Grinding Burn Inspection on Hardened Parts: Check Method and Scope
Share
A grinding alarm, surface discoloration, or unusual Barkhausen-noise signal can trigger investigation, but none alone proves the depth, severity, or cause of thermal damage. A review must connect the part and process question to a method that fits the material, surface and geometry, then preserve reference, coverage and acceptance evidence.
This article compares evidence paths for chemical surface-temper etch and Barkhausen-noise inspection. It does not provide acid-handling instructions, equipment settings, acceptance limits, operator qualification, root cause or product release. The drawing, process specification, validated laboratory procedure and authorized metallurgy or NDT reviewer control. For background, see heat treatment processes for undercarriage parts.
Define the suspected damage and process point
Identify the part, feature, drawing revision, material grade, heat treatment and case condition. Record the machining operation, machine and date, and whether the surface was ground, hard turned, milled or processed another way. Name the exact ground area and any later honing, superfinishing, shot peening, coating or plating.
State the engineering question without assuming the answer. It may concern localized tempering, re-hardening, a crack, residual-stress or microstructural anomaly, or an unexplained process signal. Record the trigger: power spike, coolant loss, wheel condition, feed change, dimensional result, discoloration or another deviation. Appearance and process alarms support investigation; they do not establish grinding-burn depth.
Define the lot, affected quantity, time window, adjacent operations and record owner. Preserve process data, inspection history and the state of parts before rework. A later surface operation may change the evidence, so the inspection point in the process must remain visible.
Check material, geometry and surface applicability
Compare the actual material and heat-treatment state with the proposed method's scope. The official ISO 14104:2017 page identifies Edition 3 as current and describes chemical surface-temper etch for localized overheating after grinding or post-heat-treatment machining. Its scope expressly excludes nitrided parts and stainless steels. The public page does not provide procedure details or acceptance criteria.
For Barkhausen noise, confirm that the steel is ferromagnetic and that hardness, microstructure, residual stress, coating, roughness and prior finishing are compatible with the validated application. The SAE ARP4462 revision page identifies ARP4462C, dated 28 May 2024, as current and scopes the method to ferromagnetic steel components. It does not make the technique suitable for every hardened part.
Map curvature, thickness, edges, radii, holes, shoulders and probe or viewing access. Record coatings, plating, scale, contamination, residual magnetism, roughness and part temperature. A flat reference coupon cannot automatically validate a threshold at a small radius or inaccessible root.
Do not extend chemical-etch scope to an excluded material or assume a Barkhausen probe can cover any geometry. Record the standard version, laboratory procedure, qualified part family and every deviation for engineering review.
Confirm how the material state was established. A purchase description alone may not prove delivered grade, heat treatment, case depth or surface hardness. Link material certificates, heat-treatment records and applicable measurements to the exact lot. When a method depends on properties that vary across a case or feature, record where the supporting characterization was taken and whether it represents the inspected surface.
Surface preparation also belongs in the applicability decision. Coating removal, polishing, cleaning or demagnetization can change the part or the signal and must follow an approved process. If the surface cannot be prepared within the validated method without affecting later use or evidence, keep the inspection on hold and select a qualified alternative through engineering review.
Compare what chemical etch and Barkhausen noise detect
Chemical surface-temper etch uses a controlled laboratory process to reveal contrast associated with localized surface changes and classify indications within its stated material and process scope. The result is a surface indication interpreted under the method, reference and acceptance document. This article intentionally omits reagents, concentrations, sequence, timing and handling because qualified facilities must control chemical safety and procedure details.
Barkhausen-noise inspection measures a magnetic response affected by microstructure and stress in ferromagnetic material. Its output is a relative signal within a validated setup, not automatically a hardness number, crack size or damage depth. Surface condition, probe orientation, geometry and prior processes can affect the response.
A Stresstech application note on gear-tooth testing provides vendor-specific vocabulary for equipment, references, scan access and geometry. It does not prove universal sensitivity, superiority or a KTSU test capability.
The methods therefore should not be called equivalent or substituted solely because both are used in grinding-damage work. Define what each indication means, the depth or sensitivity question, reference correlation, false-positive and false-negative controls, and any authorized follow-up. Metallography, microhardness or crack inspection may answer different questions and can require destructive sampling.
Verify references, calibration and validation
Record the equipment, probe, software and configuration; calibration and verification status; operator qualification; and laboratory procedure revision. Identify every reference sample or standard by ID, material, heat treatment, case depth, geometry, surface condition and known characterization.
