Engineering topics

Bryan Harder

Publications and source records attributed to Bryan Harder.

5 records

NASA NTRS2024

Thermochemical Interactions Between Yttria-Stabilized Zirconia and Molten Lunar Regolith Simulants

Oxygen produced through in-situ resource utilization (ISRU) is critical to maintaining a permanent human presence on the lunar surface. Molten regolith electrolysis and carbothermal reduction are two promising ISRU techniques for generating oxygen directly from lunar regolith, which is primarily a mixture of oxide minerals; however, both processes require operating temperatures of 1600 C to melt lunar regolith and dissociate the molten oxides. These conditions limit the use of many oxide refractory materials, such as Al2O3 and MgO, due to rapid degradation resulting from reactions between the refractory materials and molten lunar regolith. Yttria-stabilized zirconia (YSZ) is shown here to be a promising refractory oxide to provide containment of molten regolith while demonstrating limited reactivity. This work focuses on corrosion studies of YSZ powders and dense YSZ crucibles in contact with molten lunar maria and highlands regolith simulants at 1600 C. The interactions between YSZ and molten regolith are characterized using SEM/EDS, XRD, and EBSD. A FactSage thermochemical model is created to compare with the experimental results. These combined analyses suggest that lunar maria regolith will degrade YSZ faster than lunar highlands regolith due to the lower viscosity of the maria regolith. The feasibility of long-term molten regolith containment with YSZ is discussed based on the YSZ powder and crucible results.

Kevin Yu↗

NASA NTRS2022

Microstructure Segmentation With Deep Learning Encoders Pre-Trained on a Large Microscopy Dataset

This study examined the improvement of microscopy segmentation intersection over union accuracy by transfer learning from a large dataset of microscopy images called MicroNet. Many neural network encoder architectures were trained on over 100,000 labeled microscopy images from 54 material classes. These pre-trained encoders were then embedded into multiple segmentation architectures including UNet and DeepLabV3+ to evaluate segmentation performance on created benchmark microscopy datasets. Compared to ImageNet pre-training, models pre-trained on MicroNet generalized better to out-of-distribution micrographs taken under different imaging and sample conditions and were more accurate with less training data. When training with only a single Ni-superalloy image, pre-training on MicroNet produced a 72.2% reduction in relative intersection over union error. These results suggest that transfer learning from large in-domain datasets generate models with learned feature representations that are more useful for downstream tasks and will likely improve any microscopy image analysis technique that can leverage pre-trained encoders.

machine learning↗

NASA NTRS2022

Microstructure Segmentation with Deep Learning Encoders Pre-Trained on a Large Microscopy Dataset

This study examined the improvement of microscopy segmentation accuracy by transfer learning from a large dataset of microscopy images called MicroNet. Many neural network encoder architectures, including VGG, Inception, and ResNet, were trained on over 100,000 labelled microscopy images from 54 classes. These pre-trained encoders were then embedded into multiple segmentation architectures including U-Net and DeepLabV3+ to evaluate segmentation performance on newly created benchmark microscopy datasets. Compared to ImageNet pre-training, models pre-trained on MicroNet generalized better to out-of-distribution micrographs taken under different imaging and sample conditions and were more accurate with less training data. When training with only a single Ni-superalloy image, pre-training on MicroNet produced a 72.2 percent reduction in relative segmentation error. These results suggest that transfer learning from large in-domain datasets generate models with learned feature representations that are more useful for downstream tasks and will likely improve any microscopy image analysis technique that can leverage pre-trained encoders.

machine learning↗

NASA NTRSDate not supplied

Thermochemical Interactions of Yttria-Stabilized Zirconia and Molten Lunar Regolith Simulants

Oxygen produced from lunar resources through in-situ resource utilization (ISRU) is critical to maintaining a permanent human presence on the lunar surface. Molten regolith electrolysis and carbothermal reduction are two promising ISRU techniques for generating oxygen directly from lunar regolith, which is primarily a mixture of oxide minerals; however, both processes require operating temperatures of 1600C to melt lunar regolith and dissociate the molten oxides. These conditions limit the use of many oxide refractory materials, such as Al2O3 and MgO, due to rapid degradation resulting from reactions between the refractory materials and molten lunar regolith. Yttria-stabilized zirconia (YSZ) is a promising refractory oxide to provide containment of molten regolith while demonstrating limited reactivity. This work focuses on corrosion studies of YSZ powders and dense YSZ crucibles in contact with molten lunar mare and highlands regolith simulants at 1600C. The interactions between YSZ and molten regolith are characterized using SEM/EDS, XRD, and EBSD with an emphasis on elemental and microstructural analysis to assess reactivity and degradation of YSZ. Due to lunar regolith’s similar composition to calcium-magnesium-aluminosilicates (CMAS) and YSZ’s usage as a thermal barrier coating, these interactions can serve to inform YSZ/CMAS behavior by simulating cases of elevated CMAS/YSZ ratios and for higher than intended gas turbine temperatures.

Kevin Yu↗

NASA NTRSDate not supplied

Degradation of Yttria-Stabilized Zirconia in Molten Regolith Electrolysis Applications

The generation of oxygen from lunar resources is an enabling technology to support a sustained human presence on the lunar surface. Molten regolith electrolysis is a promising process for directly electrolyzing lunar regolith, which is primarily a mixture of oxide minerals. During electrolysis, metallic products (Fe, Si) are formed at the cathode and oxygen is produced at the anode. In order to perform electrolysis, the regolith must be molten to enable ionic transport of metallic cations and oxygen anions, requiring an operating temperature of 1600C. Common refractory oxide materials, such as Al2O3 and MgO, undergo rapid degradation due to the high operating temperature and contact with the corrosive molten regolith. This work investigates yttria-stabilized zirconia (YSZ) for molten regolith electrolysis applications. Corrosion studies are presented to elucidate the degradation mechanism of YSZ powders and crucibles exposed to lunar regolith simulants at 1600C. Two regolith simulants are used to represent regolith from the lunar maria (JSC-1A) and the lunar highlands (LHS-1). Degradation of YSZ in contact with regolith simulants is evaluated using SEM/EDS, XRD, and EBSD. Furthermore, a modified molten regolith electrolysis reactor is presented utilizing YSZ at the anode to enable facile collection of oxygen produced during electrolysis.

Kevin Yu↗