Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “component degradation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Theoretical Understanding of Effects of Operating Modes on the Performance Durability of Solid Oxide Cells: A Comparison between Potentiostatic and Galvanostatic Operations

In solid oxide cells (SOCs), the choice between galvanostatic (constant current) and potentiostatic (constant voltage) modes does not significantly affect the performance of SOCs as long as the cell components remain unchanged and intact. However, the degradation of cell components, which leads to changes in electrochemical and physical properties of the cell, elevates the importance of the selected operating mode. This paper aims to investigate the effects of galvanostatic and potentiostatic operating modes on the evolving properties of SOCs and their subsequent influence on performance durability. Employing non-equilibrium thermodynamic analysis, a crucial approach for understanding the degradation phenomena within an active electrochemical system, this study aims to provide in-depth insights into how these operating modes affect the longevity and efficacy of SOCs. Key findings include: In cases where oxygen electrode (OE) degradation is accelerated by higher partial pressure of oxygen ( p O 2 ), operating under constant voltage electrolysis can mitigate the high p O 2 at the OE|electrolyte (OE|EL) interface. Conversely, if OE degradation occurs more rapidly under a lower p O 2 , constant current electrolysis is more effective in suppressing degradation by achieving a high p O 2 at the OE|EL interface. For degradation of the fuel electrode (FE) due to higher p O 2 , constant current electrolysis is beneficial for more stable performance, which helps maintain low p O 2 at the FE|EL interface. When FE degradation is accelerated by lower p O 2 , constant voltage electrolysis can avert low p O 2 at the FE|EL interface. In practical scenarios, more complex degradation mechanisms come into play, especially when p O 2 significantly deviates from initial conditions. Degradation in one electrode can influence p O 2 in the other electrode, a phenomenon more pronounced in potentiostatic than in galvanostatic electrolysis.

Electrochemistry↗
DOE OSTI2023

Conclusions from 3 years of continuous capture plant operation without exchange of the AMP/PZ-based solvent at Niederaussem – insights into solvent degradation management

A many times heard mantra of solvent degradation management in amine-based post combustion capture is “keep the solvent clean” to minimize solvent consumption. It is assumed that the amine losses would decrease by the removal of metals, degradation products, and reactive trace components which are captured from the flue gas, like NO 2 (as potentially driving components of the amine degradation besides dissolved O 2 ). However, this theoretical hypothesis – based on results from laboratory experiments typically generated with fresh amines – disregards the complexity of the solvent matrix, interaction of potential metal catalysts with degradation products and oxidizing agents, and specific chemical requirements which must be fulfilled before a degradation mechanism can proceed. Degradation of the solvent CESAR1 (aqueous solution of 3.0 molar 2-amino-2-methylpropan-1-ol (AMP) and 1.5 molar piperazine (PZ)) is investigated in a unique long-time test campaign (testing time up to now 40 months; 24/7 operation) without replacement of the solvent inventory at the capture pilot plant at the lignite-fired power plant in Niederaussem. Three solvent management strategies with different effect mechanisms are investigated and evaluated: (a) removal of only anionic compounds and trace elements (within 75 days solvent inventory treated two times) and anionic as well as cationic compounds and trace elements (114 days, inventory treated four times) from the solvent by ion exchange, (b) adsorptive removal of trace elements from the solvent by active carbon in 35% of the operating time, and (c) removal of >80% NO 2 by flue gas pretreatment with thiosulfate/sulfite solution (dosing for 2,000 h). The results of the testing program clearly show that “solvent cleanliness” is not a well-defined parameter and that results from laboratory tests, tests without fully representative industrial flue gasses, and short-term testing of monoethanolamine cannot be generalized for other solvents and industrial application. Furthermore, these results showcase that specific degradation management considering solvent, capture plant and flue gas quality is reasonable. Overshooting efforts for solvent management are contra-productive and produce unnecessary waste streams, efficiency losses and costs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Development of a high-fidelity multi-cycle model of the NuScale small modular reactor using VERA

