Precision Lens Manufacturing for Rigid Medical Endoscopes

Two batches of lenses can satisfy the same component drawing yet produce different results after assembly. One may sit near a favourable combination of tolerances; another may expose a sensitive relationship between surface geometry, centration and spacing. An acceptable incoming inspection report does not explain how those parts will behave together inside a narrow, extended optical train.

That is the manufacturing problem in a rigid optical endoscope. The objective, relay optics, proximal optics and their mechanical references contribute to one image path. The question is whether the manufactured assembly preserves the intended image performance, not simply whether each lens can be polished to its drawing. This article develops that connection without describing a particular manufacturer’s optical prescription or process recipe.

1. Begin with the image requirement, not the polishing machine

Start with the required optical behaviour: resolution and contrast, centre-to-edge consistency, geometric distortion, transmission, unwanted light and focus. Where visible and near-infrared imaging are intended, include spectral transmission and chromatic behaviour in the allocation. Define field positions, working conditions and the boundary between the optical scope and the camera system before asking manufacturing to control a dimension.

Small optical diameters limit room for handling and locating features. A long shaft places several optical and mechanical interfaces between the distal image and proximal output. Relay optics transfer the intermediate image along that shaft; maintaining their relative position is part of the imaging problem. Rod-lens-based designs are one architectural approach, illustrated by the manufacturer educational material in 12, not a description of every rigid endoscope. The following blocks describe functions, not a lens prescription.

CONCEPTUAL OPTICAL ARCHITECTURE

1. Distal windowOptical entrance boundary
2. Objective opticsForms the intermediate image
3. Relay optical trainTransfers the image along the shaft
4. Eyepiece / proximal opticsDelivers the image to the interface

Mechanical reference structure
Locating and alignment relationships connect the functional groups.

Component error + assembly error + mechanical reference error → combined optical behaviour

Translate the image requirement into an optical tolerance allocation, then into manufacturable characteristics and inspection evidence. This is not a demand to tighten every tolerance. Sensitivity analysis should identify which variations materially change the required field performance and which have little influence within the proposed range. Spend process and measurement effort accordingly.

A drawing that demands exceptional surface form while leaving the optical-to-mechanical reference ambiguous may control the wrong risk. Conversely, a relatively insensitive feature does not need a severe limit merely because a machine can achieve it. Review the allocation with the people who will locate, measure and assemble the parts. Their constraints belong in the design decision before production tooling is committed.

2. Translate optical design into manufacturable and inspectable tolerances

A nominal model describes an ideal optical arrangement. A production drawing must also define what can vary, how the feature is referenced and how acceptance will be determined. Material, diameter, centre thickness, edge geometry, clear aperture and surface geometry need to remain connected to the intended assembly. A missing reference cannot be repaired by adding more decimal places. Material selection must account for refractive index and dispersion, transmission in the intended bands, and the relevant thermal and processing conditions; a glass name alone does not establish assembled performance 7.

ISO 10110-1:2019 1 supplies a common drawing language for optical elements and systems. It is not a medical-endoscope manufacturing standard or a set of ready-made product acceptance limits. Select the relevant indications and agree the applicable editions. The associated parts distinguish surface form, centring and tilt, imperfections, texture and wavefront deformation rather than treating them as one quality grade.

For each critical characteristic, identify the measured aperture, reference geometry, instrument or method, and acceptance rule. Where appropriate, transmitted wavefront can describe the combined optical effect of an element or subassembly beyond separate surface specifications. ISO 10110-14:2018 6 addresses wavefront-deformation indications; it does not require every component to carry such a tolerance. The wavelength and measurement configuration still matter.

Different specifications describe different failure mechanisms

Possible consequences depend on sensitivity, location and configuration; they are not diagnoses or universal acceptance criteria.

Specification	What it controls	Possible system-level consequence if poorly controlled
Surface form	Departure from the intended surface shape	Aberration and loss of resolution or contrast
Surface texture	Residual roughness and waviness	Scatter, redistributed light or contrast loss
Surface imperfections	Localized defects under an agreed convention	Unwanted light or local obstruction, depending on position
Centring / tilt	Relationship of optical features to reference axes	Field asymmetry or off-axis degradation
Thickness / spacing	Axial relationships between optical surfaces	Focus or aberration changes across the field
Coating	Specified spectral and reflection behaviour	Throughput change or reflection artefacts
Transmitted wavefront	Combined departure from the intended wavefront	Image degradation not explained by one surface alone
Cleanliness	Contamination at surfaces and interfaces	Scatter, haze or compromised bonded interfaces

Metrology belongs in this discussion before a tolerance is released. Check uncertainty, repeatability, calibration, reference artefacts, fixture influence and operator dependence. A tolerance narrower than the measurement system can reliably discriminate is not useful for an unqualified pass/fail decision. Improve the measurement or agree a defensible acceptance strategy; do not report additional digits as additional knowledge.

