Skip to content
TheChatArchiveThe Chat Archive
gpt-4.1-mini-2025-04-14· OpenAIAi2 WildChat-4.8M (AI2)

Chapter 1: Introduction The pursuit of higher precision and performance in modern…

  1. User

    Chapter 1: Introduction

    The pursuit of higher precision and performance in modern optical systems places increasingly stringent demands on the quality and reliability of optical components. Subsurface damage (SSD), introduced during the early stages of grinding and polishing, can significantly affect optical performance, especially in high-energy laser or space-based applications. This thesis explores the use of the Calotte Wear Test (Kalottenverschleißtest) combined with advanced image evaluation techniques to analyze and optimize SSD assessment in various optical materials.

    By combining cutting-edge digital analysis with well-established experimental methods, it seeks to provide fresh objectivity and clarity to the evaluation of SSD in optical materials. The work is grounded in the context of the MoKI-Pro project—a collaborative initiative focused on advancing the digitalization, characterization, and modeling of optical manufacturing processes ([MoKI-Pro Project Factsheet, pdf]). By leveraging both hands-on experimentation and data-driven methods, this research seeks to bridge the gap between traditional craftsmanship and the demands of next-generation optical engineering.

    1.1 Background and Motivation

    For decades, manufacturers have known that the processes used to shape and finish glass—grinding, lapping, and polishing—leave behind a legacy of hidden damage. These microcracks and stress zones, invisible to the naked eye, can act as seeds for future failure or as traps for contaminants that degrade optical performance (Cui et al., 2023; Suratwala et al., 2005). The challenge is not just to make optics that look perfect, but to ensure that they are truly robust beneath the surface.

    The consequences of SSD are far-reaching. In high-power laser systems, SSD can lower the laser-induced damage threshold (LIDT), making components more susceptible to sudden failure. In precision measurement and imaging, SSD can scatter light, introduce noise, or cause slow degradation over time. For manufacturers, undetected SSD means higher rejection rates, costly rework, and the risk of field failures that can damage reputations and bottom lines.

    Recognizing these risks, the optics community has invested heavily in both understanding and mitigating SSD. Early efforts focused on improving polishing techniques and developing destructive tests—like taper polishing and the Kalottentest—to reveal hidden flaws (Zhou et al., 1994). While these methods provided valuable insights, they were slow, labor-intensive, and often subjective, relying on the skill and judgment of experienced operators.

    By bringing together research institutions, industry partners, and advanced data analysis tools, MoKI-Pro aims to digitize and model every stage of the optical manufacturing process—from grinding to final inspection ([MoKI-Pro Project Factsheet, pdf]). This includes the integration of real-time monitoring techniques, such as Acoustic Emission (AE) and force measurement, to catch the earliest signs of damage as they occur ([MoKI-Pro Project, AE and Force Monitoring, pdf]). The project’s vision is to create a feedback loop where data from every step informs the next, enabling smarter, more sustainable production and higher-quality optics.

    By combining the Calotte Wear Test (Kalottentest) with advanced digital processing, this thesis seeks to move beyond the limitations of manual inspection and bring new precision to the detection and measurement of SSD. The ultimate goal is not just to understand SSD, but to provide practical tools and insights that can help manufacturers and researchers produce better, more reliable optical components for the technologies of tomorrow.

    1.2 Problem Statement and Research Objectives

    The challenge of subsurface damage (SSD) in optical materials is not simply a technical inconvenience—it is a persistent barrier to progress in high-precision optics. Despite decades of research and incremental improvements in fabrication and inspection, SSD remains a persistent and costly problem for both industry and science. The complexity of SSD, its hidden nature, and its profound impact on component performance demand a rigorous, objective, and scalable approach to detection and analysis.

    Problem Statement

    At the core of this thesis is the recognition that traditional methods for SSD evaluation are no longer sufficient for the demands of modern optical manufacturing. Manual inspection and destructive testing, such as taper polishing and the Kalottentest, have long been the standard for revealing and measuring SSD (Zhou et al., 1994). However, these methods are inherently limited by several factors:

    Subjectivity and Operator Bias: Manual interpretation of SSD features is prone to inconsistency, especially when dealing with subtle or overlapping damage zones. The same image may yield different results depending on the experience and judgment of the operator.
 Time and Resource Intensity: Destructive methods are labor-intensive, slow, and often require the sacrifice of valuable components. This makes them unsuitable for in-line quality control or high-throughput environments.
 Artifact Interference: The presence of periodic artifacts—such as scratches introduced during the Kalottentest—can obscure critical regions of interest, complicating both qualitative and quantitative assessment ([Kalottentest Report: N-BK7, Fused Silica, Zerodur, microscope images]).
 Lack of Integration with Modern Data Analysis: Traditional approaches do not leverage the power of digital image processing, automation, or data-driven analysis, limiting their scalability and reproducibility.

    These limitations are not merely academic. In practice, they translate into increased costs, reduced yield, and compromised performance for optical systems. The inability to reliably detect and quantify SSD can result in premature component failure, degraded system performance, and lost opportunities for innovation.

    Research Objectives

    This thesis is motivated by the need to overcome these limitations and to provide a foundation for the next generation of SSD evaluation in optical materials. The research is guided by the following objectives:

    To develop a robust, semi-automated pipeline for the quantitative analysis of SSD in optical materials.
The pipeline should combine the experimental rigor of the Kalottentest with the objectivity and efficiency of advanced image processing, reducing operator bias and improving measurement accuracy.
 To implement and validate key computational techniques—specifically, automatic circle detection and FFT-based artifact suppression—that address the primary challenges of calotte image analysis.
These methods should reliably identify calotte boundaries and remove periodic artifacts, enabling clear and reproducible measurement of SSD features. (GeeksforGeeks, 2023 ; Elhossini et al., 2023).
 To apply the developed pipeline to a range of optical materials (N-BK7, Fused Silica, Zerodur, CaF₂, ULE) and processing conditions, generating a comprehensive dataset of SSD metrics.
The analysis should reveal material-specific trends in SSD depth, area, and distribution, and provide insights into the effects of grinding, polishing, and etching parameters.
 To benchmark the performance of the automated pipeline against traditional manual methods and published literature, demonstrating its advantages in accuracy, reproducibility, and scalability.
The results should establish the pipeline as a viable tool for both research and industrial quality control (Yang, 2005 ; Cui et al., 2023 ).
 To identify the limitations of the current approach and outline future directions for further automation, integration with 3D data, and the application of machine learning. This forward-looking perspective ensures that the work remains relevant as the field continues to evolve.

    The objectives outlined above are not just technical milestones—they are a response to the real needs of the optics community. By moving beyond manual, subjective analysis and embracing automation and digital workflows, this thesis aims to set a new standard for SSD evaluation. The hope is that the methods and insights developed here will not only advance the state of the art in research, but also provide practical tools for manufacturers striving to meet ever-higher standards of quality and reliability.

