench and temper” or “Induction harden surface B only” |
| Depth and definition | “Case 0.8 mm” | “Effective case depth 0.6–0.9 mm” |
| Limit hardness | Not stated | “To 513 HV (50 HRC equivalent) per SAE J423” or “CHD to 550 HV 1 per ISO 18203” |
| Method and load | Not stated | “Vickers microhardness traverse per ASTM E384, HV 1” (state whichever force you choose) |
| Location | Not stated | “Measured at pitch line and root fillet, section A-A” |
| Stage | Not stated | “Depth applies after finish grinding” |
| Surface and core | “58 HRC” | “Surface 58–62 HRC equivalent; core 30–45 HRC at section center” |
| Sampling | Not stated | “One part or matching coupon sectioned per heat treat load; report traverse” |
Pick one system and stay in it. If the drawing cites SAE J423, give the limit per SAE J423 as 50 HRC equivalent or its HV value; if it cites ISO, use CHD, SHD or NHD with the ISO limit. Mixing an ISO depth symbol with an HRC limit is a common way to get two labs reporting two different depths from the same section. Mask threads and other surfaces that must stay soft by naming them, and say how: copper plating or a stop-off paint for carburizing, or no hardening within a stated zone for induction.
Who does what on a Saola order with a case-hardened part?
Turning and milling run at our lead CNC plant; case hardening is not on that plant’s process list. Say on the RFQ that the part is case hardened and attach the callout, and the reply will state whether and where the heat treatment and the case depth test can be arranged, with the production site and country of origin named on the quote.
Soft machining before hardening is done on the lathes and machining centers at our lead CNC plant, as described on the CNC turning and CNC milling pages. That plant’s equipment list includes grinding/polishing and a hardness tester, so a surface hardness check can be part of the inspection plan; whether precision finish grinding of hardened surfaces can be done there or has to be arranged elsewhere is stated on the quote, which matters when the case depth is required after finish grinding. A sectioned microhardness traverse needs a metallurgical lab, and the quote will say where that test would be done rather than assume it. Any coating after hardening goes through the surface finishing route at partner plants we manage, and the quote names where that work happens.
The most useful package is the drawing with the full callout on it, the surfaces to be ground after hardening and the stock you expect to leave on them, and the surfaces that must stay soft. Material certification is available on request, and ISO 9001 certification is in progress, not held. Send it through the quote form.
Frequently asked
What is the difference between effective case depth and total case depth?
Effective case depth stops where hardness falls to a stated limit, so it measures the useful, load-carrying case. Total case depth runs to where the case can no longer be told from the core, so it is always the larger number. A drawing should say which one it means.
Why do two labs report different case depths on the same part?
Usually because they used different limit hardness values, different indenter loads, or different locations on the part. Agree on all three in writing, and ask for the full hardness traverse rather than a single depth figure, so the curves can be compared.
Can effective case depth be checked without destroying a part?
Not as an acceptance measurement under the usual methods, which need a polished cross section. Nondestructive methods such as eddy current or ultrasonic techniques can sort parts once they are calibrated against sectioned parts, but the sectioned traverse remains the reference.
Should the case depth be specified before or after grinding?
After grinding, on the surfaces that carry load, because that is the case the part runs with. Tell the heat treater the grind stock so the as-hardened depth is set deep enough to leave the minimum after grinding on the high side of a distorted part.
Does a higher surface hardness make up for a shallow case?
No. Under contact load the damaging stress sits below the surface, and if the material at that depth is soft the case fails from underneath regardless of the surface reading. Depth and surface hardness are separate requirements.
Sources
- SAE J423, Methods of Measuring Case Depth (stabilized, scope), SAE International, 2023
- EN ISO 2639, Steels: Determination and verification of the depth of carburized and hardened cases (withdrawn, abstract), Austrian Standards listing, 2003
- EN ISO 18203:2022 (ISO 18203:2016), Steel: Determination of the thickness of surface-hardened layers (withdrawn 2026), UNI listing, 2022
- ISO 18203:2026, Steel, Determination of the thickness of surface-hardened layers, ISO, 2026
- ISO 14104:2017, Gears, Surface temper etch inspection after grinding, Chemical method, ISO, 2017
- ASTM E384-22, Standard Test Method for Microindentation Hardness of Materials (scope), ASTM International, 2022
- ASTM E140, Standard Hardness Conversion Tables for Metals, ASTM International, current edition, accessed October 2026
- SAE AMS2759/7F, Gas and Vacuum Carburizing and Heat Treatment of Carburizing Grade Steel Parts (scope), SAE International, 2026
- SAE AMS2759, Heat Treatment of Steel Parts, General Requirements, SAE International, accessed October 2026