3. Surface generation, polishing and finishing: different errors, different consequences

In a conventional abrasive route, blank preparation establishes workable geometry; generating and grinding approach the required shape; finer processing prepares the surface for polishing. Polishing and finishing must then deliver the specified form and residual texture without unacceptable defects. SCHOTT describes several available optical processing routes 10; no single route should be assumed for all endoscope elements.

The transitions matter because each operation can leave a different error signature. A shape deviation, residual tool pattern and handling scratch require different investigations. Intermediate inspection helps determine when an error entered the route. Extending polishing is not a universal remedy: the result may alter form, edge behaviour or remaining thickness while addressing a different surface condition.

Three meanings hidden inside “surface quality”

Surface form is departure from the intended shape, commonly considered at lower spatial frequencies. ISO 10110-5:2026 2 covers its drawing indication. Surface texture describes residual structure after removal of form, including roughness and waviness at higher and intermediate spatial frequencies; ISO 10110-8:2019 5 provides the corresponding framework. Measurement bandwidth and detrending must be stated so that laboratories compare the same surface information.

Surface imperfections are localized defects, such as scratches, pits or edge chips, evaluated under the selected convention. ISO 10110-7:2017 4 addresses these indications. Neither a scratch assessment nor a roughness value describes the whole surface. A polished element can look clean and still fail its form requirement. Equally, a small visible imperfection does not automatically establish poor imaging: its location, extent, optical function and acceptance specification govern the decision.

Aspheric surfaces can offer additional aberration correction where the optical design benefits 9. They do not guarantee zero distortion or higher resolution. Their manufacture requires a suitable form-measurement strategy and control of the relationship between the aspheric surface and its reference axis. A good on-axis surface measurement does not remove the need to check how the element is centred and installed.

How manufacturing error can appear in the image

Possible contributors only — image appearance alone does not establish root cause. Eliminate target, fixture, coupler and camera effects before attributing a result to manufacturing.

Manufacturing / assembly contributor	Possible image behaviour	Engineering check
Decentration / tilt	Left-right field asymmetry	Component centration and assembled alignment
Surface form deviation	Resolution or contrast degradation	Form, wavefront where relevant, and assembled image test
Surface roughness	Scatter or veiling loss of contrast	Texture bandwidth and controlled scatter evaluation
Coating problem	Throughput change, reflection or ghost	Spectral measurement, interface and surface inspection
Contamination	Haze, flare or localized unwanted signal	Cleanliness and assembled optical inspection
Relay-stack error	Field inconsistency or optical-axis shift	Spacing, references and finished-scope evaluation

4. Centring, edging and spacing: where tolerance accumulation begins

Once the optical surfaces exist, their relationship to the locating geometry must be established. Centring connects an optical reference to the mechanical reference used during assembly; edging establishes the relevant perimeter. Diameter, edge thickness and seating geometry influence how the part is positioned, but an accurately measured diameter alone does not establish centration. General component relationships are discussed in 7.

ISO 10110-6:2025 3 covers indications for centring and tilt of elements, subassemblies and assemblies. That scope is useful: the reference question continues after individual lens inspection. Wedge, surface-axis relationships and tilt are related but not interchangeable descriptions. Specify which relationship is being controlled rather than treating “centred” as a property independent of a datum.

Along the shaft, lens thickness, spacers, mechanical seats and group positions contribute to axial relationships. Radial location, tilt and mechanical straightness contribute to alignment. Their optical effects depend on where they occur and on the sensitivity of the design. A stack can be dimensionally within its overall length limit while containing an unfavourable distribution of internal spacing errors.

TOLERANCE ACCUMULATION — CONCEPT ONLY

• Component A± form / location variation
• Component B± thickness / tilt variation
• Spacer± axial variation
• Mechanical seat± reference variation
• Relay group± alignment variation

↓ Combined through design sensitivity
Optical-axis / spacing deviations
Potential change in assembled image performance

Worst-case analysis asks what an allowed combination can produce at the limits. Statistical analysis asks how outcomes may be distributed under stated assumptions about variation and correlation. They are different questions. Shared tooling or a common reference bias can correlate errors, so an assumption of independent random contributions needs evidence. Neither approach justifies adding unlike errors without considering optical sensitivity.

An assembly adjustment may compensate for one combination while leaving another characteristic worse. Record which compensators are permitted and verify the resulting field, not only centre focus. If production depends on selective matching, that selection becomes part of the controlled process; it cannot remain an undocumented explanation for why a demonstration instrument performs well.