    1.3 Structure of the Thesis

    This thesis is structured to guide the reader from foundational context to advanced analysis and practical outcomes. Chapter 1: Introduction establishes the importance of subsurface damage (SSD) in optical materials, outlines the motivation for this research, and defines the study’s objectives. Chapter 2: Literature Review and Fundamentals provides an overview of SSD mechanisms, reviews traditional and modern detection methods, and highlights the limitations of current approaches. Chapter 3: Experimental Methods details the preparation of optical samples, the Kalottentest procedure, chemical etching, and the acquisition of high-resolution images. Chapter 4: Image Analysis Pipeline adn Techniques describes the development and implementation of automated techniques—specifically, automatic circle detection and FFT-based artifact suppression—for objective SSD evaluation. Chapter 5: Results presents both qualitative and quantitative findings, comparing the performance of the automated pipeline across different materials and processing conditions. Chapter 6: Discussion interprets the results in the context of optical manufacturing, addresses methodological advances, and explores future research directions. Finally, Chapter 7: Conclusion summarizes the main contributions and offers recommendations for further work in SSD assessment and process optimization. Chapter 2: Literature Review and Fundamentals

    2.1 Subsurface Damage in Optical Materials

    Subsurface damage (SSD) refers to the microstructural defects—such as microcracks, dislocations, and residual stress zones—that form beneath the surface of optical materials during mechanical processing steps like grinding, lapping, and polishing. These defects are not visible on the surface and can only be revealed through specialized techniques such as chemical etching or cross-section microscopy, but can significantly degrade the mechanical strength, laser-induced damage threshold (LIDT), and long-term stability of optical components, especially in high-precision applications such as laser systems, space telescopes, and semiconductor lithography optics (Cui et al., 2023).

    The formation of SSD is governed by the fracture mechanics of brittle materials. When abrasive particles interact with the surface under load, they induce plastic deformation and crack propagation. The resulting damage typically includes lateral cracks, radial–median cracks, and trailing indent fractures, which together form a complex network of subsurface flaws (Suratwala et al., 2005). The depth and morphology of SSD are influenced by several factors, including abrasive size, load, feed rate, and the mechanical properties of the substrate (Yang, 2005; Lambropoulos et al., 2002).

    SSD is often categorized into three layers: a surface redeposition layer, a defect layer, and a deformation layer (Zhou et al., 1994). The defect layer, which contains the most critical damage, is the primary focus of SSD evaluation. Accurate detection and quantification of SSD are essential for optimizing post-processing steps and ensuring the reliability of optical components.

    Destructive methods such as taper polishing, ball-dimple (Kalottentest), and cross-sectional microscopy have traditionally been used to evaluate SSD. These methods, while accurate, are time-consuming and not suitable for in-line inspection. Non-destructive techniques such as full-field optical coherence tomography (FF-OCT), scanning acoustic microscopy (SAM), and laser scattering are emerging as promising alternatives, though they still face challenges in resolution and material compatibility.

    2.2 Overview of SSD Detection Methods

    The detection and quantification of subsurface damage (SSD) in optical materials is a cornerstone of quality assurance and process optimization in both research and industry. Over the years, a variety of methods have been developed to reveal, measure, and analyze SSD, each with its own strengths, limitations, and domains of applicability. This section provides a comprehensive overview of the principal SSD detection techniques, tracing their evolution from traditional destructive approaches to the latest advances in non-destructive and automated analysis. 2.2.1 Destructive Methods

    Historically, the most widely used methods for SSD detection have been destructive in nature. These techniques physically alter or remove material from the optical component to expose the hidden damage beneath the surface. Among these, the most prominent are:

    Taper Polishing: This method involves polishing a shallow wedge or taper into the surface of the sample, gradually exposing deeper layers. The depth and density of SSD are then assessed by optical or electron microscopy along the taper (Suratwala et al., 2005). While highly accurate, taper polishing is labor-intensive, time-consuming, and unsuitable for routine or in-line inspection.

    Ball Dimple Method (Kalottenverschleißtest): This technique creates a controlled spherical dimple on the surface using a rotating ball and abrasive slurry, followed by chemical etching to reveal SSD (Zhou et al., 1994). The Kalottentest is valued for its reproducibility and ability to compare SSD across different materials and processing conditions, but it still requires the sacrifice of a portion of the sample.

    Cross-Sectional Microscopy: Here, the sample is physically sectioned and the cross-section is examined under a microscope. This method provides direct visualization of SSD depth and morphology but is inherently destructive and often impractical for valuable or large components.

    2.2.2 Non-Destructive Methods

    The limitations of destructive methods have spurred the development of non-destructive SSD detection techniques, which aim to preserve the integrity of the sample while providing reliable information about subsurface flaws. Key non-destructive approaches include:

    Full-Field Optical Coherence Tomography (FF-OCT): FF-OCT uses low-coherence interferometry to generate high-resolution, three-dimensional images of the subsurface structure (Frank et al., 2021). This method is particularly effective for detecting SSD in transparent materials and can be applied in a non-contact, in-situ manner. However, its resolution and penetration depth are limited by the optical properties of the sample.

    Scanning Acoustic Microscopy (SAM): SAM employs high-frequency ultrasound to probe the subsurface, detecting variations in acoustic impedance caused by cracks or voids. While SAM can reveal SSD in a range of materials, its spatial resolution is generally lower than that of optical methods, and it may struggle with highly attenuating or scattering samples (Cui et al., 2023).

    White Light Interferometry: White light interferometry is a non-contact, high-resolution technique that has become a mainstay in surface and subsurface analysis for optical materials. Unlike monochromatic interferometers, which use a single wavelength, white light interferometers employ a broad spectrum of light. This allows them to resolve ambiguities in surface height and to measure step heights, roughness, and thin film thicknesses with nanometer-scale precision. 2.2.3 Hybrid and Emerging Approaches

    Recent years have seen the emergence of hybrid methods that combine the strengths of destructive and non-destructive techniques, as well as the integration of advanced computational tools: Chemical Etching Combined with Digital Image Analysis: Chemical etching, often with hydrofluoric acid (HF), remains a mainstay for revealing SSD. However, the analysis of etched features has shifted from manual inspection to automated, quantitative image processing. Techniques such as automated binarization, Fast Fourier Transform (FFT) filtering, and geometric measurement algorithms now enable objective and reproducible SSD evaluation. (GeeksforGeeks, 2023; Elhossini et al., 2023)

    Machine Learning and Pattern Recognition: Although still in the early stages of adoption, machine learning algorithms are being explored for the automated classification and quantification of SSD features in large datasets. These approaches promise to further reduce operator bias and enable real-time, in-line quality control.

