Precision Polishing: How Surface Texture and Scatter Affect an Optical System

Begin with the system requirement

The same lens can be acceptable in one instrument and unsuitable in another. Begin with wavelength band, aperture, field, illumination, detector or eye position, expected contrast and environmental conditions. A projection lens may be limited by unwanted light in nominally dark regions; a machine-vision lens may be judged by contrast at a particular object feature size. A high-power laser optic raises a different set of loss and damage questions. The polishing specification should follow these sensitivities, rather than a universally quoted surface-finish grade.

Separate requirements for transmitted wavefront, surface form, local slope, mid-spatial-frequency texture, microroughness and discrete defects. These are related but not interchangeable. A surface can have a low average roughness and still contain zonal ripple, scratches or a form error that matters to the application. Conversely, demanding an extremely low roughness everywhere may add cost without measurable system benefit.

Transfer grinding stock into a polishing plan

A usable fine-ground preform must leave enough material for polishing to remove relevant damage and correct geometry, but not so much that polishing time and form drift become uncontrolled. Material, curvature, diameter, previous abrasive, fixture and selected polishing route determine the allowance. There is no responsible universal removal depth. The handoff drawing should state center thickness, form envelope and inspection aperture with an unambiguous datum.

Track removal and condition over a representative set of parts. A first article with a good interferogram says little about repeatability if tool wear, pad condition or slurry changes across a lot. Inspection before and after polishing helps locate where a defect entered the process. Once coating or assembly is complete, that history can be hard to recover.

Form, waviness and roughness occupy different scales

Low-spatial-frequency form error changes focus and wavefront. Mid-spatial-frequency ripple can redistribute energy around an image or projected feature. Higher-frequency texture is associated with wider-angle scatter. The boundaries depend on aperture, wavelength and measurement setup. A single root-mean-square roughness value therefore hides where the error occurs in spatial frequency. Where the optical risk warrants it, specify a measurement bandwidth or power spectral density in addition to a scalar statistic.

Manufacturing controls also differ. Full-aperture form may respond to dwell and tool geometry, while local ripple can arise from sub-aperture tool paths, polishing pressure or periodic machine motion. Abrasive condition and cleaning can influence fine texture and isolated defects. If one process adjustment improves roughness but worsens form, a buyer needs both data sets before accepting the change.

Interpret scatter without overclaiming MTF

Surface texture can redirect light away from its intended path. Under restrictive assumptions for smooth surfaces, scatter increases with roughness relative to wavelength; that approximation is not a universal prediction for coated, curved, contaminated or complex assemblies. The angular distribution matters: light scattered into a critical dark zone has a different consequence from light sent outside the useful field. Measure the relevant region instead of treating total scatter as a complete answer.

Scatter may create veiling glare and lower practical scene contrast, but it is not the only cause of MTF loss. Aberrations, focus, alignment, diffraction, detector sampling and image processing also contribute. A valid comparison holds wavelength, field, aperture, focus and test method constant, then compares the polished element and the assembled system. A blanket claim that a lower roughness number will automatically raise MTF is not supported.

Measure the surface and the resulting beam

An optical profilometer or another qualified texture method should report calibration, scan area, lateral resolution, filtering and sampling locations. Interferometry or suitable form metrology addresses different error scales. Scratch and pit inspection needs a stated illumination and acceptance method. Repeated measurements at center and edge can expose nonuniform processing that one center-point number misses. The chosen instruments must resolve the specification being enforced.

When scatter is system-critical, add a defined optical test: for example angle-resolved scatter, a dark-field contrast measurement or a calibrated projection test at the operating wavelength. Record cleaning state, coating state and mount condition. Otherwise a low substrate-roughness result might be mistaken for evidence that the finished lens has low ghosting or low stray light. A coating can change both reflection and scatter, so substrate and final-assembly results should not be conflated.

Reliability includes what happens after polishing

A polished surface may later be cleaned, coated, cemented or mounted. Each step can reveal damage or add contamination, stress and interface reflections. The qualification plan should therefore include the finished configuration and relevant temperature, humidity, vibration or cleaning exposure. For a laser application, absorption and damage testing at the intended wavelength and fluence may be necessary; a camera objective may prioritize field contrast and environmental focus shift instead.

Do not infer service life or production yield from a visually smooth sample. Compare pilot and production lots using the same measurement rules, record rework and rejection causes, and keep the material and process history traceable. The goal is not an impressive single-number finish but a stable, inspectable output that the next process can use.

A useful supplier-review checklist

Ask for the measured wavelength and clear aperture; the form and roughness limits with their spatial bandwidths; the target removal from grinding; and the handling of scratch, pit and edge damage. Request sample metrology reports with calibration and repeatability information, plus the method used to correlate component data with final-system contrast or scatter. A claim of a specific nanometer roughness is incomplete without its instrument and filtering conditions.

Also distinguish a prototype capability from demonstrated lot stability. Ask what happens when a process drifts, which data are shared with the lens integrator, and whether the acceptance plan changes after coating or assembly. These questions support a realistic comparison of price, schedule and risk without assuming every supplier follows the same route.

Define a project inquiry with ALPHA OPTIK

Provide drawings, material and coating requirements, wavelength, aperture, target form and texture bands, annual volume, environmental envelope and the optical acceptance test. ALPHA OPTIK can evaluate a proposed scope against those inputs and confirm feasible manufacturing and verification arrangements for that project. This invitation is not a statement that any named equipment, numerical tolerance, certification or mass-production result has already been demonstrated.