With growing renewables penetration, there is increased interest in flexible power operation for nuclear reactors. For multi-unit SMRs, in particular the NuScale SMR, which is an integral pressurized water reactor, there are opportunities to optimize flexible power operation across multiple units to limit the degradation of structural and control components. Here, we focus on degradation of in-core components, specifically the control rods and reactor pressure vessel. To perform these studies a high-fidelity, multi-cycle representation of the NuScale SMR is required, with a detailed representation of the structural and control components. To this end, the NuScale SMR has been modelled using the Virtual Environment for Reactor Applications (VERA) software. The entire transition to equilibrium is simulated, from Cycle 1 through to the equilibrium cycle. The equilibrium cycle model shows a good agreement with the NuScale design certification application (DCA) results, with differences attributable to a combination of using public domain data for the present study, and methodological differences. K-effective, power distributions, reactivity coefficients, and boron letdown curves are compared and all found to closely match. This shows that the VERA model is suitable for further studies. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2022

Session 4: ASME Code Rules

General review of code rules What do they cover. What they don’t. Underlying assumptions Probability of failure Material property parameters Stresses and loads Degradation What is NonmetallicsWork Group (NWG) doing? Failure in components Redefining failure Degradation rules Oxidation degradation Irradiation degradation Molten Salt Issues Abrasion/Erosion

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2021

Using In Situ High-Energy X-ray Diffraction to Quantify Electrode Behavior of Li-Ion Batteries from Extreme Fast Charging

Extreme fast charging (XFC, =15 min charging time) of Li-ion batteries (LIBs) has been proposed as an immediate target to increase the commercial appeal of electric vehicles. However, XFC of LIBs is associated with the degradation of battery performance and safety concerns. Quantitative and simultaneous characterization of various components during cell degradation represents a major experimental challenge. In this work, we outline a methodology for the use of spatially resolved, high-energy X-ray diffraction as a quantitative, in situ method of mapping the degradation of LIBs. We use this approach to study the battery cell capacity loss, both locally (mm scale) and globally over the entire cell (cm scale). Specifically, our workflow allows us to quantify the total amount of plated Li on the anode, as well as its spatial correlation to the structural properties of the anode and cathode. The method complements existing optical methods to resolve the spatial heterogeneity of local degradation mechanisms such as Li plating and provides simultaneous insights into concomitant anode state-of-charge variability. We apply it to commercially relevant single-layer pouch cells with the graphite anode and the LiNi 0.5 Mn 0.3 Co 0.2 O 2 cathode. Our results show that Li plating occurs heterogeneously on the graphite anode and that it is spatially correlated to the extent of anode lithiation. In this work, we anticipate that the described workflow will allow for understanding multiscale degradation in energy-storage devices beyond LIBs, where quantitative analysis at a local and global length scale can be performed without the necessity to tear down the device, due to the applicability of high-energy X-rays to probe in situ degradation.

25 ENERGY STORAGE↗
DOE OSTI2023

Survey of Aging and Monitoring Concerns for Cables and Splices Due to Cable Repair and Replacement

The purpose of this report is to survey aging and monitoring concerns for electrical cable splices in nuclear power plants (NPPs) in long term operation. As portions of existing electrical cable runs in nuclear are replaced over time due to localized events, the total number of splices in NPPs is expected to increase. Relative to cables, the body of knowledge regarding aging of splices and splices in combination with aging cables in nuclear service environments in long-term operations is low. A few reports have considered the aging of cable system components other than cables (Jacobus 1990; Nelson 1998; Villaran and Lofaro 2002), but the nuclear industry has two decades of operating experience since these were published to further enlighten this issue. Herein we discuss electrical cables and splices commonly found in U.S. nuclear power plants, their qualification in safety-related application, and methods for monitoring their health condition. Common environmental stresses that can give rise to cable and splice failure are discussed. The Nuclear Regulatory Commission (NRC) Licensee Event Reports (LER) database was used to identify documented issues of cable and splice failure. The trend in the resultant data over time is considered to see if failures are increasing as plants age. Observations and conclusions of this work include: 1. Cables and splices are highly reliable components. Occurrence rates for events of interest were low and nearly constant over the last 20 years. 2. Common-cause failure for evaluated cable events of interest was observed to primarily be associated with loose connections, which may manifest associated with workmanship issues, thermal cycling, and/or vibration. 3. Replacement of cables is more common than repair, leading to an increase in proportion of new generation cables in the plant over time. 4. Splices on degraded cables have been observed to be problematic. Due to aging NPP infrastructure, including electrical cables, it is expected that such issues will continue to increase. 5. Condition monitoring approaches, while shown to be fruitful for cables, have been shown to be insensitive to degradation of splice sleeves, which are critical to the continued performance of splices. Additional condition monitoring (CM) work is needed to evaluate methods which are sensitive to the degradation of splice components. 6. Extended Material Degradation Assessment (EMDA) knowledge gaps for electrical cables (Bernstein et al. 2014) have not been investigated for splices but may represent similar concerns such as for the accelerated aging process historically used in environmental qualification.

