How to Measure Center Thickness (CT) Without Touching the Lens

The Physics of Non-Contact Metrology and Low Coherence Interferometry

In the hierarchy of optical tolerances, Center Thickness (CT) is often underestimated. While Surface Form Error (Power/Irregularity) gets the glory of Zernike polynomials, CT is the fundamental scalar that dictates the position of the nodal points. In multi-element assemblies-like Microscope Objectives or VR Pancake Lenses-a cumulative CT error of just 50 microns can act as a “Spacing Error,” inducing massive Spherical Aberration and shifting the Effective Focal Length (EFL) outside the compensator’s range.

Historically, CT was measured using Tactile Metrology: Hall-effect probes, LVDTs, or simple micrometers.

The Problem: Tactile measurement is destructive by nature.

1. Scratching: Touching the vertex of a polished asphere risks scratching the most critical optical zone.
2. Deformation: For soft materials (Contact Lenses, Intraocular Lenses, or thin VR plastics), the probe pressure ($>0.2N$) compresses the material, yielding a false “thinner” reading.
3. Coating Damage: Hard probes can crack sensitive Anti-Reflective (AR) dielectric stacks.

To solve this, the industry has migrated to Non-Contact Optical Metrology. This section explores the dominant physics behind this shift, specifically focusing on Low Coherence Interferometry (LCI).

The Optical Path Length (OPL) Equation

All optical non-contact thickness measurements are derived from the behavior of light passing through a medium. Unlike a caliper that measures physical distance ($d$), optical sensors measure Optical Path Length (OPL) or Time of Flight.

To derive physical thickness ($t$), we must solve the fundamental equation:

OPL = n_g × t

Where:

• t is the physical thickness.
• n_g is the Group Refractive Index of the material at the measurement wavelength.

The Critical Distinction:

Most optical designers work with the Phase Index (n). However, metrology sensors use broadband light pulses. The speed at which the “packet” of light travels is governed by the Group Index, not the Phase Index.

n_g = n – λ(dn/dλ)

For dispersive glasses (high Abbe number), the difference between n and n_g is negligible. For high-index polymers (Polycarbonate/EP), the difference is significant (>1%). Using the wrong index is the #1 source of error in non-contact CT measurement.

Low Coherence Interferometry (LCI)

The gold standard for non-contact thickness measurement is Low Coherence Interferometry (often called OCT in medical fields).

The Architecture:

LCI utilizes a Michelson Interferometer topology coupled with a broadband light source (Super-Luminescent Diode – SLD).

1. Source: Light is emitted (typically 1310nm or 850nm).
2. Splitter: A fiber coupler splits light into a Reference Arm (internal mirror) and a Sample Arm (the lens).
3. Reflection: Light reflects off the front surface (S1) and the back surface (S2) of the lens.
4. Interference: The reflected light recombines with the reference light.

The Coherence Gate:

Unlike laser interferometry (which has a long coherence length), an SLD has a very short coherence length (typically 10-20µm).

Interference fringes only occur when the optical path difference between the Reference Arm and the Sample Arm is near zero (within the coherence length).

By scanning the Reference Mirror (Time Domain LCI) or using a Spectrometer (Spectral Domain LCI), the system detects two distinct interference bursts:

• Peak 1: Reflection from Front Surface.
• Peak 2: Reflection from Back Surface.

The Calculation:

The system measures the distance between the two interference peaks in air (Δ_air).

t = Δ_air / n_g

Advantages of LCI:

• Sub-Micron Accuracy: Typical repeatability is ±0.1 µm.
• Insensitive to Vibration: Because the measurement is differential (S2 – S1), vibrations of the lens itself cancel out.

• Layer Discrimination: LCI can see inside a stack. It can measure the lens thickness, the adhesive gap, and the second lens thickness simultaneously, provided the refractive indices are different.

Time Domain vs. Spectral Domain LCI

1. Time Domain (TD-LCI)

• Mechanism: A physical mirror moves inside the sensor probe.
• Range: Can measure very thick lenses (up to 400mm) because the mechanical mirror can travel far.
• Speed: Slower (limited by mirror speed, ~10Hz – 50Hz).
• Use Case: Large telescope optics, thick prisms.

