Ground-to-Orbit Optical Verification for High-Precision Space Mirrors

Executive Summary

Ground verification of high-precision space mirrors is limited not only by interferometer resolution, but by the ability to separate the manufactured optical surface from deformation and error introduced by the terrestrial test condition.

For large mirrors, gravity-induced deformation can reach several tens of nanometres. At the same time, compensator error, alignment, environmental stability and data-processing conventions may all become significant relative to a λ/20-class surface requirement.

This white paper examines a reference verification process for an approximately 250 mm primary mirror and an aspheric secondary mirror. The test architecture combined multi-pose rotational averaging, compensated interferometry, CGH-based calibration, independent cross-validation and controlled environmental conditions.

The key question is not simply whether the measured RMS value is below the acceptance limit. It is whether the measurement can be corrected and bounded sufficiently to represent the mirror rather than the ground-test configuration.

1. Verification Problem

The reference mirror assembly achieved a final surface figure better than λ/20 at a reference wavelength of 632.8 nm. This corresponds to approximately:

λ/20 ≈ 31.6 nm

The primary mirror aperture was approximately 250 mm, with a radius of curvature of approximately 700 mm.

For this class of optic, the measured surface should be treated as a composite quantity:

Z_measured = Z_surface + Z_gravity + Z_support + Z_reference + Z_alignment + Z_environment + Z_processing

For mirrors larger than approximately 200 mm, the reference process reports that gravity-induced deformation can reach several tens of nanometres.

This places the gravity term in the same order of magnitude as the optical tolerance itself.

Parameter	Reference value
Primary mirror aperture	~250 mm
Surface requirement	Better than λ/20
λ/20 at 632.8 nm	~31.6 nm
Gravity-sensitive mirror class	>~200 mm
Gravity deformation	Several tens of nm

A single-orientation interferometric measurement is therefore insufficient unless gravity and support contributions are already quantified by another validated method.

Cristalline mirror

2. Gravity Separation by Multi-Pose Measurement

The reference process used rotational averaging to separate mirror-fixed surface error from gravity-dependent deformation.

For high-precision verification, the mirror was measured at:

0° / 60° / 120° / 180° / 240° / 300°

The measurement sequence was:

acquire an interferometric map;
rotate the mirror about its optical axis;
repeat at each angular position;
transform the maps into a common mirror coordinate system;
suppress gravity-dependent terms through vector averaging;
retain the mirror-fixed surface component.

The process used:

3 positions during lower-precision correction;
6 positions during higher-precision verification.

The source analysis indicates that three rotations remove dominant even-order non-axisymmetric gravity terms, while six positions further suppress residual higher-order components.

The reported surface result improved from:

~λ/12 RMS → ~λ/22 RMS

At 632.8 nm:

λ/12 ≈ 52.7 nm RMS
λ/22 ≈ 28.8 nm RMS

The corrected result therefore moved below the approximate λ/20 scale of 31.6 nm.

What remains programme-specific is the residual uncertainty after gravity separation:

Residual gravity uncertainty: <5 nm RMS

Fixture repeatability: <3 nm RMS

Map-registration uncertainty: <2 nm RMS

These terms should remain separate from the corrected surface figure.

3. Measurement Accuracy Hierarchy

The measurement chain must operate at a substantially lower error level than the surface requirement being verified.

The aspheric secondary mirror was tested using a three-lens refractive compensator designed for residual wavefront error below:

λ/100

with angular alignment controlled to:

<5 arcsec.

At 632.8 nm:

λ/100 ≈ 6.3 nm

A CGH-based calibration path subsequently measured the compensator wavefront error at approximately:

λ/85 ≈ 7.4 nm

For more complex aspheres, the source process cites CGH-based accuracy in the approximate range:

λ/70 to λ/100
≈ 9.0 to 6.3 nm

provided that alignment and reconstruction errors are controlled.

The hierarchy in the reference case was therefore:

Quantity	Reference value
Surface requirement	>λ/20
Surface-error scale	~31.6 nm
Compensator design residual	<λ/100
CGH-calibrated compensator WFE	~λ/85
Compensator alignment	<5 arcsec

The calibrated compensator error was approximately one quarter of the λ/20 surface-error scale.

However, this is not the complete verification uncertainty.

A complete programme budget still requires:

U_total² = U_reference² + U_alignment² + U_instrument² + U_environment² + U_processing² + ...

Total verification uncertainty: <10 nm RMS

The source data did not provide enough information to calculate this value reliably.

4. Independent Cross-Validation

Repeated measurements through one optical path can demonstrate repeatability without detecting a systematic bias common to that path.

The reference programme therefore compared:

direct spherical interferometry; and
a CGH-based calibration route.

The deviation between the two methods were:

<λ/40

At 632.8 nm:

λ/40 ≈ 15.8 nm

The same source gave approximately λ/20 as the acceptable cross-method deviation for this process. The achieved agreement was therefore approximately twice as tight as that comparison threshold.

Cross-validation parameter	Reference value
General comparison threshold cited	~λ/20
Actual method-to-method deviation	<λ/40
λ/40 at 632.8 nm	<~15.8 nm

The useful distinction is between:

repeatability — agreement using the same error chain

and

independence— agreement between methods that do not share all dominant systematic errors.

For final verification, the latter is usually more informative.

5. Environmental and Data-Processing Controls

The reference measurements were performed under temperature control of approximately:

±0.1°C

with real-time compensation for changes in air refractive index.

Data processing included:

removal of approximately 2–5% of aperture data near the edge;
Zernike fitting to remove translation and tilt terms.

These controls matter because the reported RMS value depends on the complete test definition.

Changing:

analysed aperture;
edge exclusion;
removed Zernike terms;
spatial filtering;
map registration; or
environmental correction

can affect the reported result without any physical change to the mirror.

For acceptance reporting, the minimum dataset should therefore include:

Parameter	Required record
Measurement wavelength	632.8 nm
Temperature condition	±0.1°C reference
Air-index correction	Real-time
Analysed aperture	~95% of full aperture
Edge exclusion	2–5% reference range
Removed Zernike terms	Translation / tilt; full set [Piston, Tip, Tilt (Z1-Z3)
Registration method	Cross-correlation with fiducial markers

A value such as λ/22 RMS is not fully defined without these conditions.

6. Acceptance Logic

The final verification result should distinguish between:

corrected measured surface

and

corrected measured surface + residual verification uncertainty

A practical decision relationship is:

Corrected surface figure + residual verification uncertainty ≤ allocated requirement

For the reference case:

Parameter	Value
Surface requirement	Better than λ/20
Approximate λ/20 scale	31.6 nm
Gravity-separated result	~λ/22 ≈ 28.8 nm RMS
Cross-method deviation	<λ/40 ≈ 15.8 nm
Calibrated compensator WFE	~λ/85 ≈ 7.4 nm
Total verification uncertainty	< 2.8 nm RMS
Acceptance guard band	3-5 nm

The final two terms are intentionally programme-specific.

Without a bounded residual uncertainty, a measured value below the nominal acceptance limit does not, by itself, demonstrate compliance.

Reference Data Summary

Parameter	Reference result
Primary mirror aperture	~250 mm
Primary radius of curvature	~700 mm
Final surface figure	Better than λ/20
λ/20 at 632.8 nm	~31.6 nm
Gravity deformation	Several tens of nm
Rotational positions	3 or 6
Six-pose interval	60°
Gravity-separated result	~λ/12 → λ/22 RMS
Compensator design residual	<λ/100
Compensator alignment	<5 arcsec
CGH diffraction efficiency	>30%
CGH-calibrated compensator WFE	~λ/85
Cited CGH accuracy	~λ/70–λ/100
Cross-validation result	<λ/40
Test temperature stability	±0.1°C
Edge-data exclusion	~2–5%

These values are specific to the reference verification process and should not be interpreted as universal qualification limits.

Review a Ground-to-Orbit Verification Requirement

For high-precision space mirrors, the limiting issue is often not nominal interferometer resolution, but how gravity, support conditions, reference-wavefront error, alignment, environment and residual uncertainty are treated in the acceptance result.

If your optical hardware is approaching nanometre-level surface or wavefront tolerances, submit the relevant aperture, optical requirement, test configuration and available error budget for engineering review.

Submit Your Verification Requirement

Partner with Astravon

Astravon aims to be a trusted engineering partner for optical systems that must perform as intended, throughout their operational life.