42 ENGINEERING↗
DOE OSTI2023

Microstructure and in-service degradation of baffle former bolts – in-core components of light water reactors

This report presents a microstructure evaluation of the baffle former bolt, a commercial pressurized water reactor component. Using SEM, EDS, and EBSD analysis the material was identified as AISI 316 steel with an annealed austenite microstructure. Findings include abnormal grain growth in a fraction of the grains and the absence of retained ferrite. Notable observations include an overlapping of pre- and post-irradiation deformation, including defect-free channel formation as an active deformation mechanism in the near-surface layer of BFB. Specimen surfaces, exposed to high-temperature, high-pressure water, exhibited signs of in-service corrosion degradation. EBSD and EDS analyses highlighted intergranular corrosion, possible grain boundary oxidation with depths less than 3 µm, and, unexpectedly, short cracks filled with Cr-rich oxides measuring around 5-6 µm. The presence of defect-free channels from EBSD data suggests episodes of high mechanical stresses during service. The specimens provide a potentially unique insight into the in-service degradation of NPP components, strain localization, and crack initiation processes. An intact corrosion layer reveals a complex component's in-service history. Further research is underway for a detailed analysis.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2026

Explaining System-Level Prognostics with Established Machine Learning Methods

System-level prognostics is crucial for ensuring reliability and enabling predictive maintenance in complex systems with interconnected components. This study presents a framework that integrates data-driven methods to predict the remaining useful life (RUL) of a subsystem under multiple and concurrent faults within a nuclear power plant system with explainable artificial intelligence (XAI). A nuclear power plant (NPP) operation was simulated to model the degradation behavior of NPP components, and four machine learning models—Gradient Boosting Regressor (GBR), Support Vector Regressor (SVR), Fully Connected Neural Network (FCNN), and Long Short-Term Memory (LSTM)—were evaluated for prognostics with a novel system RUL parameter. The LSTM model demonstrated potential superior repeatability, while SHAP (SHapley Additive exPlanations) for explainability provided consistent and trustworthy global explanations. In contrast, LIME (Local Interpretable Model-agnostic Explanations) offered localized interpretability but showed reduced stability for sequential data. Key findings include the interplay between component-level degradation and system-wide performance, with LSTM effectively capturing these dynamics through sequence-level predictions. The XAI techniques enhanced transparency by identifying critical features influencing model predictions and aligning with domain knowledge. Furthermore, this framework has significant implications for improving trust and understanding in predictive maintenance, particularly in safety-critical industries like nuclear energy.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2022

Bridging Equipment Reliability Data and Robust Decisions in a Plant Operation Context