1. Spectral Domain (SD-LCI / FD-OCT)

• Mechanism: No moving parts. A spectrometer analyzes the spectral interference pattern (Fourier Transform).
• Range: Limited by spectrometer resolution (typically < 10mm – 20mm).
• Speed: Ultra-fast (> 20 kHz).
• Use Case: Contact lenses, VR lenses, mobile phone lenses.

Handling Anti-Reflective (AR) Coatings

A common challenge in LCI is the presence of coatings.

• Thin Films (< 1µm): Standard LCI cannot resolve the top and bottom of an AR coating. The coating appears as a single reflection. However, the coating induces a Phase Shift in the reflected wave.
• The Error: This phase shift can cause a “Virtual Peak Shift” of 10-50nm. For standard CT tolerancing (±10µm), this is negligible. For ultra-precision optics (EUV lithography), algorithms must correct for the Fresnel phase coefficients of the coating stack.

Geometry Constraints: The “Cosine Error”

LCI measures the thickness along the ray path.

To measure Center Thickness, the probe beam must be:

1. Centered: Hitting the exact vertex.
2. Normal: Perpendicular to the surface.

If the lens is de-centered by distance x, or tilted by angle θ:

• The beam measures a chord, not the true diameter (Sagitta error).
• The measured thickness increases due to the oblique path: t_measured = t_true / cos(θ’). (Where θ’ is the refracted angle inside the material).

Therefore, high-end LCI systems are integrated with Auto-Centering Stages. The system scans the lens to find the apex (peak point) before capturing the thickness value, ensuring the vector is perfectly normal to the surface.

Low Coherence Interferometry has revolutionized CT measurement by turning a mechanical problem into a frequency analysis problem. By decoupling the measurement from physical touch, it allows for 100% inspection of delicate surfaces.

However, LCI has a weakness: it requires knowledge of the refractive index. If you don’t know the material (e.g., a competitor’s lens), you cannot measure its physical thickness, only its optical thickness.

In the next section, we will examine Chromatic Confocal Sensing, a technology that competes with LCI and solves different problems, particularly for opaque and steep surfaces.

Chromatic Confocal Sensing and Laser Triangulation

In Part 1, we established Low Coherence Interferometry (LCI) as the interferometric gold standard for measuring transparent materials. However, the optical manufacturing floor is diverse. Engineers often encounter steep aspheres, rough surfaces, or opaque substrates where LCI struggles.

This section explores the two primary alternatives: Chromatic Confocal Sensing (CCS) and Laser Triangulation. We will dissect their operating principles, their material interactions, and how they handle the “Dual Surface” problem of Center Thickness.

Chromatic Confocal Sensing (CCS) Principle

Confocal microscopy is famous for its depth-sectioning ability. Chromatic Confocal technology adapts this for industrial metrology by removing the mechanical Z-scan and replacing it with “Color Coding.”

The Mechanism:

1. Source: A white light source (LED or Halogen) is launched into a fiber.
2. Hyper-Chromatic Lens: The sensor head contains a special lens with deliberately massive Longitudinal Chromatic Aberration.
   • Blue light (400nm) focuses close to the lens.
   • Red light (700nm) focuses far from the lens.
3. The Spectrum: The light focuses along a “Optical Pen” line.
4. Reflection: When a surface sits at a specific distance, only the wavelength that focuses exactly on that surface is reflected back into the fiber aperture (pinhole principle). All other wavelengths are out of focus and rejected.
5. Spectrometer: A detector analyzes the returning light. If it sees a peak at 550nm (Green), it knows the surface is at position Z = 5mm.

Measuring Thickness with CCS:

For a transparent lens, the CCS sensor detects two spectral peaks:

• Peak λ₁: Reflection from Surface 1.
• Peak λ₂: Reflection from Surface 2.

The system calculates the distance between the two focal points.

