Search citation statements
Paper Sections
Citation Types
Year Published
Publication Types
Relationship
Authors
Journals
Erosion is a phenomenon involving material removal due to fluid entrained particles impinging on material surfaces. Experiments are crucial to investigate erosion severity and the impact of different physical parameters. However, conducting these experiments is expensive and time-consuming, and collecting high-quality data is challenging due to the small scales of material removal and sensitivity to external environment changes. Over the years, research at the Erosion/Corrosion Research Center (E/CRC) at the University of Tulsa has focused on investigating solid particle erosion in pipelines and fittings. Experiments are conducted using a large-scale boom tower flow loop, and erosion data is collected with high-accuracy, temperature-compensated ultrasonic transducers. This has resulted in more than 200 sets of high-quality experimental data covering various pipe diameters, particle diameters, sand concentrations, and flow velocities. Mining the inherent patterns and physical laws within this data is valuable for guiding future process condition selections and experimental tests. With the existing data covering a wide range of conditions, machine learning methods can be utilized to augment the data, ensuring the newly generated data adheres to physical laws and has similar distributions. Conditional Wasserstein Generative Adversarial Networks (CWGANs) are a class of machine learning models designed for generating conditioned data. CWGANs can be applied to generate additional training data when the available data is limited, significantly expanding the erosion databank for subsequent mechanistic modeling. This work proposes the Conditional GAN Synergized SPPS model (ConGANergy) framework for engineering data augmentation, which utilizes CWGANs to expand erosion experimental datasets. We conducted experiments to evaluate the performance of the proposed framework by comparing the newly generated data against original experimental data and data obtained through a semi-mechanistic erosion prediction model Sand Production Pipe Saver (SPPS) developed at E/CRC. This approach provides engineers and researchers with a powerful tool to expand databases and build robust mechanistic models, leveraging the inherent patterns within high-quality experimental data while adhering to the physical laws governing erosion phenomena.
Erosion is a phenomenon involving material removal due to fluid entrained particles impinging on material surfaces. Experiments are crucial to investigate erosion severity and the impact of different physical parameters. However, conducting these experiments is expensive and time-consuming, and collecting high-quality data is challenging due to the small scales of material removal and sensitivity to external environment changes. Over the years, research at the Erosion/Corrosion Research Center (E/CRC) at the University of Tulsa has focused on investigating solid particle erosion in pipelines and fittings. Experiments are conducted using a large-scale boom tower flow loop, and erosion data is collected with high-accuracy, temperature-compensated ultrasonic transducers. This has resulted in more than 200 sets of high-quality experimental data covering various pipe diameters, particle diameters, sand concentrations, and flow velocities. Mining the inherent patterns and physical laws within this data is valuable for guiding future process condition selections and experimental tests. With the existing data covering a wide range of conditions, machine learning methods can be utilized to augment the data, ensuring the newly generated data adheres to physical laws and has similar distributions. Conditional Wasserstein Generative Adversarial Networks (CWGANs) are a class of machine learning models designed for generating conditioned data. CWGANs can be applied to generate additional training data when the available data is limited, significantly expanding the erosion databank for subsequent mechanistic modeling. This work proposes the Conditional GAN Synergized SPPS model (ConGANergy) framework for engineering data augmentation, which utilizes CWGANs to expand erosion experimental datasets. We conducted experiments to evaluate the performance of the proposed framework by comparing the newly generated data against original experimental data and data obtained through a semi-mechanistic erosion prediction model Sand Production Pipe Saver (SPPS) developed at E/CRC. This approach provides engineers and researchers with a powerful tool to expand databases and build robust mechanistic models, leveraging the inherent patterns within high-quality experimental data while adhering to the physical laws governing erosion phenomena.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.