A reference should represent the decision being made. A coupon of different grade, hardness, case condition, curvature or finish may be useful for an instrument check yet insufficient to justify the same acceptance threshold. Preserve the evidence that established the reference condition and the limits of that correlation.
Keep setup parameters, scan direction and speed, probe force or positioning controls, inspection environment, check frequency, repeats and deviations in the controlled laboratory record. This public checklist does not publish equipment settings. Record repeatability or validation results and how known acceptable and known affected conditions were distinguished for the particular application.
Operator competence and automated software do not replace validation. If a software version or probe changes, determine whether the reference correlation and acceptance rule remain applicable. A green screen or passed system check proves only the state defined by that control.
Control drift throughout the inspection interval. Record checks before, during and after the lot when the procedure requires them, along with response to an out-of-control reference. Parts examined since the last acceptable check may need a defined hold and re-evaluation. Do not discard a failed verification or rerun it until it passes without investigating the reason.
Validation should cover the decisions claimed by the control plan. Detection of a deliberately characterized surface condition does not by itself establish defect sizing, fatigue effect or the causes of a production event. State the validated response, population, range and known blind spots so users do not expand a screening method into a universal material assessment.
Map inspection coverage and process timing
Create a feature map that shows every inspected surface, scan path or viewed area, overlap and orientation, plus unscanned zones. State whether the lot received 100 percent inspection or a sampling plan, and link each result to part and lot identity. “Checked” is incomplete when critical radii, shoulders or reworked zones are not shown.
For chemical viewing, retain controlled observation conditions, image IDs and location references under the laboratory method. For Barkhausen data, retain raw or reviewed files, scan orientation, reference checks and equipment version. Do not infer a result for areas the probe could not reach or that were outside the viewing plan.
Place inspection in the manufacturing sequence. Was it completed immediately after grinding, before superfinish, after a corrective pass, or before coating? If later honing, shot peening, polishing or rework occurred, record whether the control plan requires renewed inspection and which evidence was affected. Preserve before-and-after versions rather than overwriting the original coverage.
Route indications to confirmation and disposition
Assign each indication an ID, part and lot, exact location, raw result or image, repeat result and reviewer. Separate detection from characterization, cause and disposition. An indication can be reproducible while its depth, mechanical significance, production cause and lot impact remain open.
Use the authorized quality and engineering plan to select any confirmation. That may include a repeated scan, another NDT method, metallography, microhardness or crack inspection, with competent personnel and safe specimen control. Do not section, etch, polish or otherwise alter a production part from a public article.
Link the applicable specification and acceptance limit, create a nonconformance record where required, contain the affected population, and record rework, scrap, use-as-is or other disposition only from the designated authority. A detected indication does not itself authorize rejection, and a repeat with no signal does not automatically release the lot.
Keep root-cause and corrective action as a later process with its own evidence. Track machine, wheel, coolant, feed, heat-treatment and handling hypotheses without converting correlation into cause. Final release must name the approved criteria, evidence, reviewer, lot scope and date.
Use the inspection-scope checklist to expose three states
| Review field | Normal | Missing | Conflict |
|---|---|---|---|
| Question and part | Material, heat treatment, feature, process point and lot are defined | Case condition or suspect operation is unknown | Report describes a different part revision or finishing state |
| Method applicability | Current method scope fits material, surface and geometry with deviations recorded | Ferromagnetic response, coating or access has not been checked | Chemical-etch scope is proposed for nitrided or stainless material |
| Reference and validation | Representative known reference, equipment, operator and correlation are traceable | Reference characterization or repeatability evidence is absent | Threshold comes from a different material, case or geometry |
| Coverage and timing | Surface map, unscanned areas, sampling, process point and reinspection are explicit | Critical radius or post-rework status is not recorded | Report claims full coverage while the map excludes a feature |
| Indication and disposition | Raw result, confirmation, criterion, NCR, authority and lot release are distinct | Acceptance source or disposition owner is absent | Detection is presented as root cause or automatic release/rejection |
A normal row supports the narrow inspection question. A missing row identifies the next evidence and owner. A conflict remains open until technical review resolves it. None of these states should be converted into pass or fail without the authorized acceptance criterion.
The useful record connects damage question, material and geometry, method version, representative reference, actual coverage, indications and authorized disposition. That chain is more defensible than treating one contrast pattern or signal number as proof of grinding burn across an entire lot.