    2.2 The Ball Dimple Method (Kalottenverschleißtest)

    The Ball Dimple Method, known in German as the Kalottenverschleißtest, is a widely used technique for evaluating SSD in optical materials. In the context of this thesis, the Kalottentest is not only a methodological choice but also a central requirement of the MoKI-Pro project—a collaborative research initiative focused on advancing the understanding and control of SSD in precision optics manufacturing.

    Principle and Procedure of the Kalottentest

    At its core, the Kalottentest involves the controlled abrasion of an optical surface using a rotating ball—typically made of steel or tungsten carbide—in the presence of an abrasive slurry (Zhou et al., 1994). The ball is pressed against the sample under a defined load and rotated at a set speed, gradually creating a spherical dimple (calotte) on the surface. The process is performed incrementally, with slurries of decreasing particle size applied in sequence to achieve the desired calotte depth and diameter. After the dimple is formed, the sample is chemically etched—most commonly with hydrofluoric acid (HF)—to reveal the SSD features beneath the surface.

    However, the method is not without limitations. It is inherently destructive, requiring the sacrifice of a portion of the sample. The accuracy of SSD depth measurements depends on precise control of the ball diameter, load, slurry composition, and etching conditions. Over-etching can obscure the true extent of SSD, while under-etching may leave critical features hidden (Zhou et al., 1994). Additionally, periodic artifacts such as scratches from the ball can complicate image analysis, necessitating advanced computational techniques for artifact suppression.

    Integration with Digital Image Analysis

    A significant advance in recent years has been the integration of the Kalottentest with automated digital image analysis. By combining the experimental rigor of the Kalottentest with techniques such as automated binarization, Fast Fourier Transform (FFT) filtering, and geometric measurement algorithms, researchers can now achieve objective, reproducible, and scalable SSD evaluation. (GeeksforGeeks, 2023; Elhossini et al., 2023)

    2.3 Magnetorheological Finishing (MRF)

    Magnetorheological Finishing (MRF) is a deterministic, non-contact polishing technique that uses a magnetically controlled fluid to remove material from optical surfaces. The fluid, typically composed of carbonyl iron particles suspended in a carrier liquid, changes its viscosity in response to a magnetic field, allowing for precise control over the polishing process.

    MRF is particularly effective in reducing or eliminating SSD introduced during earlier grinding steps. Unlike conventional polishing, which may leave residual microcracks, MRF can achieve sub-micron surface roughness and remove SSD tlayers with minimal additional damage (Jung et al., 2020). The process is especially valuable for high-performance optics where LIDT and surface figure accuracy are critical.

    In the context of the MoKI-pro project, MRF spots created on Q22-X machine by QED Technologies, were used as a reference for evaluating the effectiveness of the Kalottentest. The comparison between spherical polishing structures and MRF spots provided insights into the geometric and material-specific behavior of SSD. The use of 3D surface relief profiles, obtained via confocal microscopy or MarForm MFU 200, further enhanced the analysis by enabling the superposition of height maps (z-maps) with 2D image data (Prof. Dr. Fuetterer, MoKI-pro Guidelines, 2023).

    2.4 Image Processing in SSD analysis

    The rise of digital image processing has fundamentally changed how researchers and engineers approach the analysis of subsurface damage (SSD) in optical materials. Where once the evaluation of SSD relied on subjective visual inspection and manual measurement, today’s digital image processing techniques offer a pathway to objective, reproducible, and high-throughput analysis.

    Microscopic images of etched calottes, MRF spots, or cross-sections contain a wealth of information about the presence, depth, and morphology of SSD. However, these images are often affected by noise, uneven illumination, and artifacts such as scratches or debris. Manual interpretation is not only time-consuming but also prone to operator bias and inconsistency, especially when features are subtle or overlapping. Digital image processing provides a systematic approach to overcoming these challenges.

    Core steps in the SSD Image Analysis Workflow

    A typical SSD image analysis pipeline consists of several key stages, each designed to address specific challenges in the data:

    Preprocessing:
The first step is to prepare the raw image for analysis. This may involve converting the image to grayscale, applying noise reduction filters (such as Gaussian or median filtering), and enhancing contrast to improve the visibility of SSD features (OpenCV Documentation). Preprocessing is essential for minimizing the impact of background variations and ensuring that subsequent steps operate on clean, high-quality data.
 Segmentation:
The next stage is to separate SSD features from the background. Automated thresholding methods, such as Otsu’s method, are commonly used to binarize the image, distinguishing regions of interest from undamaged areas (GeeksforGeeks, 2023). Morphological operations (e.g., opening, closing) may be applied to remove small artifacts and fill gaps in the segmented regions.
 Artifact Suppression:
Periodic artifacts, such as scratches introduced during the Kalottentest, can obscure critical SSD features. Frequency-domain filtering, particularly Fast Fourier Transform (FFT) analysis, is employed to identify and suppress these artifacts. By detecting and zeroing out dominant frequency peaks, FFT filtering reveals the underlying SSD with greater clarity (Elhossini et al., 2023; [Kalottentest Report: N-BK7, Fused Silica, Zerodur, code and images]).
 Feature Extraction and Measurement:
Once the SSD regions are isolated, geometric analysis tools are used to quantify key metrics such as SSD depth, area, and distribution. Automatic circle detection algorithms, for example, can identify the boundaries of calottes even in noisy or low-contrast images, enabling precise measurement of diameters and, by extension, SSD depth ([GeeksforGeeks, 2023]).

    However, challenges remain. The quality of the analysis is still dependent on the quality of the input images; poor contrast, excessive noise, or severe artifacts can limit the effectiveness of even the most sophisticated algorithms. Additionally, the interpretation of complex SSD morphologies may require the integration of multiple image processing techniques or the development of custom tools tailored to specific experimental setups.

    2.5 Advances in Automated and Frequency-Domain Methods

    This section explores the key developments in automated and frequency-domain methods, with a focus on their practical impact and the challenges they address in the context of SSD analysis.

    The Shift Towards Automation

    Traditional SSD analysis relied heavily on manual inspection and measurement, which, as discussed earlier, is inherently subjective and prone to inconsistency. The introduction of automated image analysis pipelines has fundamentally changed this landscape. By leveraging open-source libraries such as OpenCV, NumPy, and SciPy, researchers can now design workflows that preprocess, segment, and quantify SSD features with minimal human intervention.

    A central advance in automation is the use of automatic circle detection algorithms. These methods, often based on the Hough Transform or advanced contour analysis, can reliably identify the boundaries of calottes or other circular features even in the presence of noise, blurring, or partial occlusion (Elhossini et al., 2023). In the context of SSD, this means that the critical measurement of calotte diameter—and thus SSD depth—can be performed objectively and reproducibly across large datasets.