Correction: Like LCI, CCS requires Refractive Index correction. The light traveling to the second surface is slowed down by the lens material.

t = (Z₂ – Z₁) × CorrectionFactor(n, λ)

Advantages of CCS over LCI

While LCI is generally more accurate for thickness, CCS has distinct geometric advantages:

1. Numerical Aperture (NA): CCS sensors typically have a higher NA. They can accept light from steeper angles (up to ±40°). LCI sensors often fail if the surface slope exceeds ±5° (specular reflection is lost).
   • Application: Measuring the CT of high-curvature meniscus lenses or measuring thickness at the edge of a steep lens. 
2. Spot Size: CCS focuses light to a tiny spot (2µm – 5µm). LCI spot sizes are typically larger (20µm – 50µm). 
   • Application: Measuring micro-optics or diffractive structures.
3. Material Independence (Surface): Since CCS does not rely on interference fringes, it is robust against rougher surfaces (e.g., ground glass surfaces before polishing).

Laser Triangulation: The Differential Method

Laser Triangulation is the workhorse of general industrial metrology but is tricky for transparent optics.

Mechanism:

A laser beam hits the surface, and the diffuse (or specular) reflection is imaged onto a CMOS array at an angle. The position of the spot on the pixel array correlates to the height (Z).

The Challenge with Glass:

1. Transparency: The laser penetrates the glass. You get two spots (top and bottom).
2. Refraction: The beam path to the bottom surface is bent by the glass.
3. Specular Reflection: Triangulation usually relies on diffuse scattering. Polished lenses act as mirrors. Measuring them requires “Specular Mode” sensors where the receiver is at the exact angle of reflection (Angle of Incidence = Angle of Reflection).

Differential Thickness Measurement:

For opaque lenses (e.g., Infrared Germanium) or when high accuracy isn’t required, engineers use Opposing Sensors.

• Sensor A measures the position of the Top Surface (Z_A). Sensor B measures the position of the Bottom Surface (Z_B). The system is calibrated with a master gauge block of known thickness (T_master).
• Thickness = T_master + (Z_A – Z_A0) + (Z_B – Z_B0)

Pros/Cons:

• Pros: Does not require knowing the Refractive Index (since it measures the outer positions). Low cost.
• Cons: Requires mechanical alignment of two sensors (colinearity). Any vibration of the lens affects the reading (unlike single-ended LCI/CCS). Accuracy is typically ±1µm to ±5µm (worse than LCI).

Material Constraints: Absorption and Dispersion

Selecting the right sensor depends heavily on the lens material.

1. Infrared Optics (Germanium / Silicon)

• Problem: These are opaque to visible light. Standard CCS and LCI (at 850nm) cannot see the back surface.
• Solution: Interferometry at 3-5µm or 10.6µm. Or, use dual-head tactile/triangulation.
• Note: Silicon is transparent at 1310nm, so standard telecom-band LCI works well for Silicon optics (wafer level optics).

1. Colored Glass (Filter)

• Problem: A blue filter glass absorbs red light.
• Impact on CCS: If the lens is thick, the red wavelengths needed to measure the back surface might be absorbed.
• Solution: Use LCI (Infrared usually passes through colored glass) or check the transmission curve against the sensor range.

1. Step Height (Binary Optics)

• Sometimes “Center Thickness” involves a step (e.g., Fresnel lens center).
• CCS: Excellent for measuring the step height of the center zone.
• LCI: Can get confused by the phase jump of a sharp step.

Technology Selection Matrix

Feature	Low Coherence Interferometry (LCI)	Chromatic Confocal (CCS)	Laser Triangulation (Dual)
Primary Output	Optical Thickness ($n \cdot t$)	Optical Depth	Surface Position (Z)
Accuracy (Typical)	< 0.1 µm	0.2 µm – 1.0 µm	1.0 µm – 5.0 µm
Requires Refractive Index?	Yes	Yes (Single head)	No (Dual head)
Angular Acceptance	Low (Specular only)	High (±30° typical)	Medium
Measuring Speed	Ultra-Fast (>20kHz)	Fast (2-10kHz)	Fast
Cost	High	Medium-High	Low-Medium
Best For:	Precision Polished Lenses, Contact Lenses, Coating Stacks	Steep Aspheres, Micro-optics, Rougher surfaces	Opaque parts, General QA

 

While Low Coherence Interferometry rules the realm of ultra-precision and multi-layer stacks, Chromatic Confocal is the “All-Terrain Vehicle” of optical metrology-capable of handling steep slopes and rougher textures that would scatter an interferometer’s signal. Laser triangulation remains a niche tool for optics, mostly reserved for wafer thickness or opaque housing alignment.

