Skip to main content
SHARE
Publication

Implementing bubbly mercury material model (R-P model) in the pulse simulation to predict strain on the Spallation Neutron Source target vessel with gas injection

by Hao Jiang, Drew E Winder, Charlotte N Barbier, Bernard W Riemer
Publication Type
Journal
Journal Name
Results in Physics
Publication Date
Page Numbers
106511 to 106511
Volume
49
Issue
49

The Spallation Neutron Source (SNS) creates neutrons by striking a liquid mercury target with short pulses of high-energy protons. SNS targets have been operated with helium microbubble gas injection into portions of the flowing mercury since 2017. The measurements during in situ testing showed that the strain response of the stainless steel target vessel significantly decreased with gas injection. Strain reductions ranged from 40% to 75% for targets operating with gas-injection rates between 2 and 3 standard liters per minute. These strain reductions have allowed SNS to operate reliably at 1.4 MW over the last few years and are expected to significantly improve the fatigue life of the target vessel. The achieved mitigation of the pulse response and the associated cavitation damage with gas injection is a cornerstone of the upgrade basis to the redesigned mercury vessel for planned higher power operation.

Existing methods can simulate the structural response of the target operated without gas injection. Developing simulation techniques that account for the benefits of gas injection adds a considerable challenge. A combined mercury/bubble material model based on the Rayleigh–Plesset equation was developed to improve simulation of the response of a structure containing liquid and gas by incorporating bubble growth volume feedback. This paper details the implementation of the mercury/bubble material model in the pulse simulation of an existing target design. The newly developed simulation technique combines the current no-gas mercury material model for no-gas regions and the bubbly mercury material model for the regions with gas bubbles. The results of the strain response simulation using the new techniques are promising. The challenges of the technique development are also discussed. For example, the bubble size distributions are crucial for the simulations but remain an area of active research.