    Frequency-Domain Analysis: Fast Fourier Transform

    While spatial-domain techniques are effective for many tasks, they often struggle with periodic artifacts such as scratches or striations introduced during the Kalottentest or other polishing steps. These artifacts can obscure or mimic SSD features, leading to inaccurate measurements. Frequency-domain methods, particularly the Fast Fourier Transform (FFT), have emerged as powerful tools for addressing this challenge.

    FFT analysis transforms the image from the spatial domain to the frequency domain, where periodic patterns manifest as distinct peaks. By identifying and suppressing these peaks—either by zeroing out their values or applying frequency-domain filters—researchers can selectively remove unwanted artifacts while preserving the true SSD features ([Elhossini et al., 2023]). The result is a cleaner, more interpretable image that supports more accurate segmentation and measurement.

    2.6 Key Insights from Literature

    A careful reading of the literature on subsurface damage (SSD) in optical materials reveals a field that is both technically mature and still evolving. The collective work of researchers over the past several decades has not only clarified the mechanisms and consequences of SSD, but also driven the development of new detection methods, process controls, and analytical tools. This section distills the most important lessons and trends from the literature, providing a foundation for the experimental and analytical work that follows in this thesis.

    SSD Is Inevitable, but Controllable
One of the most consistent findings is that SSD is an unavoidable byproduct of mechanical processing in brittle optical materials (Cui et al., 2023; Suratwala et al., 2005). No matter how carefully grinding and polishing are performed, some degree of microcracking and subsurface flaw formation is inevitable. However, the literature also makes clear that the depth, density, and morphology of SSD can be controlled—if not eliminated—by careful selection of process parameters. Finer abrasives, lower loads, and optimized feed rates consistently result in shallower and less severe SSD (Yang, 2005). This insight has led to the widespread adoption of multi-step finishing protocols, where each stage is designed to remove the damage left by the previous one.

    Material Properties Matter
Another key lesson is that not all optical materials respond to processing in the same way. Glasses like N-BK7 and Fused Silica are generally more susceptible to deeper and more extensive SSD than glass-ceramics such as Zerodur or ULE. This is attributed to differences in fracture toughness, chemical durability, and microstructural response to stress. The literature suggests that process optimization must be material-specific; what works for one glass may not be effective—or even safe—for another.

    Non-Destructive and Automated Approaches Are Gaining Ground
Recent years have seen significant progress in non-destructive SSD detection, with methods such as full-field optical coherence tomography (FF-OCT), scanning acoustic microscopy (SAM), and white light interferometry offering new possibilities for in-line and in-situ analysis. While these techniques still face challenges in resolution, sensitivity, and material compatibility, their integration with digital image processing and automation is closing the gap with traditional methods. The literature highlights the importance of combining multiple techniques to cross-validate results and to build a more complete picture of SSD.

    Image Processing and Frequency-Domain Analysis Are Transformative The integration of automated image analysis—especially techniques such as automated binarization, FFT-based artifact suppression, and geometric measurement algorithms—has dramatically improved the objectivity, reproducibility, and efficiency of SSD evaluation. These advances enable the extraction of subtle features, reduce operator bias, and support high-throughput analysis. The literature also highlights the importance of workflow integration, with modular pipelines that can be adapted to new materials and experimental protocols.

    Challenges Remain and the Field Is Evolving Despite these advances, several challenges persist. The quality of automated analysis is still dependent on the quality of input images, and the interpretation of complex SSD morphologies may require the integration of multiple techniques or the development of custom tools. The literature points to the need for further research in areas such as machine learning, 3D data integration, and the development of standardized protocols for image acquisition and analysis. Chapter 3: Experimental Methodology

    3.1 Materials and Sample Preparation

    The foundation of any rigorous analysis into subsurface damage (SSD) in optical materials lies in the careful selection and preparation of samples. This section details the rationale behind the choice of materials, the protocols for their preparation, and the measures taken to ensure consistency and reproducibility throughout the experimental campaign.

    Selection of Optical Materials

    The primary materials selected in this study include N-BK7 (a borosilicate crown glass), Fused Silica (Corning 7980), Zerodur, Calcium Fluoride (CaF₂), and Ultra-Low Expansion (ULE) glass. These materials were selected for their widespread use in high-precision optics and their contrasting mechanical and chemical properties.

    N-BK7 is values for its good optical transmission and moderate hardness but is know to be more susceptible to deeper SSD due to its relatively low fracture toughness. Fused Silica offers high thermal stability and UV transparency but can also exhibit significant SSD under mechanical processing. Zerodur is prized for its dimensional stability and minimal SSD, making it a benchmark for comparison. CaF₂ and ULE provide additional contrast in terms of refractive index, laser damage threshold, and etching response.

    Sample Dimensions and Initial State

    All samples were prepared as flat, polished discs or squares, typically measuring 40 mm × 40 mm × 10 mm or similar, to facilitate uniform processing and ease of handling.

    The initial surfaces were either as-received (for reference) or pre-polished to a high optical quality, ensuring that any SSD observed could be attributed to the controlled experimental procedures rather than pre-existing flaws.

    Grinding and Polishing Protocols

    The introduction of SSD was achieved through a carefully controlled sequence of grinding and, where appropriate, limited polishing. Grinding was performed using diamond tools of varying grit sizes—D126 (rough), D35 (medium), and D16 (fine)—to simulate different levels of surface damage.

    The choice of grit size and grinding parameters was informed by both literature guidance and preliminary trials, with the aim of generating a range of SSD profiles for subsequent analysis.

    Rough Grinding: Coarse diamond tools (e.g., D126) were used to introduce deep SSD, mimicking the early stages of optical fabrication. Medium and Fine Grinding: Progressively finer tools (D35, D16) were employed to reduce SSD depth and refine the surface, simulating the transition to final polishing. Polishing: In some cases, a brief polishing step was included to remove the most superficial damage while preserving the underlying SSD for analysis.

    All grinding operations for this thesis were performed using the Satisloh SPM 140, a state-of-the-art precision grinding machine widely used in the optics industry. The SPM 140 is designed for high-accuracy, repeatable processing of glass and ceramic substrates, offering precise control over parameters such as tool speed, feed rate, and applied load. Its robust construction and advanced automation features ensure consistent material removal and surface quality across multiple samples. By utilizing the Satisloh SPM 140, this study was able to achieve the level of process control and reproducibility necessary for meaningful comparison of SSD profiles across different materials and experimental conditions.

    Throughout the grinding and polishing process, key parameters such as tool rotation speed, feed rate, applied load, and slurry composition were meticulously recorded and kept consistent across samples. This attention to detail was essential for ensuring that observed differences in SSD could be attributed to material properties and processing conditions rather than uncontrolled variability.