In the final section, we will integrate these technologies into the production line. How do we measure CT automatically? How do we handle wet contact lenses? And what is the error budget analysis?

The Critical “Hidden” Variable: Calculating Group Index (n_g) from Sellmeier Coefficients

The most common source of error in non-contact thickness measurement is not the sensor; it is the user input. Most optical engineers are accustomed to working with the Phase Refractive Index (n_d) found in standard catalogs (like Schott or Ohara), which is typically defined at the Helium d-line (587.6 nm).

However, Low Coherence Interferometry (LCI) uses broadband light pulses (typically in the Infrared range, e.g., 1310 nm). The speed of this light “packet” is governed by the Group Index (n_g), not the Phase Index. Using standard n_d for an LCI measurement will result in a thickness error of 1% to 5%, which is catastrophic for precision optics.

The Practical Workflow: You do not need to perform complex calculus manually. Modern LCI software does not ask for the refractive index; it asks for the Dispersion Formula.

1. Open the glass datasheet (Schott, Ohara, CDGM).
2. Locate the Sellmeier Coefficients (often labeled as B₁, B₂, B₃ and C₁, C₂, C₃).
3. Input these coefficients directly into the metrology software.
4. The software calculates the exact Group Index for the specific center wavelength of the sensor probe.

The Physics: The relationship between Group Index and Phase Index is defined by the dispersion (slope) of the material:

n_g = n – λ · (dn / dλ)

To find n and dn/dλ, we use the Sellmeier Equation (Standard Form):

n²(λ) = 1 + [B₁·λ² / (λ² – C₁)] + [B₂·λ² / (λ² – C₂)] + [B₃·λ² / (λ² – C₃)]

Note: Ensure you check the units! Some catalogs use micrometers (µm) for λ, while others use nanometers (nm).

Table: The Cost of Using the Wrong Index (at λ = 1310 nm) The following table demonstrates the error magnitude if a user mistakenly inputs the standard Phase Index (n) instead of the correct Group Index (n_g) for a 5.000mm lens.

Material	Phase Index (n) @ 1310nm	Group Index (n_g) @ 1310nm	Difference (%)	Measurement Error on 5mm Lens
N-BK7 (Crown)	1.503	1.519	~ 1.06%	53 µm
N-SF11 (Flint)	1.748	1.792	~ 2.51%	125 µm
Polycarbonate	1.570	1.595	~ 1.59%	79 µm
Germanium	4.020	4.140	~ 2.98%	149 µm

 

Conclusion: As shown above, neglecting the Group Index calculation introduces errors (>50µm) that far exceed the typical manufacturing tolerance (± 10µm). Always use Sellmeier coefficients.

Applications, Automation, and Error Budgets

In Parts 1 and 2, we covered the physics of LCI and Chromatic Confocal sensors. In this final section, we move to the practical application. Measuring Center Thickness (CT) in a lab is one thing; measuring it inside a Diamond Turning Machine or on a high-speed Contact Lens production line is another.

This section covers advanced use cases, the specific challenges of “Wet Metrology,” and the mathematical error analysis required to certify a measurement.

Wet Metrology: The Contact Lens Challenge

Measuring the CT of a dry soft contact lens (hydrogel) is useless because the lens shrinks and warps when dry. It must be measured in a hydrated state (in saline solution). This presents a unique optical problem.

The Index Matching Problem:

• Refractive Index of Saline (n_saline) ≈ 1.335.
• Refractive Index of Hydrogel (n_lens) ≈ 1.38 – 1.42.
• Contrast: The difference (Δn) is very small (< 0.08).

