Accepted Manuscripts

Brantley Mills, Adam Hetzler and Oscar Deng
J. Verif. Valid. Uncert   doi: 10.1115/1.4041837
A thorough code verification effort has been performed on a reduced order, finite element model for 1D fluid flow convectively coupled with a 3D solid, referred to as the 'advective bar' model. The purpose of this effort was to provide confidence in the proper implementation of this model within the SIERRA/Aria thermal response code at Sandia National Laboratories. The method of manufactured solutions is applied so that the order of convergence in error norms for successively refined meshes and timesteps is investigated. Potential pitfalls that can lead to a premature evaluation of the model's implementation are described for this verification approach when applied to this unique model. Through observation of the expected order of convergence, these verification tests provide evidence of proper implementation of the model within the codebase.
TOPICS: Fluid dynamics, Errors, Finite element model
Aytekin Gel, Avinash Vaidheeswaran, Jordan Musser and Charles Tong
J. Verif. Valid. Uncert   doi: 10.1115/1.4041745
Establishing the credibility of computational fluid dynamics (CFD) models for multiphase flow applications is increasingly becoming a mainstream requirement. However, the established Verification and Validation (V&V) Standards have been primarily demonstrated for single phase flow applications. Studies to address their applicability for multiphase flows have been limited. Hence, their application may not be trivial and require a thorough investigation. We propose to adopt the ASME V&V 20 Standard and explore its applicability for multiphase flows through several extensions by introducing several best practices. In the current study, the proposed VVUQ framework is presented and its preliminary application is demonstrated for the simulation of granular discharge through a conical hopper, a commonly employed process in industry. As part of the proposed extensions for the V&V process, a detailed survey of subject matter experts including the CFD modelers and experimentalists was conducted. The results from the survey highlighted the need for a more quantitative assessment of the importance rank in addition to a sensitivity study before embarking on simulation and experimental campaigns. Hence, a screening study followed by a global sensitivity was performed to identify the most influential parameters for the CFD simulation as the first phase of the process, which is reported in the current paper. The results show that particle-particle coefficients of restitution and friction are the most important parameters. The identification of these parameters is important to quantify their effect on the quantities of interest, and improve the confidence level in the numerical predictions.
TOPICS: Simulation, Multiphase flow, Uncertainty quantification, Computational fluid dynamics, Particulate matter, Matter, Flow (Dynamics), Friction
Matteo Diez, Riccardo Broglia, Danilo Durante, Angelo Olivieri, Emilio F. Campana and Frederick Stern
J. Verif. Valid. Uncert   doi: 10.1115/1.4041372
The objective of the present work is the application of uncertainty quantification (UQ) methods for statistical assessment and validation of experimental and computational ship resistance and motions in irregular head waves, using both time series studies and a stochastic regular wave UQ model solved by a metamodel-based Monte Carlo method. Specifically, UQ methods are used for: (1) statistical assessment and validation of experimental and computational modeling of input irregular waves versus analytical benchmark values; (2) statistical assessment of both experimental and computational ship resistance and motions in irregular waves; (3) validation of computational ship resistance and motions in irregular waves versus experimental benchmark values; (4) statistical validation of both experimental and computational stochastic regular wave UQ model for ship resistance and motions versus irregular-wave experimental benchmark values. Methods for problem (1) include Fourier analysis for wave energy spectrum moments, analysis of the auto-covariance matrix and block-bootstrap methods for the uncertainty of wave elevation statistical moments, along with block-bootstrap methods for the uncertainty of mode and distribution. The uncertainty of wave height statistical estimators is evaluated by the bootstrap method. The same methodologies are used to evaluate statistical uncertainties associated to ship resistance and motions in problem (2). Errors and confidence intervals of statistical estimators are used to define validation criteria in problem (3) and (4). The contribution of the present work is the application and integration of UQ methodologies for the solution of problems from (1) to (4). Results are shown for the Delft catamaran.
TOPICS: Waves, Computational fluid dynamics, Ships, Uncertainty quantification, Uncertainty, Computer simulation, Time series, Wave energy, Errors, Fourier analysis, Monte Carlo methods

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In