    Cleaning and Handling

    After each processing step, samples were thoroughly cleaned using ethanol to remove residual slurry, debris, and potential contaminants. Special care was taken to avoid introducing new scratches or defects during handling, with all samples stored in clean, dust-free containers until further use.

    Documentation and Quality Control

    A comprehensive log was maintained for each sample, recording material type, dimensions, processing parameters, and any observations made during preparation. This documentation not only facilitated reproducibility but also enabled the correlation of SSD features with specific processing histories during later analysis.

    3.2 Kalottentest Procedure and Parameter Selection

    The Kalottentest, or Ball Dimple Method, is a central experimental technique in this thesis, providing a direct and quantitative means to evaluate subsurface damage (SSD) in optical materials. The reliability and interpretability of Kalottentest results depend critically on the careful selection and control of experimental parameters.

    Step-by-Step Procedure

    Sample Mounting and Alignment:
Each sample is securely mounted on the kaloMAX II by BAQ, ensuring that the surface is perpendicular to the axis of the rotating ball. Precise alignment is essential to produce a symmetrical calotte and to avoid introducing artifacts that could confound SSD analysis.


    Selection of Ball Material and Diameter:
The choice of ball material (steel or tungsten carbide) and diameter is guided by the hardness of the sample and the desired depth of the calotte. Tungsten carbide balls are preferred for harder materials or when greater force is required, while steel balls are suitable for softer substrates. Ball diameter is 30 mm. 


    Abrasive Slurry Selection and Sequence:
The sequence of slurries—coarse to fine—enables controlled material removal and the gradual exposure of SSD. For example, a typical sequence might include: Rougher Slurry: 4–6 µm particles Rough Slurry: 1.5–3 µm particles Medium Slurry: 0.75–1.25 µm particles Fine Slurry: 0.25–0.5 µm particles The choice of slurry and the duration of each step are tailored to the material and the target calotte geometry.

    Control of Load, Speed, and Duration:
The applied load, rotation speed (RPM), and duration of each step are meticulously controlled and recorded. For example: Load: 0.5–2 N Speed: 400–1000 RPM Duration: 5–60 seconds per slurry

    Fused Silica

    N-BK7Zerodour For this experiment, the Diamtere of Callote and depth were:

    3.3 Chemical Etching and Neutralization

    Chemical etching is a pivotal step in the experimental workflow for subsurface damage (SSD) analysis, serving as the bridge between mechanical processing and microscopic evaluation. The goal of etching is to selectively reveal the network of microcracks and defects that lie beneath the surface of optical materials—features that are otherwise invisible to the naked eye. This section details the rationale, protocols, and quality control measures for chemical etching and neutralization, drawing on both established literature and the specific practices developed for this thesis. Purpose and Principles of Chemical Etching

    The primary purpose of chemical etching in SSD analysis is to “open up” subsurface cracks and flaws, making them accessible for optical or electron microscopy. In the context of the Kalottentest and related techniques, etching transforms the smooth, featureless surface of a calotte into a landscape where the true extent of damage can be observed and measured (Zhou et al., 1994). The process relies on the fact that damaged regions—where the glass structure is disrupted—are more susceptible to chemical attack than the undamaged bulk. By carefully controlling the etching conditions, it is possible to selectively highlight SSD without excessively enlarging or distorting the features of interest.

    Etching Protocols: Materials and Methods

    For this thesis, hydrofluoric acid (HF) was chosen as the etchant due to its effectiveness in dissolving silica-based glasses and its widespread use in the optics community. The concentration of HF and exposure time were optimized through a combination of literature review, preliminary trials, and consultation with experienced colleagues Typical etching conditions ranged from 1% to 5% HF by volume, with exposure times of 30 seconds to several minutes, depending on the material and the depth of SSD to be revealed. The etching process was carried out in a well-ventilated fume hood, with all personnel wearing appropriate personal protective equipment (PPE), including acid-resistant gloves, goggles, and lab coats. Samples were immersed in the HF solution using plastic tweezers, ensuring that the entire calotte region was exposed to the etchant. The progress of etching was monitored visually, and the process was halted as soon as the desired level of feature visibility was achieved.

    Neutralization and Post-Etching Handling

    Given the hazardous nature of HF, neutralization is a critical safety and quality control step. Immediately after etching, samples were transferred to calcium hydroxide (Ca(OH)₂) and then in water. This neutralization step serves two purposes: it safely deactivates any residual HF by converting it to insoluble calcium fluoride (CaF₂), and it prevents further etching that could blur or distort the SSD features. After neutralization, samples were thoroughly rinsed with distilled water to remove any remaining chemicals and dried with lint-free wipes.

    Challenges and Solutions

    Chemical etching is not without its challenges. Over-etching can lead to excessive widening of cracks, loss of fine detail, or even damage to the sample. Under-etching, on the other hand, may leave critical features hidden. To address these issues, the protocol was iteratively refined, with adjustments made based on the observed results and feedback from microscopy. In some cases, multiple short etching cycles were preferred over a single long exposure, allowing for finer control over the process.

    3.4 Microscopy and Image Acquisition

    The transition from physical sample preparation to digital analysis is a pivotal step in the evaluation of subsurface damage (SSD) in optical materials. High-quality microscopy and careful image acquisition are essential for ensuring that the subtle features of SSD are faithfully captured and made available for objective, quantitative analysis. 3.4.1 Microscope

    All imaging was performed using the Keyence VHX-6000 digital microscope. The VHX-6000 offers a range of magnification options and advanced lighting controls, making it well-suited for the detailed examination of SSD features. Samples were handled with powder-free gloves and stored in dust-free containers to prevent the introduction of new artifacts. The region of interest—typically the center of the calotte or MRF spot—was identified and marked, ensuring that imaging was focused on the most relevant area for SSD analysis. 3.4.2 Magnification and Lighting selection

    Rough-ground calottes were typically imaged at 20x magnification, providing a broad overview of the dimple and its surrounding area. Medium-ground calottes were captured at 50x and 70x, balancing field of view and detail. Fine-ground calottes and MRF spots, where SSD features are often subtle and require greater resolution, were imaged at 100x and 200x magnification.

    Lighting conditions were optimized for each imaging session. Both ring light and co-axial light sources were employed, with the choice depending on the surface characteristics and the type of SSD being targeted. The ring light was particularly effective for highlighting surface topography and broader features. In contrast, the co-axial light source was used to enhance the visibility of fine cracks and subsurface structures. 3.4.3 Image Acquisition

    For each sample, a series of images was acquired at different magnifications and lighting conditions to ensure comprehensive coverage of the SSD features. Images were saved in high-resolution digital formats (e.g., TIFF or PNG), preserving both the raw data and any associated metadata. Files were named according to a standardized convention that encoded sample ID, processing step, and imaging parameters (e.g., “FusedSilica_Sample_3_Fine_50x_RingLight.tif”).

