0
Research Papers

# Validation Facility and Model Development for Nuclear Fuel Assembly Response to Seismic Loading

[+] Author and Article Information
Noah A. Weichselbaum

Mechanical and Aerospace Engineering,
The George Washington University,
Washington, DC 20052
e-mail: weichselbaum@gwu.edu

Morteza Rahimi Abkenar

Civil and Environmental Engineering,
The George Washington University,
Washington, DC 20052
e-mail: rahimi_m@gwu.edu

Marcos Vanella

Mechanical and Aerospace Engineering,
The George Washington University,
Washington, DC 20052

Majid T. Manzari

Civil and Environmental Engineering,
The George Washington University,
Washington, DC 20052
e-mail: manzari@gwu.edu

Elias Balaras

Mechanical and Aerospace Engineering,
The George Washington University,
Washington, DC 20052
e-mail: balaras@gwu.edu

Philippe M. Bardet

Mechanical and Aerospace Engineering,
The George Washington University,
Washington, DC 20052
e-mail: bardet@gwu.edu

Manuscript received January 31, 2015; final manuscript received June 30, 2015; published online September 3, 2015. Assoc. Editor: Jovica R. Riznic.

ASME J of Nuclear Rad Sci 1(4), 041005 (Sep 16, 2015) (11 pages) Paper No: NERS-15-1013; doi: 10.1115/1.4031031 History: Received January 31, 2015; Accepted July 07, 2015; Online September 03, 2015

## Abstract

A joint experimental and numerical campaign is conducted to provide validation dataset of high-fidelity fluid–structure interaction (FSI) models of nuclear fuel assemblies during seismic loading. A refractive index-matched (RIM) flow loop is operated on a six-degree-of-freedom shake table and instrumented with nonintrusive optical diagnostics. The test section can house up to three full-height fuel assemblies. To guarantee reproducible and controlled initial conditions, special care is given to the test section inlet plenum; in particular, it is designed to minimize secondary pulsatile flow that may arise due to ground acceleration. A single transparent surrogate $6×6$ fuel subassembly is used near prototypical Reynolds number, $Re=105$ based on hydraulic diameter. To preserve dynamic similarity of the model with prototype, the main dimensionless parameters are matched and custom spacer grids are designed. Special instruments are developed to characterize fluid and structure response and to operate in this challenging shaking environment. In parallel to the earlier experiments, we also conducted fully coupled direct numerical simulations, where the equations for the fluid and the structure are simultaneously advanced in time using a partitioned scheme. To deal with the highly complex geometrical configuration, which also involves large displacements and deformations, we utilize a second-order accurate, immersed boundary (IB) formulation, where the geometry is immersed in a block-structured grid with adaptive mesh refinement (AMR). To explore a wide parametric range, we will consider several subsets of the experimental configuration. A typical computation involves 60,000 cores, on leadership high-performance computing facilities (i.e., IBM Blue-Gene Q–MIRA).

<>

## Figures

Fig. 1

Layout of experimental facility. Test section is fixed on the shake table with support structure

Fig. 2

Spacer grid design

Fig. 3

Acrylic rod in pure para-cymene (top), acrylic rod in para-cymene/cinnamic aldehyde solution (bottom)

Fig. 4

Full structural model

Fig. 5

Ground motion records used in the seismic analyses: top: El Centro; bottom: Santa Monica

Fig. 6

Computed mode shapes. Left: first three modes of prototypical bundle; center: first three modes of surrogate bundle; right: first mode of full facility

Fig. 7

Displacements (in mm) of fuel rods under sinusoidal motion for case α

Fig. 8

Displacement (in mm) of the structure under El Centro record for cases α and C with 5% of damping

Fig. 9

Schematic of fuel rod subjected to seismic ground acceleration aB(t)

Fig. 10

Eulerian and Lagrangian grids in two dimensions. A Lagrangian marker is related to Eulerian stencil on each velocity component

Fig. 11

Schemes for consistent forcing when surfaces span different grid-blocks, Lagrangian IB forcing: (a) Using duplicated marker particles (virtual particles), assigned to the guard-cell regions of the block (mvl, mvl+1 for block A). (b) Using inverse guard-cell filling after forcing by markers within the domain of each block, and subsequent addition of volume forces on each block boundaries

Fig. 12

Computational setup for 2×2 fuel rod assembly. The uniform Eulerian grid (fluid grid) is split on a processor grid distributed on the y–z axes (rectangular blocks). The rods are split on an arbitrary number of segments and distributed among processors

Fig. 13

Computational setup for 2×2 spacer grid + fuel rod assembly. Both spacer grid and fuel rods’ surfaces are triangulated, split in a specified amount of bodies and their boundary condition is treated using immersed boundaries

Fig. 14

2×2 spacer grid + fuel rod assembly: Instantaneous flow field for nondimensional time t*=39.0 for direct simulation. Vorticity in span-wise y-direction is shown on different slices around two rods and the spacer grid assembly

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections