The Nuclear Regulatory Commission (NRC) regulation 10 CFR Part 50 Appendix S requires consideration of soil-structure interaction (SSI) in nuclear power plant (NPP) analysis and design. ...Soil-structure interaction analysis for NPPs is routinely carried out using guidance provided in the ASCE Standard 4-98 titled “Seismic Analysis of Safety-Related Nuclear Structures and Commentary”. This Standard, which is currently under revision, provides guidance on linear seismic soil-structure-interaction (SSI) analysis of nuclear facilities using deterministic and probabilistic methods. A new appendix has been added to the forthcoming edition of ASCE Standard 4 to provide guidance for time-domain, nonlinear SSI (NLSSI) analysis. Nonlinear SSI analysis will be needed to simulate material nonlinearity in soil and/or structure, static and dynamic soil pressure effects on deeply embedded structures, local soil failure at the foundation-soil interface, nonlinear coupling of soil and pore fluid, uplift or sliding of the foundation, nonlinear effects of gaps between the surrounding soil and the embedded structure and seismic isolation systems, none of which can be addressed explicitly at present.
Appendix B of ASCE Standard 4 provides general guidance for NLSSI analysis but will not provide a methodology for performing the analysis. This paper provides a description of an NLSSI methodology developed for application to nuclear facilities, including NPPs. This methodology is described as series of sequential steps to produce reasonable results using any time-domain numerical code. These steps require some numerical capabilities, such as nonlinear soil constitutive models, which are also described in the paper.
•Linear and nonlinear SSSI analyses of buildings using SASSI and LS-DYNA.•Subjectively compared results with observations from centrifuge experiments.•Simulations and experiments predict minimal SSSI ...effects in global response.•However, adjacent restraint can effect nonlinear footing response considerably.
The influence of structure-soil-structure interaction (SSSI) in low- to medium-rise buildings is investigated through numerical simulations, and observations are compared with those from previous studies that analyzed data from a set of centrifuge experiments of similar models. The buildings include a one-story moment-resisting frame building on spread footings and a two-story shear wall building on a basemat. The numerical simulations are performed using the industry-standard, frequency-domain, linear analysis code SASSI, and the time-domain nonlinear finite-element analysis code, LS-DYNA. In LS-DYNA the simulations are performed with and without geometric nonlinearities (gapping, sliding and uplift) to understand their effects on SSSI. Three plan arrangements of the buildings are considered to characterize the influence of relative location on SSSI: (1) an in-plane SSSI (iSSSI) arrangement, in which the two buildings are placed adjacent to each other along a line parallel to the direction of ground shaking, (2) an anti-plane arrangement (aSSSI), in which the two buildings are placed adjacent to each other along a line perpendicular to the direction of ground shaking, and (3) a combined in-plane-anti-plane (cSSSI) arrangement, in which two shear wall buildings are placed at two adjacent sides of the frame building on footings. Results from the numerical simulations in SASSI and LS-DYNA show that SSSI has negligible effect on the global spectral accelerations of the buildings in these arrangements. The numerical simulations agree with experimental observations in this regard. Numerical investigations into the SSSI response of the frame building on footings placed adjacent to a deep basement show that the presence of the deep basement reduces uplift in the footings and results in smaller peak spectral accelerations at the roof, underscoring the potential importance of geometric nonlinearities (gapping, sliding and uplift) in SSSI and foundation design.
Soil-structure interaction (SSI) analysis is generally a required step in the calculation of seismic demands in nuclear structures, and is currently performed using linear methods in the frequency ...domain. Such methods should result in accurate predictions of response for low-intensity shaking, but their adequacy for extreme shaking that results in highly nonlinear soil, structure or foundation response is unproven. Nonlinear (time-domain) SSI analysis can be employed for these cases, but is rarely performed due to a lack of experience on the part of analysts, engineers and regulators. A nonlinear, time-domain SSI analysis procedure using a commercial finite-element code is described in the paper. It is benchmarked against the frequency-domain code, SASSI, for linear SSI analysis and low intensity earthquake shaking. Nonlinear analysis using the time-domain finite-element code, LS-DYNA, is described and results are compared with those from equivalent-linear analysis in SASSI for high intensity shaking. The equivalent-linear and nonlinear responses are significantly different. For intense shaking, the nonlinear effects, including gapping, sliding and uplift, are greatest in the immediate vicinity of the soil-structure boundary, and these cannot be captured using equivalent-linear techniques.
•Described a method of performing time-domain nonlinear SSI analysis.•Benchmarked this method against the established frequency-domain code, SASSI.•Performed fully nonlinear SSI analyses of two buildings using LS-DYNA.•Compared LS-DYNA results to those calculated using SASSI.•Observed significant differences for cases with highly nonlinear behavior.