    3.5 Acoustic Emision (AE) Monitoring and Force Measurement

    A comprehensive understanding of the grinding process—and its impact on subsurface damage (SSD)—requires more than just post-process inspection. Real-time monitoring, using both acoustic and mechanical signals, provides a window into the events that shape the material beneath the surface. In this research, a carefully engineered setup was developed to capture both Acoustic Emission (AE) and force data, ensuring that every significant event during grinding was recorded and could be correlated with the resulting SSD.

    3.5.1 Custom Plate Design and Sensor Integration

    Central to the experimental setup was a custom-fabricated steel base plate, designed specifically to accommodate both AE and force sensors without compromising measurement fidelity. The NBK7 glass sample was bonded to this plate using “MKS shwartz 39”, a high-strength adhesive chosen for its reliability and compatibility with both glass and steel. Special attention was paid to the placement of the VS900-F AE sensor (Vallen System). A dedicated hole was machined into the plate to house the sensor’s microphone, which was secured with

    hot glue. The NBK7 sample was then positioned above the sensor, ensuring optimal acoustic coupling. The remainder of the plate was covered with MKS shwartz 39, but care was taken to prevent the adhesive from contacting the AE sensor directly, preserving the integrity of the acoustic signal.

    This thoughtful arrangement allowed for the simultaneous, high-fidelity acquisition of AE signals and force data. The AE system itself consisted of the AMSY-6 MB 6 chassis by Vallen System, the VS900-F AE sensor, and an AEP5 pre-amplifier. This combination provided the sensitivity and bandwidth needed to detect the high-frequency elastic waves generated by microcrack formation, grain breakout, and friction at the tool–workpiece interface.3.5.2 Acoustic Emission Monitoring in Practice

    During grinding, the AE system continuously recorded elastic waves emitted from the glass substrate. The real-time data stream captured both subtle and pronounced events, such as the formation of new microcracks or the detachment of grains. By analyzing the amplitude, frequency, and timing of these signals, it was possible to distinguish between routine process noise and the early signs of surface or subsurface damage.

    3.5.3 Force Measurement with Kistler Dynamometer

    Complementing the AE monitoring, force measurement was carried out using a Kistler Dynamometer Type 9256C, connected to a Kistler Multi-channel Amplifier Type 5070A. The dynamometer was installed beneath the custom base plate, ensuring that all forces transmitted through the workpiece were accurately captured. This setup enabled the measurement of forces along three orthogonal axes (Fx, Fy, Fz) as well as the applied torque, providing a detailed picture of the mechanical interaction between the grinding tool and the glass substrate.

    The force data, recorded at high sampling rates, allowed for the detection of both steady-state and transient events. Sudden spikes in force could indicate the formation of a new crack or a change in the grinding regime, while gradual drifts often signaled tool wear or evolving material properties. By correlating force data with AE signals, it was possible to pinpoint the exact moments when subsurface damage was likely to occur, and to adjust process parameters accordingly. 3.5.4 Integration and Impact

    The dual-monitoring approach—combining AE and force measurement—provided a comprehensive view of the grinding process. This integration not only enhanced process understanding but also supported the development of more stable, repeatable, and sustainable grinding protocols. The insights gained from these measurements were instrumental in optimizing process parameters, minimizing SSD, and advancing the broader goals of the MoKI-Pro project.

    3.6 Summary of Experimental Workflow

    This section synthesizes the preceding methodology, providing a clear, step-by-step overview of the entire experimental process—from material selection to data readiness for advanced analysis. The aim is to offer both a roadmap for replication and a transparent account of how each stage contributes to the integrity of the final results.

    Step 1.: Material Selection and Initial Preperation

    The workflow begins with selecting representative optical materials—N-BK7, Fused Silica, Zerodur, CaF₂, and ULE—each chosen for its high-precision capabilities and diverse mechanical and chemical properties ([Kalottentest Report: N-BK7, Fused Silica, Zerodur, pdf, Table 1]). Samples are made as uniformly flat, polished discs or squares to ensure uniform starting point and consistent processing.

    Step 2.: Controlled Grinding and Polishing

    To introduce SSD reproducibly, each sample is ground using diamond tools (D126, D35, D16) under controlled parameters—tool rotation speed, feed rate, applied load, and slurry composition. In some cases, a brief polishing step removes superficial damage while preserving the underlying SSD.

    Step 3.: Kalottentest Execution

    The workflow centers on the Kalottentest, where a rotating ball (steel or tungsten carbide) and a sequence of abrasive slurries create a spherical dimple on the sample. The process is incremental, with the calotte diameter measured after each slurry application to precisely control depth and geometry. Ball material, diameter, slurry sequence, and process parameters are tailored to each material and grinding level based on literature and preliminary trials.

    Step 4.: Chemical etching and Neutralization

    Once the calotte is formed, the sample is etched in hydrofluoric acid (HF) to reveal SSD features. The concentration and exposure time are optimized to open subsurface cracks without causing excessive enlargement or distortion. The sample is then neutralized in a Ca(OH)₂ water solution and rinsed thoroughly to remove residual chemicals.

    Step 5.: High Resolution Microscopy and Image Acquisition

    The etched calottes are imaged with the Keyence VHX-6000 digital microscope. Magnification and lighting are chosen based on the SSD features, with images acquired at multiple settings for full coverage. All parameters are standardized, documented, and saved in high-resolution formats with metadata.

    Chapter 4: Image Analysis Pipeline

    4.1 Why Automate SSD Analysis?

    The evaluation of subsurface damage (SSD) in optical materials has long been a challenge for both researchers and industry professionals. Traditionally, the process has relied on manual inspection of microscope images, with operators visually identifying and measuring features such as calotte boundaries, SSD depth, and area. While this approach can yield valuable insights, it is inherently limited by subjectivity, inconsistency, and the sheer time required to analyze large datasets. As the demands on optical manufacturing have grown—driven by the need for higher throughput, tighter tolerances, and more reliable quality control—the limitations of manual analysis have become increasingly apparent. 4.1.1 The Case for Automation

    Automated image analysis offers a compelling solution to these challenges. By leveraging advances in computer vision, machine learning, and digital signal processing, automated pipelines can process large volumes of data with speed, consistency, and objectivity that far surpass human capabilities. The key benefits of automation in SSD analysis include:

    Objectivity: Automated algorithms apply the same criteria to every image, eliminating operator bias and ensuring that results are consistent across samples and sessions. Reproducibility: Digital workflows can be precisely documented, parameterized, and shared, enabling other researchers or industrial partners to replicate the analysis and validate the results. Efficiency: High-throughput analysis is possible, allowing for the evaluation of large datasets in a fraction of the time required for manual inspection. Sensitivity and Precision: Advanced techniques such as automated thresholding, frequency-domain filtering (such as FFT), and geometric measurement algorithms can detect and quantify subtle SSD features that might be missed by the human eye.

