Technical Cleanliness Reports for Machined Parts: Read the Limits
Share
A technical-cleanliness report can list particle counts, size classes and a cleanliness code without showing whether the laboratory tested the surfaces that matter to the buyer. A result per component cannot be compared directly with a limit per unit area, and a clean-looking filter says little if the extraction method could not reach an internal passage.
Review therefore follows a chain: inspected part and surface, particle definition and normalization, extraction and contamination controls, analysis method, reported result, and contractual acceptance limit. Dimensional requirements remain with the existing guide to machined track-roller tolerances. This article addresses residual particulate cleanliness. It does not extend the result to film contamination, operating-fluid cleanliness, cleanroom classification or cosmetic defects.
Define the surfaces and particles of interest
Begin with the controlled part number, drawing revision, supplier lot and sample identity. Record the sample quantity, test date and laboratory. Then identify the functionally relevant surfaces or internal volume named by the customer specification. A general statement such as “complete part tested” is not adequate when plugs, blind passages, assembled covers or excluded regions change what extraction can reach.
List included and excluded surfaces explicitly. For a machined component, the scope might distinguish external machined faces, threaded holes, oil passages, sealed cavities and surfaces covered during testing. The report should show how the laboratory interpreted the specification. Never extend a valid result from tested surfaces to an untested internal passage.
Record the particle definition used by the method. The customer specification may address particles above a stated reporting threshold, specific size classes, the largest particle, total residual mass or selected material categories. Capture the smallest relevant size and the dimension used to classify particles, such as a stated length or another defined measure. Do not assume that two reports use the same particle dimension because their tables look alike.
Normalization is part of the result. Record whether values are reported per component, per defined surface area, per internal volume, per sample group or on another approved basis. A count per component cannot be compared with a count per 100 cm² until the inspected scope and normalization basis genuinely match. Dividing one by an estimated area does not create an equivalent result without an authorized calculation method.
Identify the customer specification and revision that supply the acceptance criteria. ISO’s public page for ISO 16232:2018 says the standard addresses inspection and comparable reporting of particulate contamination on functionally relevant road-vehicle components. ISO confirmed the edition in 2024. The standard does not provide a universal cleanliness limit for every component, and its scope should not be applied automatically to an unrelated product.
The report should preserve the laboratory, test date, sample receipt condition, packaging and any delay between receipt and extraction. Particles can be lost or added through handling, transport and storage. If the packaging was open, damaged or visibly contaminated, record that limitation rather than treating the laboratory result as an unquestioned picture of the production state.
Check extraction and sample preparation
Particle analysis depends on recovering contamination from the specified surfaces and transferring it to a medium or filter without adding an uncontrolled background. Record how samples were unpacked, handled and prepared. The laboratory environment, tools, containers, extraction medium and filter all need the contamination controls required by the approved method.
Review the blank level. A blank test estimates contamination contributed by the extraction and analysis process rather than the part. The report should identify the applicable blank procedure, result, timing and acceptance basis. If blank contribution is absent or excessive under the controlled method, the component result may not be decision-ready.
Record the extraction technique and parameters actually used. Pressure rinsing, agitation, ultrasound, air-based methods or combinations can recover particles differently depending on geometry and material. The applicable method may specify medium, volume, pressure, flow, time, nozzle access, orientation or other variables. This article does not select one technique for every machined part.
Ask how internal surfaces were reached. A passage that the medium cannot access is outside the demonstrated extraction scope even if the part was immersed or rinsed externally. Fraunhofer’s component and assembly cleanliness laboratory page notes that standardized or customer methods may be used and that internal surfaces commonly require extraction. It also describes analytical options that provide different size and composition information. The page does not prescribe an acceptance limit or required technique for the buyer’s part.
Review the evidence that extraction is sufficiently effective and controlled for the test object. The approved procedure may use repeat extractions and a decay evaluation or another qualification approach. Record the qualification reference, part geometry, parameter set and date. A method qualified on an open housing may not cover a narrow blind passage without evidence.
Preserve every repeated extraction result. Combining several filters or extracts should follow the controlled procedure and remain visible in the record. Do not discard an early high result because a later extraction is cleaner. The sequence may be essential to showing recovery behavior.
Check filtration details, including filter material, diameter and pore size where required, and connect them to the particle range and analysis method. The extraction medium and filter must be sufficiently clean and compatible with the parts and analysis. Record rinsing, transfer loss, spilled medium, torn filters, contamination events and any deviation.