•Performed equivalent linear and nonlinear site response analyses using industry-standard numerical programs.•Considered a wide range of sites and input ground motions.•Noted the practical issues ...encountered while using these programs.•Examined differences between the responses calculated from different programs.•Results of biaxial and uniaxial analyses are compared.
Site response analysis is a precursor to soil-structure interaction analysis, which is an essential component in the seismic analysis of safety-related nuclear structures. Output from site response analysis provides input to soil-structure interaction analysis. Current practice in calculating site response for safety-related nuclear applications mainly involves the equivalent linear method in the frequency-domain. Nonlinear time-domain methods are used by some for the assessment of buildings, bridges and petrochemical facilities. Several commercial programs have been developed for site response analysis but none of them have been formally validated for large strains and high frequencies, which are crucial for the performance assessment of safety-related nuclear structures. This study sheds light on the applicability of some industry-standard equivalent linear (SHAKE) and nonlinear (DEEPSOIL and LS-DYNA) programs across a broad range of frequencies, earthquake shaking intensities, and sites ranging from stiff sand to hard rock, all with a focus on application to safety-related nuclear structures. Results show that the equivalent linear method is unable to reproduce the high frequency acceleration response, resulting in almost constant spectral accelerations in the short period range. Analysis using LS-DYNA occasionally results in some unrealistic high frequency acceleration ‘noise’, which can be removed by smoothing the piece-wise linear backbone curve. Analysis using DEEPSOIL results in abrupt variations in the peak strains of consecutive soil layers. These variations are found to be a consequence of the underlying hysteresis rules. There are differences between the site response predictions from equivalent linear and nonlinear programs, especially for large strains and higher frequencies, which are important for nuclear applications. The acceleration predictions from nonlinear programs are reasonably close for most cases, but the peak strain predictions can be significantly different despite using identical backbone curves. Variability in the predictions of different site response analysis programs is significant for large strains and at higher frequencies, underlining the need for the validation of these programs. Biaxial horizontal site response analyses are also performed for the stiff soil site using LS-DYNA. Results from these analyses show that the inclusion of the orthogonal component of the ground motion in site response analysis can significantly influence the acceleration response.
While multifidelity modeling provides a cost-effective way to conduct uncertainty quantification with computationally expensive models, much greater efficiency can be achieved by adaptively deciding ...the number of required high-fidelity (HF) simulations, depending on the type and complexity of the problem and the desired accuracy in the results. We propose a framework for active learning with multifidelity modeling emphasizing the efficient estimation of rare events. Our framework works by fusing a low-fidelity (LF) prediction with an HF-inferred correction, filtering the corrected LF prediction to decide whether to call the high-fidelity model, and for enhanced subsequent accuracy, adapting the correction for the LF prediction after every HF model call. The framework does not make any assumptions as to the LF model type or its correlations with the HF model. In addition, for improved robustness when estimating smaller failure probabilities, we propose using dynamic active learning functions that decide when to call the HF model. We demonstrate our framework using several academic case studies (including some high-dimensional problems) and two finite element model case studies: estimating Navier-Stokes velocities using the Stokes approximation and estimating stresses in a transversely isotropic model subjected to displacements via a coarsely meshed isotropic model. Across these case studies, not only did the proposed framework estimate the failure probabilities accurately, but compared with either Monte Carlo or a standard variance reduction method, it also required only a small fraction of the calls to the HF model.
•Low-fidelity models used to estimate rare events with few high-fidelity model calls.•Active learning using Gaussian Process adaptively decides HF model calls.•Flexibility over LF model choice: reduced physics/DOFs or poorly trained surrogate.•Active learning and Subset Simulation variance reduction method are coupled.•Example demonstrations include Navier-Stokes and solid mechanics problems.
Microreactors present promising opportunities to open new nuclear energy markets. However, it is expected that the economic competitiveness of this new class of reactors will hinge on potential cost ...reductions via mass production. It is therefore critical to begin assessing important considerations for the factory production of microreactors. An overview of the important aspects of the general layout of a microreactor factory, along with best practices to be incorporated early in the design process, is provided in this study. Then, a detailed use case is considered and modeled using a dedicated tool that can map workflows and activities within a factory. The end product is a 242 000 sq. ft. factory model that can ramp up production from 10 to 100 units per year.
