Search

Thermal Barrier Coatings (TBCs), and plasma spray coatings in general, require critical control over the deposited thickness to achieve reliable coating performance. Currently, the plasma spray industry quantifies thickness by sampling the part before and after TBC deposition. Approximate thickness is thus inferred from previous runs. However, process variability can allow errors to propagate in this result that leads to wasted time and resources, and can ultimately lead to non-reliant coatings. To this end, an in situ optical fringe profilometer is developed that enables coating thickness measurements across a 2-dimensional surface. The initial profilometer concept is explored through requirements and trade studies, leading to a hardware and algorithm design family and prototype build to capture and compare real-world data to simulation and model predictions. This initial result shows a viable path-forward and the ability to achieve micrometer-scale depth resolution. Modifications and alterations to the in situ profilometer are then explored to improve the performance limits achievable. In specific, industrial spray coatings operate by dropping fine-grain media into a high pressure gas line aimed through a plasma torch to impart enough thermal and kinetic energy to stick to the part surface. This presents a challenging operational environment for an optical depth measurement sensor, working with a variable high-temperature blackbody stray light source; constant part rotation and plasma gun movement; and a non-isolated vibration environment. As such, the concept of the profilometer is further adapted specific to this end-purpose, by developing and reviewing both dual-fringe projection and plenoptic imaging. These techniques allow an improvement to both the system micro- and macroscopic depth retrieval limits, allowing a method to solve for an extended range of phase ambiguities and relax object focus requirements (respectively). The end result is a system concept an

A snapshot imaging Mueller matrix polarimeter (SIMMP) is theoretically described and empirically demonstrated through simulation. Spatial polarization fringes are localized onto a sample by incorporating polarization gratings (PGs) into a polarization generator module. These fringes modulate the Mueller matrix (MM) components of the sample, which are subsequently isolated with PGs in an analyzer module. The MM components are amplitude modulated onto spatial carrier frequencies which, due to the PGs, maintain high visibility in spectrally broadband illumination. An interference model of the SIMMP is provided, followed by methods of reconstruction and calibration. Lastly, a numerical simulation is used to demonstrate the system's performance in the presence of noise.

In this work, image processing algorithms are presented for an advanced sensor classification known collectively as imaging modulated polarimetry. The image processing algorithms presented are novel in that they use frequency domain based approaches, in comparison to the data domain based approaches that all previous algorithms have employed. Under the conditions on the data and imaging device derived in this work, the frequency domain based demodulation algorithms will optimally reduced reconstruction artifacts in a least squared sense. This work provides a framework for objectively comparing polarimeters that modulate in different domains (i.e. time vs. space), referred to as the spectral density response function. The spectral density response function is created as an analog to the modulation transfer function (or the more general transfer function for temporal devices) employed in the design of conventional imaging devices. The framework considers the total bandwidth of the object to be measured, and then can consider estimation artifacts that arise in both time and space due to the measurement modality that has been chosen. Using the framework for objectively comparing different modulated polarimeters (known as the spectral density response function), a method of developing a Wiener filter for multi-signal demodulation is developed, referred to as the polarimetric Wiener filter. This filter is then shown to be optimal for one extensive test case. This document provides one extensive example of implementing the algorithms and spectral density response calculations on a real system, known as the MSPI polarimeter. The MSPI polarimeter has been published extensively elsewhere, so only a basic system description here is used as necessary to describe how the methods presented here can be implemented on this system.

A white-light snapshot channeled linear imaging (CLI) polarimeter is demonstrated by utilizing polarization gratings (PGs). The CLI polarimeter is capable of measuring the two-dimensional distribution of the linear Stokes polarization parameters by incorporating two identical PGs, in series, along the optical axis. In this configuration, the general optical shearing functionality of a uniaxial crystal-based Savart plate is realized. However, unlike a Savart plate, the diffractive nature of the PGs creates a linear dependence of the shear versus wavelength, thus providing broadband functionality. Consequently, by incorporating the PG-based Savart plate into a Savart plate channeled imaging polarimeter, white-light interference fringes can be generated. This enables polarimetric image data to be acquired at shorter exposure times in daylight conditions, making it more appealing over the quasi-monochromatic channeled imaging polarimeters previously described in the literature. Furthermore, the PG-based device offers significantly more compactness, field of view, optical simplicity, and vibration insensitivity than previously described white-light CLI polarimeters based on Sagnac interferometers. Included in this paper are theoretical descriptions of the linear (S0, S1, and S2) and complete (S0, S1, S2, and S3) channeled Stokes imaging polarimeters. Additionally, descriptions of our calibration procedures and our experimental proof of concept CLI system are provided. These are followed by laboratory and outdoor polarimetric measurements of S0, S1, and S2.