Measurement with LCI: When using LCI, the reflection intensity (R) is governed by the Fresnel equation:

R = [(n₂ – n₁) / (n₂ + n₁)]²

Because Δn is tiny, the reflection from the lens surfaces is extremely weak (< 0.1%).

• Solution: High-sensitivity LCI sensors are required.
• Cuvette (Wet Cell): The lens is placed in a saline-filled cuvette.
• The Signal: The sensor sees 4 peaks:
  1. Air/Cuvette wall.
  2. Saline/Front Lens Surface (Weak).
  3. Back Lens Surface/Saline (Weak).
  4. Cuvette wall/Air.

The “Sag” Error:

In a wet cell, the lens floats. If it tilts, the measurement increases (Cosine error). Advanced systems use “Self-Centering” cuvettes or rapid scanning to find the true vertex minimum.

Measuring Multi-Layer Stacks (VR Pancake Lenses)

Modern VR optics are “Laminates” (Lens + Glue + Polarizer + Glue + Lens).

Measuring the Total Thickness (TT) is insufficient. Engineers need the thickness of each individual layer to control the optical path.

LCI Tomography:

LCI is uniquely suited for this. Because each interface (Glue/Plastic) creates a reflection peak, the LCI readout looks like an ultrasound scan (A-Scan).

• Peak A: Top Surface.
• Peak B: Interface Lens 1 / Glue.
• Peak C: Interface Glue / Lens 2.
• Peak D: Bottom Surface.

The Challenge: The refractive indices of the Glue (n=1.49) and the Lens (n=1.52) are often very close.

• Peak Overlap: If the glue layer is thin (< 10µm), Peaks B and C might merge.
• Resolution: High-end SD-OCT systems with broad bandwidth sources (> 100nm bandwidth) can achieve axial resolution of < 2µm, allowing separation of these thin adhesive layers.

Integration into Manufacturing (In-Situ Metrology) To achieve “Closed Loop” manufacturing, CT sensors are mounted directly onto the lathe or molding machine.

Diamond Turning Machines (DTM):

• Setup: A Chromatic Confocal probe is mounted on the tool post.
• Process: After cutting the front surface, the machine probes the center (Z_front). Then, it probes the fixture (Z_back).
• Feedback: The machine calculates the actual thickness remaining. If the lens is 5µm too thick, the machine automatically programs a “Correction Pass” to remove the excess material.
• Benefit: Zero handling errors. No need to unmount the lens for measurement (which loses centration).

Injection Molding:

• Setup: Sensors measure the mold cavity gap before injection, or the lens immediately upon ejection.
• Thermal Drift: The main error source is heat. A hot lens (80°C) has a different refractive index and thickness than a cold lens (20°C). The metrology system must apply a Thermal Expansion Coefficient (CTE) correction algorithm.

Integration into Manufacturing (In-Situ Metrology)

To achieve “Closed Loop” manufacturing, CT sensors are mounted directly onto the lathe or molding machine.

Diamond Turning Machines (DTM):

• Setup: A Chromatic Confocal probe is mounted on the tool post.
• Process: After cutting the front surface, the machine probes the center (Z_front). Then, it probes the fixture (Z_back).

• Feedback: The machine calculates the actual thickness remaining. If the lens is 5µm too thick, the machine automatically programs a “Correction Pass” to remove the excess material.
• Benefit: Zero handling errors. No need to unmount the lens for measurement (which loses centration).

Injection Molding:

• Setup: Sensors measure the mold cavity gap before injection, or the lens immediately upon ejection.
• Thermal Drift: The main error source is heat. A hot lens (80°C) has a different refractive index and thickness than a cold lens (20°C). The metrology system must apply a Thermal Expansion Coefficient (CTE) correction algorithm.

Beyond the Single Lens: Measuring Air Spacing in Optical Assemblies

While measuring the Center Thickness (CT) of a single component is critical for manufacturing, the most powerful application of Low Coherence Interferometry (LCI) lies in Assembly QA. Modern optical systems-such as microscope objectives, camera modules, and VR pancake stacks-are complex assemblies where multiple lenses are stacked with precision spacer rings.