5. Coatings, cleanliness and optical interfaces

ISO 9211-2:2024 13 provides optical-coating specification context; ISO 9211-3:2024 14 addresses environmental durability categories. Neither establishes a finished endoscope’s validated reprocessing performance.

A coating specification concerns spectral band, incidence conditions, substrate and reflection behaviour, as well as durability appropriate to the application. Antireflection treatment can reduce reflections within its design conditions 11; it does not eliminate every reflection in a multi-surface assembly. A witness sample, a model and a measurement on the actual optic provide different evidence and should be identified as such.

For designs intended to transmit visible and NIR images, evaluate the required bands rather than relying on the word “broadband.” Coating behaviour contributes to the assembled spectral response, but glass, bonded interfaces, geometry and the rest of the imaging system also contribute. An NIR coating alone cannot guarantee fluorescence-system performance. The fluorescence imaging article provides the wider system context.

Cleanliness is an optical variable. Particles and films may scatter light, produce haze or affect an interface before bonding. Once enclosed, contamination can be difficult to locate without disturbing the assembly. Define the inspection condition after cleaning and protect it through transfer, joining and closure. Separate removable contamination from permanent surface damage before deciding whether to rework a part.

Where cemented or bonded optics are used, adhesive-layer thickness, voids, contamination, cure stress and axis relationships need appropriate control. Spectral transmission and thermal stability belong in the interface assessment. A passing lens inspection does not prove that bonding preserved the subassembly’s optical behaviour. These are engineering review topics, not a prescription for an adhesive, cure schedule or manufacturing recipe.

Manufacturing cleaning is also distinct from device reprocessing. A treatment suitable for a loose optical element cannot be transferred automatically to an assembled medical instrument. Evaluate the finished construction and validated instructions rather than extrapolating from an individual material’s resistance.

6. From individual lenses to a rigid-endoscope optical train

The assembly connects the objective, relay path and eyepiece to a common set of mechanical references. Tube straightness, seats, spacers, runout, orientation and joining stress can affect that relationship. The relay must transfer the image through the available optical apertures; its position cannot be assessed solely by whether the parts fit inside the shaft. Assembly and optical tolerancing are linked disciplines 8.

Consider a conceptual example: individually compliant relay elements are installed with a small group offset relative to the objective. Incoming reports can all pass while the finished instrument shows asymmetric off-axis performance. The example identifies a plausible interaction, not a diagnosis for any particular image. Repeat the measurement and check external alignment before attributing the effect to the relay.

The distal window is another optical and mechanical interface. Its form where relevant, reflection, contamination, joining or sealing, and mounting stress can contribute to assembled behaviour. Window material and construction are design choices; there is no assumption here that all rigid scopes use sapphire or one sealing arrangement. Inspect the relevant state after integration, not only before mounting.

COMPONENT → ASSEMBLY → IMAGE QUALITY

1. Component levelSurface form · centring · thickness · coating · cleanliness
2. Assembly levelSpacing · mechanical references · relay alignment · objective alignment
3. Finished-scope levelResolution · contrast · field consistency · distortion · transmission · stray light

A set of individually compliant lenses does not guarantee a compliant optical assembly. Check the optical axis and relevant field behaviour at useful assembly stages while the source of a change can still be isolated. Preserve orientation and configuration records. Replacing several groups together may restore a passing image, but it provides weak evidence about the actual process failure.

The finished scope is itself an optical system; it is also one component of a camera-based endoscopy system. Keep these boundaries explicit. A processor-enhanced display image cannot replace characterization of the optical train, while a bench result on the train cannot establish performance with every camera or coupler.

7. Manufacturing quality across the intended lifecycle

Room-temperature final inspection establishes an initial condition. Where retention of performance is a product requirement, compare critical characteristics before and after the intended validated reprocessing exposure. Thermal cycling, moisture, chemicals, handling and pressure where applicable may challenge different interfaces. Their relevance depends on the device and its validated processing method; not every rigid scope follows the same sterilization route.

The investigation can include seal integrity, distal-window condition, coating stability, bonded interfaces, internal contamination and optical-axis or focus drift. Use the same measurement configuration and a stable reference for the comparison. A changed field map merits investigation, but it does not identify a failed seal, adhesive or lens by itself. Record exposure and handling history with the observations.

Lifecycle verification is a construction-level question. A material certificate, coating sample or adhesive datasheet cannot establish retention of performance for the complete scope. A process change affecting a joint or reference surface may therefore require renewed assessment beyond a component reinspection. The verification plan must define the exposure and acceptance criteria; this article supplies neither cycle counts nor a universal durability claim.

Manufacturing defects can change the image delivered to an imaging system. Their clinical significance depends on the application and cannot be inferred from manufacturing measurements alone. Optical acceptance is not evidence of improved diagnosis, fewer complications or a better surgical outcome.

8. Inspect the assembled scope, not only the components

Organize inspection at three levels: component, optical subassembly and finished scope. Each level closes a different question. Component measurements address drawing characteristics; subassembly tests address joining, references and alignment; finished-scope tests establish the assembled optical behaviour under the agreed conditions. Traceability connects them without allowing one report to substitute for another. Follow the evidence from drawing through component manufacture, coating, centring and edging, cleaning, subassembly, relay alignment, mechanical integration and final optical inspection. The actual operation sequence is design-dependent.

Quality gates across the manufacturing chain

This is an engineering planning framework, not an internal SOP or a normative test sequence.

Stage	Typical question	Evidence
Optical drawing	Are critical characteristics allocated and inspectable?	Reviewed drawing, references and measurement plan
Incoming material	Does the material match the controlled design?	Material identity and relevant batch documentation
Generated / polished lens	Are form, texture, defects and dimensions acceptable?	Identified surface and dimensional inspection records
Coated optic	Is the required spectral behaviour supported?	Relevant coating measurements with sample and setup identity
Centred / edged optic	Do optical features relate correctly to locating geometry?	Centration, edge and datum-related measurements
Optical subassembly	Did positioning and joining preserve optical behaviour?	Alignment, spacing and wavefront evidence where relevant
Finished rigid scope	Does the assembled instrument meet optical requirements?	Validated finished-product optical test results

A selected excellent instrument demonstrates what that sample achieved, not the stability of future production. Examine lot variation and revision history alongside the measurement system. Statistical process control may be useful where measurement capability and production volume support it. Do not calculate impressive capability figures from mixed revisions, selected survivors or variation dominated by the fixture.

Close the feedback loop with observed failure patterns. If field asymmetry tracks a particular assembly stage, investigate that stage against its recorded references before tightening every lens tolerance. If measurements change with the operator, resolve the measurement problem before ranking suppliers. Corrective action should address the demonstrated contributor and be checked on representative subsequent work.

For finished-scope methods and the boundary between optical and camera-system results, see How to Evaluate Optical Performance in Rigid Endoscopes. Manufacturing evidence becomes useful when it explains whether the intended image survives the complete chain: drawing, part, interface, assembly and finished instrument.

Key takeaways

1. Derive manufacturing specifications from image-performance requirements and sensitivity.
2. Separate form, texture, imperfections and centration rather than assigning one “surface quality” number.
3. Control the tolerance and reference chain throughout the long optical assembly.
4. Treat component compliance as necessary evidence, not proof of finished-scope compliance.
5. Connect component metrology, assembly control and finished-scope measurements through traceable feedback.

References

Official records and manufacturer engineering references consulted on 6 October 2026. ISO 10110 supplies drawing conventions, not endoscope acceptance limits. The tables, examples and inspection framework above are engineering interpretations; the diagrams are conceptual only. No proprietary standard procedures or product prescriptions are reproduced.

1. ISO 10110-1:2019. Optics and photonics — Preparation of drawings for optical elements and systems — Part 1: General. Official ISO record.

2. ISO 10110-5:2026. Optics and photonics — Preparation of drawings for optical elements and systems — Part 5: Surface form tolerances. Official ISO record.

3. ISO 10110-6:2025. Optics and photonics — Preparation of drawings for optical elements and systems — Part 6: Centring and tilt tolerances. Official ISO record.

4. ISO 10110-7:2017. Optics and photonics — Preparation of drawings for optical elements and systems — Part 7: Surface imperfections. Official ISO record.

5. ISO 10110-8:2019. Optics and photonics — Preparation of drawings for optical elements and systems — Part 8: Surface texture. Official ISO record.

6. ISO 10110-14:2018. Optics and photonics — Preparation of drawings for optical elements and systems — Part 14: Wavefront deformation tolerance. Official ISO record.

7. Edmund Optics. Understanding Optical Specifications. Engineering reference.

8. Edmund Optics. Integration of Optical Systems. Engineering reference.

9. Edmund Optics. All About Aspheric Lenses. Engineering reference.

10. SCHOTT. Technical details of Optical Components. Manufacturer process overview.

11. Edmund Optics. Anti-Reflection (AR) Coatings. Engineering reference.

12. KARL STORZ. Telescopes. Manufacturer educational overview of the HOPKINS rod-lens system. Cited only as an architectural example, not as an opentic design or performance claim.

13. ISO 9211-2:2024. Optics and photonics — Optical coatings — Part 2: Optical properties. Official ISO record.

14. ISO 9211-3:2024. Optics and photonics — Optical coatings — Part 3: Environmental durability. Official ISO record.

Standards note. These editions are listed as published at the time of review. Formal drawing preparation and acceptance require the applicable full standards and validated procedures; this article is not a substitute for either.