A common technique, referred to as channeled imaging polarimetry (ClP), enables the snapshot acquisition of the 2- dimensional Stokes parameters of an arbitrary scene or sample. It achieves this by amplitude modulating the Stokes parameters onto various interference-based spatial carrier frequencies. While this technique has utility. it often suffers from low signal-lo-noise ratios, in remote sensing scenarios. due to its requirement of narrow spectral bandwidth illumination 3 nm in the visible). This paper discusses one hardware implementation that can be utilized to overcome this limitation. Consequently, an overview of the theoretical and experimental development of a unique Sagnac interferometer is discussed. Both laboratory and outdoor data are included to demonstrate the instruments ability to measure polarization in arbitrary scenes. Inclusion of blazed diffraction gratings inside the interferometer enables whitelight interference fringes to be generated. By incorporating these gratings, the operational bandwidth of the interference fringes can exceed approximately 300 nm; two orders of magnitude greater than previous CIP implementations. Lastly by modifying the diffraction grating, the sensor becomes capable of snapshot multispectral imaging. This will be briefly discussed. with both a theoretical description and experimental data., Submitted to the SPIE Aerospace Conference.

The theoretical and experimental demonstration of a multispectral Sagnac interferometer (MSI) is presented. The MSI was created by including two multiple-order blazed diffraction gratings in both arms of a standard polarization Sagnac interferometer (PSI). By introducing these high-order diffractive structures, unique spectral passbands can be amplitude modulated onto coincident carrier frequencies. Extraction of the modulated multispectral images, corresponding to each passband, is accomplished within the Fourier domain. This yields a unique multispectral sensor capable of imaging all the passbands in a single snapshot. First, the theoretical operating principles of a polarization Sagnac interferometer are discussed to provide a context for the MSI. This is followed by the theoretical and experimental development of the MSI, which is an extension of a dispersion-compensated polarization Sagnac interferometer (DCPSI) based polarimeter. Indoor and outdoor testing and validation of the MSI is performed by observing vegetation, demonstrating the ability of our experimental setup to detect four distinctive spectral passbands., Submitted for publication in Optics Express Journal. The original document contains color images.

In this work, three methods of measuring the polarization state of light in the thermal infrared (3-12 microns) are modeled, simulated, calibrated and experimentally verified in the laboratory. The first utilizes the method of channeled spectropolarimetry (CP) to encode the Stokes polarization parameters onto the optical power spectrum. This channeled spectral technique is implemented with the use of two Yttrium Vanadate (YVO4) crystal retarders. A basic mathematical model for the system is presented, showing that all the Stokes parameters are directly present in the interferogram. Theoretical results are compared with real data from the system, an improved model is provided to simulate the effects of absorption within the crystal, and a modified calibration technique is introduced to account for this absorption. Lastly, effects due to interferometer instabilities on the reconstructions, including nonuniform sampling and interferogram translations, are investigated and techniques are employed to mitigate them.Second is the method of prismatic imaging polarimetry (PIP), which can be envisioned as the monochromatic application of channeled spectropolarimetry. Unlike CP, PIP encodes the 2-dimensional Stokes parameters in a scene onto spatial carrier frequencies. However, the calibration techniques derived in the infrared for CP are extremely similar to that of the PIP. Consequently, the PIP technique is implemented with a set of four YVO4 crystal prisms. A mathematical model for the polarimeter is presented in which diattenuation due to Fresnel effects and dichroism in the crystal are included. An improved polarimetric calibration technique is introduced to remove the diattenuation effects, along with the relative radiometric calibration required for the BPIP operating with a thermal background and large detector offsets. Data demonstrating emission polarization are presented from various blackbodies, which are compared to data from our Fourier transform infrared spect