The VDA QMC catalogue for VDA 19.1, third revised edition dates the current edition to February 2026 and describes component cleanliness testing with updated extraction and analysis topics. It is a paid industry guideline, not a mandatory requirement for every part. A report should name the actual method and edition agreed in the contract rather than merely state “VDA compliant.”
Read particle counts, sizes and analysis methods
Separate gravimetric and microscopic results. Gravimetry measures residual mass under the named preparation and weighing method. Microscopy counts and classifies visible particles under defined imaging and analysis rules. A total mass does not reveal the number or maximum size of particles, and a size-class count does not supply residual mass by implication.
For microscopic results, record the reporting threshold, particle-size classes, count in each class and maximum reported particle. Confirm which particle dimension defines the class and how touching or overlapping objects are separated. Filter occupancy can affect detection and classification; an overloaded filter may prevent reliable automated analysis.
ZEISS’s technical cleanliness analysis guidance explains in its measurement-system context that filter selection follows the specified particle range and that filter occupancy and classification affect automated microscopy and comparability. It does not prove particle source, contract acceptance or instrument suitability for every specification.
Record whether the report classifies particles as metallic shiny, non-shiny, fibers or other stated categories. Optical appearance is a classification under the imaging method, not chemical proof of composition. Do not identify a particle as steel, coating or seal material solely from brightness or color.
When composition is required, identify the approved analytical method, such as SEM-EDX or another technique, the particles selected, detection limitations and how results were interpreted. Composition analysis on selected particles should not be generalized to the entire filter unless the method supports that conclusion. Preserve the link between each analyzed particle, its image and its spectrum or raw result.
Review manual actions and artifacts. Operators may confirm or reclassify fibers, merged particles, filter defects, shadows or edge objects. The report should identify the software version, review rules and any manual changes required by the procedure. A software label is not particle-source evidence.
Keep raw images and data traceable to the report. If the PDF shows only a summary table, record where the filter scan, particle gallery and original results are retained. Capture measurement uncertainty or method limitations where supplied, especially near an acceptance boundary. Do not recalculate a pass when the required raw basis, normalization or classification rule is missing.
Link classification and acceptance to the specification
Make a line-by-line comparison with the contractual specification. For each limit, record the metric, normalization basis, size class, material or category restriction, maximum-particle rule and sample or lot scope. Confirm that the test method and revision match. A laboratory can follow its method correctly while producing a report that cannot answer a customer limit written on another basis.
Recompute pass or fail only when the controlled inputs are available and the buyer’s procedure authorizes it. Keep the reported value, limit, units and normalization together. Address borderline results using the contractual rule for uncertainty, rounding or guard bands. Do not invent an acceptance margin.
Link samples to the supplier lot and purchase order. State how samples were selected and what population the result represents. A clean sample does not automatically approve every unit or future lot. If a retest is requested, define the reason, sample selection, preparation state and whether the original nonconforming or incomplete record remains in the history.
Record deviations or concessions with the affected requirement, lot, result, authorization, scope and expiry. The report reviewer does not grant acceptance. A missing customer limit, inconsistent method edition or unmatched normalization stays open for the appropriate technical or quality owner.
| Scope and sample | Extraction and controls | Analysis and reported result | Specification and status |
|---|---|---|---|
| Part, revision, lot, sample, included surfaces and normalization all agree | Qualified extraction, blank, repeat recovery and filter controls are linked | Method, threshold, size classes, maximum particle and limitations are recorded | Normal: result and contractual limits are comparable for authorized review |
| Part identified, but internal passages and excluded surfaces are not stated | Blank and extraction-recovery evidence are absent | Summary count provided without raw image trace | Missing: laboratory or supplier supplies scope and method evidence |
| Report normalizes per component; specification limits per defined area | Extraction method is valid for the report's stated scope | Reported classes use different boundaries from the customer revision | Conflict: do not convert or reclassify without an approved basis |
| Sample and tested surfaces are traceable | Extraction and blank controls align | Method-compliant particle results reported | Unresolved: customer acceptance limit is absent, so no product decision is made |
Keep the original report, raw-data references, extraction qualification, blank records, filter or image records, specification revision and disposition together for the required retention period. A corrective-action trigger may follow a confirmed nonconformance, but its ownership and closure belong to the approved quality process.
A decision-ready cleanliness report is more than a particle code. It shows which surfaces and samples were examined, how contamination was recovered and controlled, how particles were measured and classified, and how comparable results were evaluated against explicit customer limits.