A bolted joint is the most widely used fastener due to advantages such as
low price and easy maintenance. However, bolt looseness caused by
vibration, corrosion, or human error induces serious accidents, including
human injury and damage to facilities and equipment. Recently the safety
issue on the bolt looseness becomes more critical, and lots of researches
have been performing to detect bolt looseness. However, most of the
techniques developed require expensive equipment or advanced
knowledge for bolt looseness detection.
In this study, we proposed a new bolt looseness detection technique
based on a Highly Nonlinear Solitary Waves (HNSW) in the granular chain
composed of a row of spherical particles. This HNSW has a high energy
density in a relatively low-frequency regime. Also, its wave velocity is small
compare to the other elastic waves. These wave characteristics are useful
for non-destructive testing.
We numerically and experimentally investigate the bolt looseness and the
corresponding vibration characteristics based on the HNSW. We develop
the granule sensor and an experimental model. Also, we develop an efficient
numerical model by combining a Discrete Element Model (DEM) for the
granule sensor and a Finite Element Model (FEM) for the bolted joint
structure. Rayleigh damping, a simplified damping model, was applied to
the analysis model to consider the effect of damping of the bolted joint
structure. We confirmed that the reflected wave of the HNSW is sensitive to
the change of the applied torque to the bolt, especially around 0~20 %
torque range compare to the recommended torque(100 %). We identify
that the applied torque changes the effective damping, and the reflected
wave sensitive to the damping changes.
From additional vibration experiments and modal analysis, we confirmed
the change of damping due to joint torque depends on the vibration mode
of the bolt and the joint structures. The damping increases with joint torque
increase in bolt deformation dominant modes, while the damping decrease
with joint torque increase in plate deformation dominant modes. The first
reflected HNSW senses the vibration characteristics of the bolt dominant
modes.
In this study, we validate that the HNSW is useful to detect the blot
looseness by sensing the change of damping in the bolted joint structure.