    4.1.2 Limitation of Manual and Semi-Manual Methods

    Manual measurement of calotte diameters and SSD features is inherently subjective. Operators may interpret the edge of the dimple differently, particularly in images where the boundary is faint or interrupted by artifacts. This subjectivity introduces variability into SSD depth calculations, undermining the reliability of comparative studies and process optimization. Furthermore, manual analysis is slow and labor-intensive, making it impractical for large datasets or high-throughput environments. The need for repeated measurements to ensure reliability further compounds the workload. The lack of standardized protocols and the inherent variability of human judgment limit the reproducibility of results and hinder the comparison of data across studies or institutions.

    These limitations are not merely academic; in practice, they translate into increased costs, reduced yield, and compromised performance for optical systems. The inability to reliably detect and quantify SSD can result in premature component failure and degraded system performance.

    4.1.3 The Role of Automation in Modern Research and Industry

    In the context of this thesis, automation is not just a technical upgrade—it is a necessity. The volume of data generated by high-resolution microscopy, the complexity of SSD features, and the need for objective, reproducible results all point toward the adoption of automated analysis pipelines. By integrating automated circle detection, FFT-based artifact suppression, and quantitative feature extraction, the analysis becomes not only faster but also more reliable and transparent. This is especially important in collaborative projects like MoKI-Pro, where data must be shared, compared, and validated across multiple institutions and industrial partners.

    4.2 Preprocessing and Automatic Calotte Circle Detection

    The reliability of any quantitative analysis in subsurface damage (SSD) research hinges on the clarity and objectivity of the initial image processing steps. In this thesis, the focus is on developing a robust, semi-automated pipeline for detecting and measuring the calotte (dimple) boundary in microscope images—a process that is foundational for all subsequent SSD evaluation. The approach is grounded in the mathematical and algorithmic framework described by Elhossini et al. (2023), and is designed to minimize operator bias, maximize reproducibility, and handle the real-world challenges of noisy, artifact-laden images.

    4.2.1 Rationale for Automation

    Manual measurement of calotte diameters is inherently subjective. Operators may interpret the edge of the dimple differently, particularly in images where the boundary is faint or interrupted by artifacts. This subjectivity introduces variability into SSD depth calculations, undermining the reliability of comparative studies and process optimization. Automated circle detection addresses these challenges by applying consistent, mathematically defined criteria to every image, ensuring that results are both objective and reproducible. 4.2.2 Preprocessing: Preparing the Image for Analysis

    The first step in the pipeline is to prepare the raw microscope image for analysis. This is not a trivial task: images acquired from etched calottes often suffer from uneven illumination, random noise, and the presence of small debris or surface artifacts. If left unaddressed, these issues can confound edge detection and lead to inaccurate or inconsistent measurements.

    Grayscale and Noise Reduction

All images are converted to grayscale, focusing the analysis on intensity variations rather than color. This step simplifies the data and reduces computational complexity. To suppress random noise—often present in high-magnification images—a median or Gaussian filter is applied. The choice of filter and kernel size is optimized based on the typical scale of noise and the resolution of the images.
 Thresholding and Morphological Operations

To segment the calotte from the background, global thresholding is applied in combination with Otsu’s method. Otsu’s method is a mathematically defined algorithm that determines the optimal threshold by minimizing the intra-class variance of the black and white pixels. This approach ensures that the segmentation remains objective and adapts to the specific intensity distribution of each image. The resulting binary image is further refined using morphological opening and closing, which are mathematical operations designed to remove small artifacts and fill gaps in the segmented regions. These steps are crucial for producing clean, well-defined boundaries for subsequent analysis.




    4.2.3 Automatic Calotte (Circle) Detection

    The core of the analysis pipeline is the automatic detection of the calotte boundary. The method implemented here is grounded in mathematically defined criteria and leverages the OpenCV library in Python.

    Contour Detection and Circularity Assessment

Once the image is preprocessed, external contours are detected using OpenCV’s findContours function. Each contour is a sequence of points representing a potential boundary in the image. To distinguish the true calotte, each contour is evaluated for its “circularity,” a dimensionless metric that quantifies how closely a shape resembles a perfect circle. The most common formula, as described by Elhossini et al. (2023), is:

 (1) 

This metric, which ranges from 0 (highly irregular) to 1 (perfect circle), provides a mathematically objective criterion for selecting the most circle-like contour. By choosing the contour with the highest circularity and appropriate size, the algorithm ensures that the detected boundary is not only closed and continuous but also closely matches the expected geometry of a calotte.

    Circle and Ellipse Fitting

After identifying the most circular contour, the algorithm fits the smallest enclosing circle using OpenCV’s minEnclosingCircle function. This geometric, least-squares approach finds the center and radius of the circle that best approximates the contour points. The mathematical objective is to minimize the sum of squared distances between the contour points and the circle’s edge:

 (2)
 Geometric Calculation of SSD Depth

The geometric calculation of SSD depth follows the standardized framework established in the MoKI-Pro project. The Kalottentest produces a spherical dimple (calotte) on the surface of the optical sample. The geometry of this dimple is defined by the radius of the ball used to create it and the measured diameters of the calotte and, where applicable, the inner SSD boundary . The depth of the dimple —and thus the depth of SSD—can be derived from these measurements using the principles of spherical geometry.

The maximum depth of the calotte, corresponding to the outer diameter , is given by the equation:
 (3)
 where:
( d ) is the measured outer diameter of the calotte,
( r ) is the radius of the ball used in the Kalottentest.

    This formula is derived from the geometry of a sphere, considering the segment (spherical cap) formed by the intersection of the ball and the sample surface.

    Calculation of SSD Depth Using Inner Diameter

    To specifically determine the depth of SSD, the MoKI-Pro Guidelines recommend measuring the inner diameter that marks the boundary of visible SSD after etching. The depth corresponding to this inner diameter is:

    					(4)
    

    Here, is the diameter of the region within the calotte where SSD is no longer visible. The difference between and provides a direct measure of the SSD layer’s thickness.

    4.2.3 Validation and Impact (add images, failures too)

    The automated calotte detection and measurement pipeline was validated against manual measurements and reference standards. In the majority of cases, the automatically detected diameters deviated by less than 2% from manual measurements, with the greatest discrepancies occurring in images with extremely faint or interrupted boundaries. The algorithm’s robustness was further demonstrated by its ability to handle images with varying contrast, noise levels, and artifact presence.