The “See-Through” Capability: LCI sensors do not stop at the first surface. Due to the high dynamic range of the sensor, the probe beam penetrates the first element, travels through the air gap, and reflects off the subsequent elements. This allows the system to generate a “Tomographic” view of the entire optical stack in a single shot.

Why Air Spacing is Critical: In a fixed-barrel assembly, the position of every optical surface is cumulative.

• The Dependency: If Element #1 is manufactured 10 µm too thick, it will physically push into the space reserved for the air gap. Consequently, the Air Spacing to Element #2 will be 10 µm too thin (assuming the mechanical spacer is rigid).
• Optical Impact: An error in air spacing often degrades performance (specifically Spherical Aberration) more severely than an error in glass thickness.

Assembly QA with LCI: Instead of measuring components individually and hoping they fit, Rotlex systems allow for Final Assembly Verification.

1. Spacer Verification: Detect if a spacer ring is missing, dirty, or seated incorrectly (tilt).
2. Gap Analysis: Since the refractive index of air is known (n ≈ 1.00027), the measurement of the air gap is absolute and highly accurate.
3. Closed Loop Assembly: For high-end “Active Alignment” stations, the LCI sensor guides the robot to adjust the Z-position of the lens elements in real-time until the target air gap is achieved.

Limitations: When LCI Cannot Measure (Coatings and Material Absorption)

While Low Coherence Interferometry is robust, it relies on the fundamental principle of light transmission. There are two specific “Elephant in the Room” scenarios where this technology hits a physics wall.

1. Metallic Mirror Coatings

If an optical element features a metallic coating (e.g., Aluminum, Silver, or Chrome), such as in Beam Splitters or Fold Mirrors, the LCI sensor is effectively blinded.

• The Physics: Metals have a complex refractive index with a high extinction coefficient (k). They reflect nearly 100% of the light at the first surface and block transmission to the substrate.
• The Consequence: The sensor will detect the Position of the coated surface with high precision, but it cannot measure the Thickness of the glass beneath it. For these parts, measurement must be performed before coating, or using a Counter-Acting (Dual Head) Confocal setup.

2. IR-Absorbing Materials (Colored Filters)

Standard LCI sensors operate in the Near-Infrared (NIR) spectrum, typically at 1310 nm or 850 nm.

• The Trap: Certain colored filter glasses (e.g., Blue or Green ionically colored glass like Schott BG or KG series) are specifically designed to absorb Infrared radiation.
• The Failure: If the measuring wavelength falls within the absorption band of the filter, the light energy dissipates into heat before it can reflect off the back surface.
• Best Practice: Before purchasing an LCI system, always overlay the Transmission Curve of your specific glass material with the Center Wavelength of the sensor probe. If transmission at 1310 nm is < 5%, the measurement will likely fail.

Calibration Strategy: The Myth of the “Golden Lens”

A common question from QA Directors is: “How do I calibrate the machine? Can I just use a ‘Golden Lens’ that I measured with a micrometer?”

The Answer is No.

Attempting to calibrate an optical thickness sensor using a physical glass lens introduces a circular error. To know the optical thickness, you must know the Refractive Index ($n_g$). However, glass batches vary. If you rely on a glass reference standard, you are calibrating your machine against a variable material property, not a fixed physical constant.

The Solution: The Air Gap Etalon

To verify accuracy without the uncertainty of refractive index, metrology labs use a NIST-Traceable Air Gap Etalon.

• What is it? An assembly of two perfectly parallel optical flats separated by a precise, rigid spacer (e.g., Zerodur or Ceramic). 
• Why Air? The refractive index of air is a physical constant (n ≈ 1.00027 at standard temperature and pressure). It does not have “batch-to-batch” variation like glass. 
• The Procedure: The LCI measures the gap between the two plates. Since n is known perfectly, any deviation in the thickness reading is purely an instrument error, allowing for precise linearization and calibration of the sensor.

Beyond the Single Lens: Measuring Air Spacing in Optical Assemblies

While measuring the Center Thickness (CT) of a single component is critical for manufacturing, the most powerful application of Low Coherence Interferometry (LCI) lies in Assembly QA. Modern optical systems-such as microscope objectives, camera modules, and VR pancake stacks-are complex assemblies where multiple lenses are stacked with precision spacer rings.

The “See-Through” Capability: LCI sensors do not stop at the first surface. Due to the high dynamic range of the sensor, the probe beam penetrates the first element, travels through the air gap, and reflects off the subsequent elements. This allows the system to generate a “Tomographic” view of the entire optical stack in a single shot.

Why Air Spacing is Critical: In a fixed-barrel assembly, the position of every optical surface is cumulative.

• The Dependency: If Element #1 is manufactured 10 µm too thick, it will physically push into the space reserved for the air gap. Consequently, the Air Spacing to Element #2 will be 10 µm too thin (assuming the mechanical spacer is rigid).
• Optical Impact: An error in air spacing often degrades performance (specifically Spherical Aberration) more severely than an error in glass thickness.

Assembly QA with LCI: Instead of measuring components individually and hoping they fit, Rotlex systems allow for Final Assembly Verification.

1. Spacer Verification: Detect if a spacer ring is missing, dirty, or seated incorrectly (tilt).
2. Gap Analysis: Since the refractive index of air is known (n ≈ 1.00027), the measurement of the air gap is absolute and highly accurate.
3. Closed Loop Assembly: For high-end “Active Alignment” stations, the LCI sensor guides the robot to adjust the Z-position of the lens elements in real-time until the target air gap is achieved.

Error Budget Analysis

When certifying a CT measurement of 2.000mm ± 0.005mm, where do the errors come from?

1. Refractive Index Uncertainty (The Killer)

This is the dominant error term.

• Formula: t = OPL / n_g
• Differentiation: dt/t = -dn/n
• Example: If n_g has an error of 0.1% (0.0015), the thickness result has an error of 0.1%.
• For a 5mm lens, 0.1% error = 5µm error.
• Mitigation: Verify the raw material batch index using a refractometer before LCI measurement.

2. Cosine Error (Alignment)

Measuring off-vertex or at an angle.

• Error ≈ t × (1 – 1/cos(θ))
• A 2° tilt on a 5mm lens introduces ~3µm error.
• Mitigation: Auto-collimation or peak-search algorithms.

3. Linearity and Calibration

The sensor itself must be linear.

• LCI relies on the spectrometer wavelength calibration.
• CCS relies on the chromatic aberration curve.
• Mitigation: Calibration against a NIST-traceable Step Height Standard or Gauge Block.

Typical Error Budget (Example for 5mm Lens)

Error Source	Probability Distribution	Magnitude (µm)	Contribution to Uncertainty
Sensor Repeatability	Normal (Gaussian)	0.05	Low
Refractive Index ($n_g$)	Rectangular	3.00	High (Dominant)
Cosine Error (Centration)	Rectangular	0.50	Medium
Thermal Expansion	Triangular	0.20	Low
Vibration/Noise	Normal	0.10	Low
Total Uncertainty (k=2)	–	~3.2 µm	–

 

Note: Without precise index control, achieving tolerances tighter than ±5µm with optical non-contact methods is physically impossible.

Measuring Center Thickness without touching the lens is a triumph of modern photonics. It enables the production of scratch-free, high-precision, and complex optical stacks that define the future of VR and medical optics.

However, the shift from Tactile to Optical metrology transfers the burden of accuracy from Mechanics to Material Science. The metrologist can no longer treat the lens as a “black box”; they must know its Group Index, its thermal state, and its layer structure.

For the highest precision, Low Coherence Interferometry stands alone, offering the ability to slice through transparency with light itself, turning the elusive dimension of “Thickness” into a precise spectral signature.

Disclaimer: 

This document is intended for educational use only. It does not represent legal, regulatory, or certification advice, and should not be interpreted as a declaration of compliance or approval by Rotlex or any regulatory authority.