Aditya Venkatraman, Ryan Michael Katona, Demitri Maestas, Matthew Roop, Philip Noell, David Montes de Oca Zapiain
{"title":"An active learning framework for the rapid assessment of galvanic corrosion","authors":"Aditya Venkatraman, Ryan Michael Katona, Demitri Maestas, Matthew Roop, Philip Noell, David Montes de Oca Zapiain","doi":"10.1038/s41529-024-00476-4","DOIUrl":null,"url":null,"abstract":"The current present in a galvanic couple can define its resistance or susceptibility to corrosion. However, as the current is dependent upon environmental, material, and geometrical parameters it is experimentally costly to measure. To reduce these costs, Finite Element (FE) simulations can be used to assess the cathodic current but also require experimental inputs to define boundary conditions. Due to these challenges, it is crucial to accelerate predictions and accurately predict the current output for different environments and geometries representative of in-service conditions. Machine learned surrogate models provides a means to accelerate corrosion predictions. However, a one-time cost is incurred in procuring the simulation and experimental dataset necessary to calibrate the surrogate model. Therefore, an active learning protocol is developed through calibration of a low-cost surrogate model for the cathodic current of an exemplar galvanic couple (AA7075-SS304) as a function of environmental and geometric parameters. The surrogate model is calibrated on a dataset of FE simulations, and calculates an acquisition function that identifies specific additional inputs with the maximum potential to improve the current predictions. This is accomplished through a staggered workflow that not only improves and refines prediction, but identifies the points at which the most information is gained, thus enabling expansion to a larger parameter space. The protocols developed and demonstrated in this work provide a powerful tool for screening various forms of corrosion under in-service conditions.","PeriodicalId":19270,"journal":{"name":"npj Materials Degradation","volume":null,"pages":null},"PeriodicalIF":6.6000,"publicationDate":"2024-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s41529-024-00476-4.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"npj Materials Degradation","FirstCategoryId":"88","ListUrlMain":"https://www.nature.com/articles/s41529-024-00476-4","RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
The current present in a galvanic couple can define its resistance or susceptibility to corrosion. However, as the current is dependent upon environmental, material, and geometrical parameters it is experimentally costly to measure. To reduce these costs, Finite Element (FE) simulations can be used to assess the cathodic current but also require experimental inputs to define boundary conditions. Due to these challenges, it is crucial to accelerate predictions and accurately predict the current output for different environments and geometries representative of in-service conditions. Machine learned surrogate models provides a means to accelerate corrosion predictions. However, a one-time cost is incurred in procuring the simulation and experimental dataset necessary to calibrate the surrogate model. Therefore, an active learning protocol is developed through calibration of a low-cost surrogate model for the cathodic current of an exemplar galvanic couple (AA7075-SS304) as a function of environmental and geometric parameters. The surrogate model is calibrated on a dataset of FE simulations, and calculates an acquisition function that identifies specific additional inputs with the maximum potential to improve the current predictions. This is accomplished through a staggered workflow that not only improves and refines prediction, but identifies the points at which the most information is gained, thus enabling expansion to a larger parameter space. The protocols developed and demonstrated in this work provide a powerful tool for screening various forms of corrosion under in-service conditions.
电偶中的电流可以确定其电阻或易腐蚀性。然而,由于电流取决于环境、材料和几何参数,因此测量的实验成本很高。为了降低成本,可以使用有限元(FE)模拟来评估阴极电流,但也需要实验输入来定义边界条件。由于这些挑战,加速预测并准确预测不同环境下的电流输出以及代表在役条件的几何形状至关重要。机器学习代用模型为加速腐蚀预测提供了一种方法。然而,购买校准代用模型所需的模拟和实验数据集会产生一次性成本。因此,我们开发了一种主动学习协议,通过校准低成本的代理模型,将示范电偶(AA7075-SS304)的阴极电流作为环境和几何参数的函数。代用模型在 FE 模拟数据集上进行校准,并计算出一个获取函数,该函数可识别出具有最大潜力改进电流预测的特定额外输入。这是通过一个交错的工作流程实现的,该流程不仅能改进和完善预测,还能确定获得最多信息的点,从而扩展到更大的参数空间。这项工作中开发和演示的规程为在役条件下筛查各种形式的腐蚀提供了一个强大的工具。
期刊介绍:
npj Materials Degradation considers basic and applied research that explores all aspects of the degradation of metallic and non-metallic materials. The journal broadly defines ‘materials degradation’ as a reduction in the ability of a material to perform its task in-service as a result of environmental exposure.
The journal covers a broad range of topics including but not limited to:
-Degradation of metals, glasses, minerals, polymers, ceramics, cements and composites in natural and engineered environments, as a result of various stimuli
-Computational and experimental studies of degradation mechanisms and kinetics
-Characterization of degradation by traditional and emerging techniques
-New approaches and technologies for enhancing resistance to degradation
-Inspection and monitoring techniques for materials in-service, such as sensing technologies