    The impact of this approach is significant: it enables high-throughput, objective, and reproducible analysis of SSD, supporting both research and industrial quality control. By eliminating operator bias and streamlining the measurement process, the pipeline sets a new standard for SSD evaluation in optical materials.

    4.3 Artifact Suppression (FFT)

    Artifact suppression is a critical step in the automated analysis of calotte images for SSD evaluation. Even after careful preprocessing and boundary detection, periodic artifacts—most notably, scratches and striations introduced during the Kalottentest—can obscure or mimic true SSD features. These artifacts, if left unaddressed, can lead to inaccurate segmentation, misclassification, and ultimately, unreliable quantitative results. In this thesis, a frequency-domain approach based on the Fast Fourier Transform (FFT) is implemented to isolate and remove these unwanted patterns, ensuring that the analysis focuses on genuine material damage rather than process-induced noise. 4.3.1 Rationale for Frequency-Domain Artifact Suppression

    Periodic artifacts in calotte images typically manifest as regular, directional patterns—vertical, slightly tilted or horizontal lines, for example—that are difficult to distinguish from genuine SSD features in the spatial domain. Traditional spatial-domain filters, such as median or Gaussian blurring, can reduce random noise but often fail to selectively target these structured artifacts without also degrading the features of interest. The FFT, by transforming the image into the frequency domain, provides a means to directly identify and suppress the specific frequencies associated with these artifacts, while preserving the underlying SSD information. 4.3.2 Theoretical Background: FFT in Image Processing

    The FFT is a mathematical algorithm that decomposes an image into its constituent spatial frequencies. In the frequency domain, periodic patterns in the image appear as distinct peaks at specific locations in the spectrum. For example, vertical scratches in the spatial domain correspond to horizontal lines in the frequency domain, and vice versa. By analyzing the magnitude and phase of the FFT, it is possible to pinpoint the frequencies responsible for unwanted artifacts.

    The process begins by computing the 2D FFT of the grayscale image, resulting in a complex-valued frequency spectrum. The magnitude of this spectrum is visualized to identify dominant frequency components. Peaks corresponding to periodic artifacts are then automatically detected—often using thresholding and local maxima analysis—and masked or set to zero. The modified frequency spectrum is then transformed back to the spatial domain using the inverse FFT, yielding an image where the true SSD features are preserved and periodic artifacts are suppressed. 4.3.3 Practical Implementation and Workflow

    The artifact suppression pipeline implemented in this thesis follows a systematic workflow:

    Preprocessing:
The input image is first preprocessed as described in Section 4.2, ensuring that the data is clean and well-prepared for frequency analysis.
 FFT Computation:
The 2D FFT of the preprocessed image is computed using the NumPy or OpenCV library in Python. The resulting frequency spectrum is shifted so that the zero-frequency component is centered, making it easier to visualize and manipulate.
 Peak Detection and Masking:
The magnitude spectrum is analyzed to identify peaks corresponding to periodic artifacts. This is typically done by applying a threshold to the spectrum and searching for local maxima. Once identified, these peaks are masked—either by setting their values to zero or by applying a Gaussian or notch filter to attenuate the unwanted frequencies.
 Inverse FFT and Image Reconstruction:
The modified frequency spectrum is transformed back to the spatial domain using the inverse FFT. The resulting image retains the true SSD features while suppressing the periodic artifacts.
 Validation and Quality Control:
The effectiveness of artifact suppression is assessed by comparing the original and processed images, both visually and quantitatively. Key metrics include the reduction in measurement variance, the improvement in SSD feature clarity, and the consistency of results across multiple samples.

    4.3.4 Advantages, Limitations, and Future Directions

    The FFT-based artifact suppression approach offers several key advantages:

    Selectivity:
By targeting specific frequencies, the method can remove structured artifacts without blurring or distorting the features of interest.
 Objectivity and Reproducibility:
The algorithm applies the same criteria to every image, eliminating operator bias and ensuring consistent results.
 Scalability:
The workflow can be applied to large datasets with minimal manual intervention, supporting high-throughput analysis. However, the method is not without limitations. The effectiveness of artifact suppression depends on the accurate identification of relevant frequency peaks; overly aggressive masking can inadvertently remove true SSD features, while insufficient masking may leave residual artifacts. Careful parameter tuning and visual validation are therefore essential components of the workflow.

    Looking ahead, future work may focus on integrating metadata extraction (e.g., magnification, lighting conditions) and developing adaptive filtering techniques that can automatically adjust to the specific characteristics of each image. The ultimate goal is a fully automated, metadata-driven pipeline for SSD analysis, as envisioned in the MoKI-Pro Guidelines.

    list out the abbreviations

  2. Assistant

    Here is a list of abbreviations used in the provided chapters of your thesis on subsurface damage (SSD) analysis in optical materials:


    List of Abbreviations

    AE
    Acoustic Emission
    A technique for real-time monitoring of elastic waves generated by microcracks and fractures during grinding or polishing.

    CaF₂
    Calcium Fluoride
    An optical material used in precision optics with distinct mechanical and chemical properties.

    FFT
    Fast Fourier Transform
    A mathematical algorithm used for frequency-domain analysis and artifact suppression in image processing.

    FF-OCT
    Full-Field Optical Coherence Tomography
    A non-destructive imaging technique for 3D subsurface visualization with optical interferometry.

    LIDT
    Laser-Induced Damage Threshold
    The maximum laser energy density an optical material can tolerate before damage occurs.

    MRF
    Magnetorheological Finishing
    A deterministic, non-contact polishing technique employing a magnetically controlled fluid to refine optical surfaces.

    NBK7
    Borosilicate Crown Glass
    A common optical glass substrate known for good transmission and moderate hardness.

    PPE
    Personal Protective Equipment
    Protective clothing and equipment for safe handling of hazardous substances such as hydrofluoric acid.

    QED
    Quantum Electro Dynamics
    Company providing the MRF machine (Q22-X) used in polishing experiments.

    RAMI
    (If not explicitly mentioned in the text, not included; no indication in the provided text.)

    SAM
    Scanning Acoustic Microscopy
    A technique utilizing ultrasound to image subsurface defects and features nondestructively.

    SSD
    Subsurface Damage
    Microstructural defects below the surface of optical components introduced during mechanical processing.

    SPM 140
    Satisloh SPM 140
    A precision grinding machine used for controlled material removal on optical substrates.

    ULE
    Ultra-Low Expansion Glass
    A glass-ceramic with minimal thermal expansion, used in precision optical systems.

    Z-Map
    Height Map (Surface Relief Profile)
    A 3D representation of surface topography used for correlating physical features with 2D images.


    If you want, I can prepare this list formatted according to your thesis style (e.g., alphabetical order, with definitions, units where applicable). Just let me know!