In order to reduce operation and maintenance (O&M) costs, nuclear power plants (NPPs) are moving from corrective and periodic maintenance to predictive maintenance strategies. Such transition requires changes on the data that needs to be retrieved and on the type of decision processes to be employed. Advanced monitoring and data analysis technologies are essential to support predictive strategies. They can in fact provide precise information about health of a component, track its degradation trends, and provide information of its expected failure time. With such information, maintenance operations for a component can be performed right before its expected failure time. This dynamic context of O&M operations requires new methods to analyze data, propagate component health information from the component to the system level, and optimize plant resources. In this respect, the risk informed asset management (RIAM) project has been tasked to develop and test this new class of methods into a risk analytics toolset. This toolset consists of data analytics tools coupled with reliability methods designed to manage plant assets and performances in a predictive maintenance context. This report shows the latest improvements on such development and the initial testing of our methods on the three main research areas that the RIAM project is focusing on. These areas are the following: equipment reliability data analytics, system reliability modeling, and plant resources optimization methods. We show how the methods developed in these areas can support predictive maintenance strategies by: 1) analyzing equipment reliability data (either in numeric and textual form), 2) assessing component and system health through an innovative margin-based reliability approach, and 3) identifying the most critical components and set optimal maintenance schedule based on plant economic and operational constraints.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Modeling Graphite Degradation (Oxidation and Irradiation)

Presentation Outline: Background and motivation Component design (ASME rules) Graphite degradation Oxidation Irradiation MOOSE modeling Oxidation modeling Graphite component stress modeling Oxidized conditions Irradiated conditions Future work

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Experimental and Analytical Verification of ASME Section IiII Division 5 Creep-Fatigue Design Rules

The continuous advancement of structural materials and the growing demands for more reliable and economical structural components in high-temperature reactor applications have necessitated the development of comprehensive design methodologies and design rules. Mechanical degradation of structural components at elevated temperatures subjected to cyclic deformation is controlled by the creep-fatigue damage. Over the past few decades, diligent research efforts have been dedicated to refining the development of elevated temperature design rules in the American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code (BPVC), Section III, Division 5 and to develop conservative design rules that can effectively guard against the risk of creep-fatigue failure. In ASME Section III, Division 5, for a design to pass the creep-fatigue acceptance criteria, creep damage and fatigue damage are evaluated separately, and these damages must not violate the bi-linear creep-fatigue interaction diagram, i.e., the so-called D-diagram. The creep-fatigue damage evaluation procedure assumes that the effects of the actual cyclic loading sequence can be bounded by assuming that the individual loading cycles are uniformly distributed throughout the component design life. In this study, creep-fatigue experiments with variable amplitudes and loading sequencies were designed and performed on Alloy 617 at high temperatures. The results were analyzed to evaluate the loading history effect on creep-fatigue damage accumulation and to verify the assumptions for the creep-fatigue evaluation design rules.

36 - MATERIALS SCIENCE↗
DOE OSTI2026

Elevated-Temperature Tribo-Corrosion Response of Eutectic High-Entropy Alloy

The combination of elevated temperature and tribo-corrosion leads to the accelerated degradation of structural components used in many extreme environments. Recently developed high-entropy alloys (HEAs) with multiple principal elements have the potential for superior degradation resistance compared with presently used structural alloys. Here, we demonstrate the microstructural stability, pitting resistance, and superior tribo-corrosion degradation resistance of the AlCoCrFeNi 2.1 eutectic HEA in comparison with duplex stainless steel 2205 in deionized water (controlled low-ionic-strength electrolyte) at 25 °C, 50 °C and 100 °C. The AlCoCrFeNi 2.1 HEA showed excellent microstructural stability and tribo-corrosion resistance at all three temperatures, an order-of-magnitude lower wear rate, and a lower coefficient of friction compared with duplex 2205 steel. The lowest wear volume loss and wear rate for both AlCoCrFeNi 2.1 and duplex steel were recorded at 50 °C, which was attributed to temperature-assisted passivation and formation of a comparatively stable tribological surface condition. These results suggest superior performance of eutectic HEAs in tribo-corrosion applications compared with currently used dual-phase steels and motivate future evaluation in ion-containing industrial water chemistries.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Phase stability and tensorial thermal expansion properties of single to high-entropy rare-earth disilicates

