Verification and validation for large eddy simulation of the turbulent flow around an underwater entity

Abstract

The objective is to analyze the errors and uncertainty in the simulation results of the complex, unsteady turbulent flow and vortices. The implicitly filtered large eddy simulation (LES) with the boundary data immersion method is used to investigate the transient turbulent flow around a fully appended underwater entity model (SUBOFF) proposed by Groves et al. [Geometric Characteristics of DARPA Suboff Models: (DTRC Model Nos. 5470 and 5471) (David Taylor Research Center, 1989)] with emphasis on verification and validation. As for the verification, the five-equation method is used to assess the LES error, including the modeling error and numerical error in the transient flow. They offset each other, leading to a small total error. As for the validation, it has been achieved in the most area when the experimental result is located at the band of ysim ± UV (ysim is the simulation result, and UV is the validation uncertainty). There is a strong relationship between the validation uncertainty and the complex vortex interaction. The validation uncertainty becomes significant, which means less accuracy of the simulation result, within the tip flow region, adverse pressure gradient region, tip vortex interacting with the boundary layer region and shear layer region. Most all of these regions occur behind the appendage and at the shape changing position of the SUBOFF.