Based on the activities and workflows needed, cost estimates for equipment and staffing needs are generated. These are expected to be first-order estimates, but would still provide guidance on the level of investment needed to reach mass production levels of microreactors. Furthermore, the potential cost reductions from scaling production are quantified. It was found that for a 100-unit factory throughput, reductions above 70% per unit cost relative to a prototype demonstration, could be observed for tasks conducted within a factory. These estimates focus solely on component fabricated at a factory and do not account for fuel costs nor any site activities. Because the analysis is design specific, not all findings are expected to be applicable across different microreactors (notably larger varieties), but it still provides a foundation establishing the basis for the mass production of these reactors.
Microreactors present promising opportunities to open new nuclear energy markets. However, it is expected that the economic competitiveness of this new class of reactors will hinge on potential cost ...reductions via mass production. It is therefore critical to begin assessing important considerations for the factory production of microreactors. An overview of the important aspects of the general layout of a microreactor factory, along with best practices to be incorporated early in the design process, is provided in this study. Then, a detailed use case is considered and modeled using a dedicated tool that can map workflows and activities within a factory. The end product is a 242 000 sq. ft. factory model that can ramp up production from 10 to 100 units per year. Based on the activities and workflows needed, cost estimates for equipment and staffing needs are generated. These are expected to be first-order estimates, but would still provide guidance on the level of investment needed to reach mass production levels of microreactors. Furthermore, the potential cost reductions from scaling production are quantified. It was found that for a 100-unit factory throughput, reductions above 70% per unit cost relative to a prototype demonstration, could be observed for tasks conducted within a factory. These estimates focus solely on component fabricated at a factory and do not account for fuel costs nor any site activities. Because the analysis is design specific, not all findings are expected to be applicable across different microreactors (notably larger varieties), but it still provides a foundation establishing the basis for the mass production of these reactors.
Tristructural isotropic (TRISO)-coated particle fuel is a robust nuclear fuel and determining its reliability is critical for the success of advanced nuclear technologies. However, TRISO failure ...probabilities are small and the associated computational models are expensive. We used coupled active learning, multifidelity modeling, and subset simulation to estimate the failure probabilities of TRISO fuels using several 1D and 2D models. With multifidelity modeling, we replaced expensive high-fidelity (HF) model evaluations with information fusion from two low-fidelity (LF) models. For the 1D TRISO models, we considered three multifidelity modeling strategies: only Kriging, Kriging LF prediction plus Kriging correction, and deep neural network (DNN) LF prediction plus Kriging correction. While the results across these multifidelity modeling strategies compared satisfactorily, strategies employing information fusion from two LF models called the HF model least often. Next, for the 2D TRISO model, we considered two multifidelity modeling strategies: DNN LF prediction plus Kriging correction (data-driven) and 1D TRISO LF prediction plus Kriging correction (physics-based). The physics-based strategy, as expected, consistently required the fewest calls to the HF model. However, the data-driven strategy had a lower overall simulation time since the DNN predictions are instantaneous, and the 1D TRISO model requires a non-negligible simulation time.
•TRISO, a robust nuclear fuel, is associated with small failure probabilities.•Active learning with multifidelity modeling to efficiently estimate TRISO failure.•Multifidelity information fusion from two low-fidelity models most efficient.•Physics-based multifidelity strategy reduces calls to high-fidelity model.•Data-driven multifidelity strategy has less overall simulation time.
The acoustic fluid-structure interaction (FSI) formulation is a practical numerical approach for the seismic analysis of fluid-filled tanks. However, there are no verification and validation studies ...reported in the literature that demonstrate the ability of an acoustic FSI numerical model to predict responses important to structural and mechanical design for intense translational and rotational earthquake inputs. Herein, an acoustic FSI formulation is implemented in the open-source Multiphysics Object-Oriented Simulation Environment (MOOSE), and is formally verified and validated using analytical solutions and code-to-code verification, and experimental data, respectively. The analytical solutions are for small amplitude, unidirectional seismic inputs. The code-to-code verification utilizes a previously verified and validated Arbitrary Lagrangian-Eulerian (ALE) numerical model in the commercial finite element code LS-DYNA. The validation studies utilize a comprehensive data set assembled from results of 3D earthquake-simulator tests of a fluid-filled vessel. The acoustic numerical model in MOOSE is verified and validated for hydrodynamic pressures and support reactions except for cases that involve significant convective response. For small amplitude inputs, numerically predicted wave heights match those of the analytical solutions. The numerical model is not verified and validated for wave height calculations under intense 3D seismic inputs. The run times for the acoustic FSI simulations in MOOSE are an order of magnitude, or more, shorter than for the corresponding ALE simulations in LS-DYNA. The utility of the MOOSE acoustic FSI implementation is demonstrated by seismic analysis of a building equipped with a fluid-filled, advanced nuclear reactor.