Temperature limitations in nickel-base superalloys have resulted in the emergence of SiC-based ceramic matrix composites as a viable replacement for gas turbine components in aviation applications. Higher operating temperatures allow for reduced fuel consumption but present a materials design challenge related to environmental degradation. Rare-earth disilicates (RE 2 Si 2 O 7 ) have been identified as coatings that can function as environmental barriers and minimize hot component degradation. In this work, single- and multiple-component rare-earth disilicate powders were synthesized via a sol-gel method with compositions selected to exist in the monoclinic C 2/m phase (β phase). Phase stability in multiple cation compositions was shown to follow a rule of mixtures and the C 2/m phase could be realized for compositions that contained up to 25% dysprosium, which typically only exists in a triclinic, P $\bar1$, phase. All compositions exhibited phase stability from room temperature to 1200°C as assessed by X-ray diffraction. The thermal expansion tensors for each composition were determined from high-temperature synchrotron X-ray diffraction and accompanying Rietveld refinements. It was observed that ytterbium-containing compositions had larger changes in the α 31 shear component with increasing temperature that led to a rotation of the principal axes. Principal axes rotation of up to 47° were observed for ytterbium disilicate. The results suggest that microstructure design and crystallographic texture may be essential future avenues of investigation to ensure thermo-mechanical robustness of rare-earth disilicate environmental barrier coatings.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Anion-Exchange-Membrane Electrolysis with Alkali-Free Water Feed

Hydrogen is a green and sustainable energy vector that can facilitate the large-scale integration of intermittent renewable energy, renewable fuels for heavy transport, and deep decarbonization of hard-to-abate industries. Anion-exchange-membrane water electrolyzers (AEM-WEs) have several achieved or expected competitive advantages over other electrolysis technologies, including the use of precious metal-free electrocatalysts at both electrodes, fluorine-free hydrocarbon-based ionomeric membranes and bipolar plates based on inexpensive materials. Contrasting the analogous proton-exchange-membrane system (PEM-WE), where pure water is circulated (no support electrolyte), the current generation of AEM-WEs necessitates the circulation of a dilute aqueous alkaline electrolyte for reaching high energy efficiency and durability. For several reasons, including but not limited to lower cost of balance-of-plant, lower operating cost and improved device's lifetime, achieving high cell efficiency and performance using an alkali-free water feed is highly desirable. In this review, we develop and build a foundational understanding of AEM-WEs operating with pure water, as well as discuss the effects of operating with natural water feeds like seawater. After a discussion of the possible advantages of pure-water-fed AEM-WEs, we cover the thermodynamic and kinetic processes involved in AEM-WE, followed by a detailed review of materials and components and their integration in the device. We highlight the influence of electrolyte composition and alkali/electrolyte-free feed on the membrane-electrode assembly, ionomers, electrocatalysts, porous transport layer, bipolar plates and operating configuration. We provide evidence for how the pure water feed engenders several issues related to the degradation of device components and propose mitigation strategies.

Electrodes↗
DOE OSTI2021

Material Interactions in Severe Accidents – Benchmarking the MELCOR V2.2 Eutectics Model for a BWR-3 MARK-I Station Blackout: Part I – Single Case Analysis

Here in this analysis, the two material interaction models available in the MELCOR code are benchmarked for a severe accident at a BWR under representative Fukushima Daiichi boundary conditions. This part of the benchmark investigates the impact of each material interaction model on accident progression through a detailed single case analysis. It is found that the eutectics model simulation exhibits more rapid accident progression for the duration of the accident. The slower accident progression exhibited by the interactive materials model simulation, however, allows for a greater degree of core material oxidation and hydrogen generation to occur, as well as elevated core temperatures during the ex-vessel accident phase. The eutectics model simulation exhibits more significant degradation of core components during the late in-vessel accident phase – more debris forms and relocates to the lower plenum before lower head failure. The larger debris bed observed in the eutectics model simulation also reaches higher temperatures, presenting a more significant thermal challenge to the lower head until its failure. At the end of the simulated accident scenario, however, core damage is comparable between both simulations due to significant core degradation that occurs during the ex-vessel phase in the interactive materials model simulation. A key difference between the two models’ performance is the maximum temperatures that can be reached in the core and therefore the maximum ΔT between any two components. When implementing the interactive materials model, users have the option to modify the liquefaction temperature of the ZrO 2 -interactive and UO 2 -interactive materials as a way to mimic early fuel rod failure due to material interactions. Through modification of the liquefaction of high melting point materials with significant mass, users may inadvertently limit maximum core temperatures for fuel, cladding, and debris components.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2023

Root Cause Analysis of Degradation in Protonic Ceramic Electrochemical Cell with Interfacial Electrical Sensors Using Data‐Driven Machine Learning

Abstract Protonic ceramic electrochemical cells (PCECs) offer promising paths for energy storage and conversion. Despite considerable achievements made, PCECs still face challenges such as physiochemical compatibility between componenets and suboptimal solid–solid contact at the interfaces between the electrolytes and electrodes. In this study, a novel approach is proposed that combines in situ electrochemical characterization of interfacial electrical sensor embedded PCECs and machine learning to quantify the contributions of different cell components to total degradation, as well as to predict the remaining useful life. The experimental results suggest that the overpotential induced by the oxygen electrode is 48% less than that of oxygen electrode/electrolyte interfacial contact for up to 1171 h. The data‐driven machine learning simulation predicts the RUL of up to 2132 h. The root cause of degradation is overpotential increase induced by oxygen electrode, which accounts for 82.9% of total cell degradation. The success of the failure diagnostic model is demonstrated by its consistency with degradation modes that do not manifest in electrolysis fade during early real operations. This synergistic approach provides valuable insights into practical failure diagnosis of PCECs and has the potential to revolutionize their development by enabling improved performance prediction and material selection for enhanced durability and efficiency.

25 ENERGY STORAGE↗
DOE OSTI2022

Research on oxygen solubility in aqueous amine solvents with common additives used for CO 2 chemical absorption

Oxidative degradation in amine based post-combustion carbon capture (PCCC) leads to solvent losses and decreased solvent performance. Oxygen (O 2 ) in the flue gas is dissolved into the aqueous solvent and can reactively degrade the amine component. Advanced aqueous amine solvents are designed to contain additional surfactant and anti-foam additives as well as two types of corrosion inhibitors. These additives contribute to improve the solvent performance while reducing foaming and corrosion of expensive equipment. This work examines the impact of these additives on oxygen solubility (or dissolved oxygen - DO) in various aqueous amines used in PCCC including ethanolamine (MEA), 1-amino-2-propanol (A2P), 2-amino-1-propanol (2A1P), 2-(methylamino)-ethanol (NMEA), 2-(ethylamine)-ethanol (2EAE), diethanolamine (DEA), methyl-diethanolamine (MDEA), dimethylethanolamine (DMEA), piperazine (PZ), 1,2-ethyldiamine (EDA), and 1,6 hexadiamine (HDA). Further, the impact of carbon loading on oxygen solubility with and without these additives was also examined. The surfactant and anti-foam additives had minimal impact on DO for all the solvents investigated. The two anticorrosion additives had opposite effects in the measured DO, 2- mercaptobenzothiazole (MBT) decreased the DO concentration, and sodium metavanadate showed an increase. These results indicate that several of these common additives can be expected to minimally affect DO related oxidative degradation in these advance amine solvents.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Reinforcement learning for adaptive maintenance policy optimization under imperfect knowledge of the system degradation model and partial observability of system states

Maintenance policy optimization usually is faced with challenges that arise from an imperfect knowledge of system degradation models and from the partial observability of system degradation states. Here, this paper proposes a reinforcement learning method to address these two challenges for a class of maintenance problems with Markov degradation processes. The reinforcement learning approach consists of a learning component and a planning component. Using sequentially collected observations, at each step of decision-making the learning component improves the knowledge of system degradation in terms of the probability distributions of the transition rates based on sequential Bayesian inference. Using the updated transition rates, at each step of decision-making the maintenance policy optimization problem is then formulated as a partially observable Markov decision problem, and the planning component computes the optimal maintenance policy that maximizes the expected cumulative reward. The proposed method is illustrated using a numerical example with repair and inspection maintenance actions. The result shows that as more observations are collected, the learning component progressively learns the true system degradation process, and the planning component adjusts the optimal maintenance policy accordingly as well, which leads to increased reward.

42 ENGINEERING↗